3YNCHRONOUS Synchronous 4RANSMISSION Transmission …FILE/ABB+SDH+booklet+5-2+final+2008.pdf · 3YNCHRONOUS 4RANSMISSION 3YSTEMS 3$( Synchronous Transmission Systems (SDH) A guide
Post on 15-Feb-2018
219 Views
Preview:
Transcript
SynchronousTransmission
Systems (SDH)
A guide to the SDH world
Foreword
Synchronous Transmission Systems
Synchronous TransmissionSystems
Nortel Networks
Issue 5.0 July 1999
' 1999 Nortel Networks References:
Diagrams and figures extracted from ITU-T Recommendations arereproduced with the kind permission of the ITU TelecommunicationsStandardisation Sector.
ii
ABB Switzerland Ltd.Edition 5.2, 2008
Foreword
Synchronous Transmission Systems
Synchronous TransmissionSystems
Nortel Networks
Issue 5.0 July 1999
' 1999 Nortel Networks References:
Diagrams and figures extracted from ITU-T Recommendations arereproduced with the kind permission of the ITU TelecommunicationsStandardisation Sector.
iiForeword
Synchronous Transmission Systems
Synchronous TransmissionSystems
Nortel Networks
Issue 5.0 July 1999
' 1999 Nortel Networks References:
Diagrams and figures extracted from ITU-T Recommendations arereproduced with the kind permission of the ITU TelecommunicationsStandardisation Sector.
ii
Synchronous Transmission Systems
iii Foreword
Foreword This book brings together a wide range of ideas and information to act as both an educational tool and a useful reference manual. It is aimed primarily at managers and engineers who need a broad understanding of the core concepts behind modern transmission networks. Since this book was first published, the world’s telecommunications networks have changed considerably. Synchronous transmission systems now form the basis of the vast majority of networks. With the introduction of Wave Division Multiplexing we have seen the addition of an optical layer in the network structure, both for high capacity long haul transport and for the efficient transfer of a wide range of protocols in the metropolitan area network. Despite the difficulties faced by many operators and manufacturers in the first few years of the 21st Century, use of the Internet and data traffic in general have continued to increase rapidly. This has had a direct impact on the way networks are being deployed. Much of this book has been re-written to reflect these changes, but the original focus on core concepts is retained. As a world leader in optical communications, line systems and multiplex equipment, Nortel has a broad expertise covering all aspects of transmission. Large investments in R&D and regular involvement in meetings of the major standards bodies have ensured that Nortel has furthered its position as a world leader in telecommunications systems. Nortel is committed to the development of a full range of synchronous and optical transmission products to operate within a Managed Transmission Network.
Synchronous Transmission Systems
Foreword iv
Synchronous Transmission Systems
iii Foreword
Foreword This book brings together a wide range of ideas and information to act as both an educational tool and a useful reference manual. It is aimed primarily at managers and engineers who need a broad understanding of the core concepts behind modern transmission networks. Since this book was first published, the world’s telecommunications networks have changed considerably. Synchronous transmission systems now form the basis of the vast majority of networks. With the introduction of Wave Division Multiplexing we have seen the addition of an optical layer in the network structure, both for high capacity long haul transport and for the efficient transfer of a wide range of protocols in the metropolitan area network. Despite the difficulties faced by many operators and manufacturers in the first few years of the 21st Century, use of the Internet and data traffic in general have continued to increase rapidly. This has had a direct impact on the way networks are being deployed. Much of this book has been re-written to reflect these changes, but the original focus on core concepts is retained. As a world leader in optical communications, line systems and multiplex equipment, Nortel has a broad expertise covering all aspects of transmission. Large investments in R&D and regular involvement in meetings of the major standards bodies have ensured that Nortel has furthered its position as a world leader in telecommunications systems. Nortel is committed to the development of a full range of synchronous and optical transmission products to operate within a Managed Transmission Network.
Synchronous Transmission Systems
Foreword iv
CH-5400 Baden (Switzerland), January 2008
Synchronous Transmission Systems
iii Foreword
Foreword This book brings together a wide range of ideas and information to act as both an educational tool and a useful reference manual. It is aimed primarily at managers and engineers who need a broad understanding of the core concepts behind modern transmission networks. Since this book was first published, the world’s telecommunications networks have changed considerably. Synchronous transmission systems now form the basis of the vast majority of networks. With the introduction of Wave Division Multiplexing we have seen the addition of an optical layer in the network structure, both for high capacity long haul transport and for the efficient transfer of a wide range of protocols in the metropolitan area network. Despite the difficulties faced by many operators and manufacturers in the first few years of the 21st Century, use of the Internet and data traffic in general have continued to increase rapidly. This has had a direct impact on the way networks are being deployed. Much of this book has been re-written to reflect these changes, but the original focus on core concepts is retained. As a world leader in optical communications, line systems and multiplex equipment, Nortel has a broad expertise covering all aspects of transmission. Large investments in R&D and regular involvement in meetings of the major standards bodies have ensured that Nortel has furthered its position as a world leader in telecommunications systems. Nortel is committed to the development of a full range of synchronous and optical transmission products to operate within a Managed Transmission Network.
Synchronous Transmission Systems
Foreword iv
Synchronous Transmission Systems
iii Foreword
Foreword This book brings together a wide range of ideas and information to act as both an educational tool and a useful reference manual. It is aimed primarily at managers and engineers who need a broad understanding of the core concepts behind modern transmission networks. Since this book was first published, the world’s telecommunications networks have changed considerably. Synchronous transmission systems now form the basis of the vast majority of networks. With the introduction of Wave Division Multiplexing we have seen the addition of an optical layer in the network structure, both for high capacity long haul transport and for the efficient transfer of a wide range of protocols in the metropolitan area network. Despite the difficulties faced by many operators and manufacturers in the first few years of the 21st Century, use of the Internet and data traffic in general have continued to increase rapidly. This has had a direct impact on the way networks are being deployed. Much of this book has been re-written to reflect these changes, but the original focus on core concepts is retained. As a world leader in optical communications, line systems and multiplex equipment, Nortel has a broad expertise covering all aspects of transmission. Large investments in R&D and regular involvement in meetings of the major standards bodies have ensured that Nortel has furthered its position as a world leader in telecommunications systems. Nortel is committed to the development of a full range of synchronous and optical transmission products to operate within a Managed Transmission Network.
Synchronous Transmission Systems
Foreword iv
Contents Contents
Synchronous Transmission Systems
Table of Contents
Foreword .......................................................................... iii
Table of Contents ............................................................. v
Introduction .......................................................................1Structure and Use of this Book .......................................................4
Transport Networks ..........................................................5Evolution of Transport Networks.....................................................5Transmission Technology .............................................................11SDH to the Operator ......................................................................19
Basics of the Synchronous Digital Hierarchy ..................21Terms and concepts .......................................................................22Example of SDH Multiplexing .....................................................28Explanation of the Multiplexing Structure ....................................29Introduction to Network Management ..........................................32Introduction to Protection ..............................................................33Introduction to Equipment Standardisation ...................................34
Elements of a Synchronous Transmission System ........35Functionality of a Network Element .............................................35Types of Network Element ...........................................................39Regenerators and Repeaters ..........................................................46Submarine Systems .......................................................................47
SDH Network Architecture and Design ...........................49Mapping User Connection Demand to a Physical Network .........49Types of VC-4 Layer Architecture ................................................51Network Optimisation and Flexibility ...........................................59VC-12 Layer Design .....................................................................60
Protection ........................................................................67Introduction ................................................................................... 67Terms ............................................................................................ 68Equipment Protection ................................................................... 71Restoration .................................................................................... 73Network Protection ....................................................................... 74Interworking of Protection Schemes ............................................ 92
Network Management .....................................................99The Physical Management Path .................................................. 101The TMN Layered Hierarchy ..................................................... 103Functionality of a Network Management System ...................... 106Network Management Platform ................................................. 111The Telecommunication Management Network ........................ 111
High Capacity Networks ...............................................115Wavelength Division Multiplexing ............................................ 117STM-64 ....................................................................................... 126Moving Beyond 2.5 Gbit/s ......................................................... 126Evolution to the Optical Network ............................................... 130
SDH - A Detailed Description .......................................135Synchronous Operation .............................................................. 135Differences Between SDH and SONET ..................................... 137The STM-1 Frame ...................................................................... 138European Multiplexing Structure ............................................... 144Concatenation ............................................................................. 153Higher Transmission Rates ......................................................... 156
Synchronisation ............................................................157The Voice Legacy ....................................................................... 158The Digital Revolution ............................................................... 159Service Sensitivity to Synchronisation ....................................... 159Cellular ....................................................................................... 162Video Services ............................................................................ 163Synchronisation Across Multiple Operators ............................... 164
v vi
Contents Contents
Synchronous Transmission Systems
Table of Contents
Foreword .......................................................................... iii
Table of Contents ............................................................. v
Introduction .......................................................................1Structure and Use of this Book .......................................................4
Transport Networks ..........................................................5Evolution of Transport Networks.....................................................5Transmission Technology .............................................................11SDH to the Operator ......................................................................19
Basics of the Synchronous Digital Hierarchy ..................21Terms and concepts .......................................................................22Example of SDH Multiplexing .....................................................28Explanation of the Multiplexing Structure ....................................29Introduction to Network Management ..........................................32Introduction to Protection ..............................................................33Introduction to Equipment Standardisation ...................................34
Elements of a Synchronous Transmission System ........35Functionality of a Network Element .............................................35Types of Network Element ...........................................................39Regenerators and Repeaters ..........................................................46Submarine Systems .......................................................................47
SDH Network Architecture and Design ...........................49Mapping User Connection Demand to a Physical Network .........49Types of VC-4 Layer Architecture ................................................51Network Optimisation and Flexibility ...........................................59VC-12 Layer Design .....................................................................60
Protection ........................................................................67Introduction ................................................................................... 67Terms ............................................................................................ 68Equipment Protection ................................................................... 71Restoration .................................................................................... 73Network Protection ....................................................................... 74Interworking of Protection Schemes ............................................ 92
Network Management .....................................................99The Physical Management Path .................................................. 101The TMN Layered Hierarchy ..................................................... 103Functionality of a Network Management System ...................... 106Network Management Platform ................................................. 111The Telecommunication Management Network ........................ 111
High Capacity Networks ...............................................115Wavelength Division Multiplexing ............................................ 117STM-64 ....................................................................................... 126Moving Beyond 2.5 Gbit/s ......................................................... 126Evolution to the Optical Network ............................................... 130
SDH - A Detailed Description .......................................135Synchronous Operation .............................................................. 135Differences Between SDH and SONET ..................................... 137The STM-1 Frame ...................................................................... 138European Multiplexing Structure ............................................... 144Concatenation ............................................................................. 153Higher Transmission Rates ......................................................... 156
Synchronisation ............................................................157The Voice Legacy ....................................................................... 158The Digital Revolution ............................................................... 159Service Sensitivity to Synchronisation ....................................... 159Cellular ....................................................................................... 162Video Services ............................................................................ 163Synchronisation Across Multiple Operators ............................... 164
v vi
1
2
3
4
5
Contents Contents
Synchronous Transmission Systems
Table of Contents
Foreword .......................................................................... iii
Table of Contents ............................................................. v
Introduction .......................................................................1Structure and Use of this Book .......................................................4
Transport Networks ..........................................................5Evolution of Transport Networks.....................................................5Transmission Technology .............................................................11SDH to the Operator ......................................................................19
Basics of the Synchronous Digital Hierarchy ..................21Terms and concepts .......................................................................22Example of SDH Multiplexing .....................................................28Explanation of the Multiplexing Structure ....................................29Introduction to Network Management ..........................................32Introduction to Protection ..............................................................33Introduction to Equipment Standardisation ...................................34
Elements of a Synchronous Transmission System ........35Functionality of a Network Element .............................................35Types of Network Element ...........................................................39Regenerators and Repeaters ..........................................................46Submarine Systems .......................................................................47
SDH Network Architecture and Design ...........................49Mapping User Connection Demand to a Physical Network .........49Types of VC-4 Layer Architecture ................................................51Network Optimisation and Flexibility ...........................................59VC-12 Layer Design .....................................................................60
Protection ........................................................................67Introduction ................................................................................... 67Terms ............................................................................................ 68Equipment Protection ................................................................... 71Restoration .................................................................................... 73Network Protection ....................................................................... 74Interworking of Protection Schemes ............................................ 92
Network Management .....................................................99The Physical Management Path .................................................. 101The TMN Layered Hierarchy ..................................................... 103Functionality of a Network Management System ...................... 106Network Management Platform ................................................. 111The Telecommunication Management Network ........................ 111
High Capacity Networks ...............................................115Wavelength Division Multiplexing ............................................ 117STM-64 ....................................................................................... 126Moving Beyond 2.5 Gbit/s ......................................................... 126Evolution to the Optical Network ............................................... 130
SDH - A Detailed Description .......................................135Synchronous Operation .............................................................. 135Differences Between SDH and SONET ..................................... 137The STM-1 Frame ...................................................................... 138European Multiplexing Structure ............................................... 144Concatenation ............................................................................. 153Higher Transmission Rates ......................................................... 156
Synchronisation ............................................................157The Voice Legacy ....................................................................... 158The Digital Revolution ............................................................... 159Service Sensitivity to Synchronisation ....................................... 159Cellular ....................................................................................... 162Video Services ............................................................................ 163Synchronisation Across Multiple Operators ............................... 164
v vi
vii viContents Contents
Synchronous Transmission Systems
Table of Contents
Foreword .......................................................................... iii
Table of Contents ............................................................. v
Introduction .......................................................................1Structure and Use of this Book .......................................................4
Transport Networks ..........................................................5Evolution of Transport Networks.....................................................5Transmission Technology .............................................................11SDH to the Operator ......................................................................19
Basics of the Synchronous Digital Hierarchy ..................21Terms and concepts .......................................................................22Example of SDH Multiplexing .....................................................28Explanation of the Multiplexing Structure ....................................29Introduction to Network Management ..........................................32Introduction to Protection ..............................................................33Introduction to Equipment Standardisation ...................................34
Elements of a Synchronous Transmission System ........35Functionality of a Network Element .............................................35Types of Network Element ...........................................................39Regenerators and Repeaters ..........................................................46Submarine Systems .......................................................................47
SDH Network Architecture and Design ...........................49Mapping User Connection Demand to a Physical Network .........49Types of VC-4 Layer Architecture ................................................51Network Optimisation and Flexibility ...........................................59VC-12 Layer Design .....................................................................60
Protection ........................................................................67Introduction ................................................................................... 67Terms ............................................................................................ 68Equipment Protection ................................................................... 71Restoration .................................................................................... 73Network Protection ....................................................................... 74Interworking of Protection Schemes ............................................ 92
Network Management .....................................................99The Physical Management Path .................................................. 101The TMN Layered Hierarchy ..................................................... 103Functionality of a Network Management System ...................... 106Network Management Platform ................................................. 111The Telecommunication Management Network ........................ 111
High Capacity Networks ...............................................115Wavelength Division Multiplexing ............................................ 117STM-64 ....................................................................................... 126Moving Beyond 2.5 Gbit/s ......................................................... 126Evolution to the Optical Network ............................................... 130
SDH - A Detailed Description .......................................135Synchronous Operation .............................................................. 135Differences Between SDH and SONET ..................................... 137The STM-1 Frame ...................................................................... 138European Multiplexing Structure ............................................... 144Concatenation ............................................................................. 153Higher Transmission Rates ......................................................... 156
Synchronisation ............................................................157The Voice Legacy ....................................................................... 158The Digital Revolution ............................................................... 159Service Sensitivity to Synchronisation ....................................... 159Cellular ....................................................................................... 162Video Services ............................................................................ 163Synchronisation Across Multiple Operators ............................... 164
v vi
6
7
8
9
10
vii vi Contents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
11
12
Contents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
Contents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
Evolving SDH to support Data-centric and Wavelength Services
Contents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
Contents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
Contents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
viiiContents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
Contents Contents
Synchronous Transmission Systems Synchronous Transmission Systems
Global Timing .............................................................................165Synchronisation Basics ...............................................................166External Timing Equipment ........................................................167Synchronising SDH Networks ....................................................168Synchronisation Status Messaging ..............................................169
Ethernet over SDH (GFP, LCAS) .................................171
vii viii
Generic Framing Procedure (GFP)................................................171 Resilient Package Ring..................................................................180
vii viii
List of Abbreviations ......................................................185
vii viii
1 1
1 Introduction
Introduction Since this book was first published, the world’s telecommunications networks have changed considerably. Network operators are responding to the end-user demand for sophisti-cated telecommunications services such as video conferencing, internet access, business continuity, remote database access and multimedia file transfer. These services require a flexible network with the availability on demand of virtually unlimited bandwidth. The far-reaching effects of the implementation of synchronous transmis-sion systems are, however, still evolving. Operators now need to deliver global, carrier-grade applications and services that merge all the disparate net-working elements and technologies into a seamless, open network, that is, a unified network. Discontinuities are driving the need for unified networks and will cause unprecedented changes, affecting virtually every type of business institu-tion. For example:
• Data traffic has now surpassed voice traffic yet the vast major-ity of network investments in the world are optimised for voice traffic.
• Optical networks make it possible to deliver ever larger streams of information over the same piece of glass, and this is at the centre of satisfying a huge demand for bandwidth. Yet access networks that can deliver this bandwidth over the first critical
1
Synchronous Transmission Systems
Introduction 2
• mile from users’ homes to the network are only now being de-ployed by cable companies, electrical utilities, wireless compa-nies and telephone companies. The prediction is that average residential user access speeds will quadruple in the next two years.
• The Internet, with its incredibly low costs per business transac-
tion, is driving a whole new business model. This in turn is driving new classes of customer relationships, but the networks needed to support these new models are often not in place.
These and other discontinuities are putting tremendous pressure on net-work users and providers. Enterprises of every kind and size are finding their internal and external bandwidth requirements are exploding. The web offers the potential for order of magnitude reductions in transaction costs, for example, in the sale of airline tickets and books over the Internet. Enterprises are driving to use Information Technology and networks to leverage the low cost of Internet commerce and high returns of supply chain management sys-tems. Carriers and service providers feel similar pressure from these disconti-nuities. Established carriers are faced with the enormous challenge of reconfiguring their networks to handle the explosive growth of data. High speed optical networks are forcing them to move switching away from the core of their networks to the edge. The line between optical and wireline switching is blurring. At the same time, carriers are looking for new ways to leverage the investment in their installed base. New carriers are starting fresh with data-optimised networks, but they have to find ways to offer services that are differentiated and provide revenue-generating value to their customers. Operators need to deliver unified networks that combine data and te-lephony, are seamless across the enterprise and carrier, delivering quality of service, and blending the lines between routing, optical, wireline, wireless, switching and IP. 1
1 Introduction
2
1 Introduction
Introduction Since this book was first published, the world’s telecommunications networks have changed considerably. Network operators are responding to the end-user demand for sophisti-cated telecommunications services such as video conferencing, internet access, business continuity, remote database access and multimedia file transfer. These services require a flexible network with the availability on demand of virtually unlimited bandwidth. The far-reaching effects of the implementation of synchronous transmis-sion systems are, however, still evolving. Operators now need to deliver global, carrier-grade applications and services that merge all the disparate net-working elements and technologies into a seamless, open network, that is, a unified network. Discontinuities are driving the need for unified networks and will cause unprecedented changes, affecting virtually every type of business institu-tion. For example:
• Data traffic has now surpassed voice traffic yet the vast major-ity of network investments in the world are optimised for voice traffic.
• Optical networks make it possible to deliver ever larger streams of information over the same piece of glass, and this is at the centre of satisfying a huge demand for bandwidth. Yet access networks that can deliver this bandwidth over the first critical
1
Synchronous Transmission Systems
Introduction 2
• mile from users’ homes to the network are only now being de-ployed by cable companies, electrical utilities, wireless compa-nies and telephone companies. The prediction is that average residential user access speeds will quadruple in the next two years.
• The Internet, with its incredibly low costs per business transac-
tion, is driving a whole new business model. This in turn is driving new classes of customer relationships, but the networks needed to support these new models are often not in place.
These and other discontinuities are putting tremendous pressure on net-work users and providers. Enterprises of every kind and size are finding their internal and external bandwidth requirements are exploding. The web offers the potential for order of magnitude reductions in transaction costs, for example, in the sale of airline tickets and books over the Internet. Enterprises are driving to use Information Technology and networks to leverage the low cost of Internet commerce and high returns of supply chain management sys-tems. Carriers and service providers feel similar pressure from these disconti-nuities. Established carriers are faced with the enormous challenge of reconfiguring their networks to handle the explosive growth of data. High speed optical networks are forcing them to move switching away from the core of their networks to the edge. The line between optical and wireline switching is blurring. At the same time, carriers are looking for new ways to leverage the investment in their installed base. New carriers are starting fresh with data-optimised networks, but they have to find ways to offer services that are differentiated and provide revenue-generating value to their customers. Operators need to deliver unified networks that combine data and te-lephony, are seamless across the enterprise and carrier, delivering quality of service, and blending the lines between routing, optical, wireline, wireless, switching and IP. 1
1 Introduction
3
Synchronous Transmission Systems
3 Introduction
The unified network architecture addresses the following needs:
• Personal networking - new levels of personal productivity, us-ers expectations and service demands are defined at this level,
• Flexible, high speed access - network access must be compara-
ble to the speeds that users experience within their own Enter-prise LAN environment, and flexible enough to deliver users’ expectations for services wherever they are needed,
• Switched infrastructures - switching technology offers reliable,
high performance infrastructures for reliable scalability to meet rapidly growing demands,
• High performance optical networks - the need to deliver huge
amounts of data at ever lower costs per megabyte of switched data,
• Applications and services - the need for networks designed to
deliver high value services and applications to meet user re-quirements.
The common factors are the massive increase in demand for bandwidth and the requirement for networks that can carry a variety of types of traffic across resilient networks. The implementation of synchronous transmission systems has given opera-tors a method of meeting these demands. Their networks are more efficient and enable savings to be made in their operating costs. Embedded man-agement channels within the transmission make the operation, administra-tion and maintenance of networks far more efficient and gives operators the opportunity to dramatically reduce the time and cost of provisioning new services. The introduction of Wavelength Division Multiplexing (WDM) techniques provides an optical layer for the synchronous network which will enable the bandwidth demands to be satisfied. This book describes the evolution, facilities and features of synchronous transmission networks and also describes how these will cope with the demands of the unified networks of the future.
Synchronous Transmission Systems
Introduction 4
Structure and Use of this Book Chapter 2 gives a brief evolution of telecommunications transport net-works andtransmission technology. If this is familiar, you may wish to omit reading this chapter. Chapter 3 describes the basics of the Synchronous Digital Hierarchy (SDH). Ifyou are familiar with the SDH standards you may wish to omit reading this chapter. Chapter 4 details the various types and uses of network elements used insynchronous transmission systems. Chapter 5 discusses the network architecture of synchronous transmis-sionsystems and uses a typical network as an example. Chapter 6 details the various protection schemes used in synchro-noustransmission systems to achieve the network resilience demanded by end users. Chapter 7 dicusses the management of synchronous networks; the ar-chitectureof the management system, its functions and facilities provided at the variouslevels in the architecture. Chapter 8 describes the high capacity transmission network. Details the use ofWDM and optical networking. Chapter 9 gives a more detailed description of the Synchronous Digi-talHeirarchy. Chapter 10 discusses the need for synchronisation across the network anddescribes some of the methods of achieving this. Chapter 11 EoSDH: Discuss solutions for transmission of IP/Ethernet over SDH networks (GFP, LCAS) Chapter 12 is a list of abbreviations.
1 Introduction
4
Synchronous Transmission Systems
3 Introduction
The unified network architecture addresses the following needs:
• Personal networking - new levels of personal productivity, us-ers expectations and service demands are defined at this level,
• Flexible, high speed access - network access must be compara-
ble to the speeds that users experience within their own Enter-prise LAN environment, and flexible enough to deliver users’ expectations for services wherever they are needed,
• Switched infrastructures - switching technology offers reliable,
high performance infrastructures for reliable scalability to meet rapidly growing demands,
• High performance optical networks - the need to deliver huge
amounts of data at ever lower costs per megabyte of switched data,
• Applications and services - the need for networks designed to
deliver high value services and applications to meet user re-quirements.
The common factors are the massive increase in demand for bandwidth and the requirement for networks that can carry a variety of types of traffic across resilient networks. The implementation of synchronous transmission systems has given opera-tors a method of meeting these demands. Their networks are more efficient and enable savings to be made in their operating costs. Embedded man-agement channels within the transmission make the operation, administra-tion and maintenance of networks far more efficient and gives operators the opportunity to dramatically reduce the time and cost of provisioning new services. The introduction of Wavelength Division Multiplexing (WDM) techniques provides an optical layer for the synchronous network which will enable the bandwidth demands to be satisfied. This book describes the evolution, facilities and features of synchronous transmission networks and also describes how these will cope with the demands of the unified networks of the future.
Synchronous Transmission Systems
Introduction 4
Structure and Use of this Book Chapter 2 gives a brief evolution of telecommunications transport net-works andtransmission technology. If this is familiar, you may wish to omit reading this chapter. Chapter 3 describes the basics of the Synchronous Digital Hierarchy (SDH). Ifyou are familiar with the SDH standards you may wish to omit reading this chapter. Chapter 4 details the various types and uses of network elements used insynchronous transmission systems. Chapter 5 discusses the network architecture of synchronous transmis-sionsystems and uses a typical network as an example. Chapter 6 details the various protection schemes used in synchro-noustransmission systems to achieve the network resilience demanded by end users. Chapter 7 dicusses the management of synchronous networks; the ar-chitectureof the management system, its functions and facilities provided at the variouslevels in the architecture. Chapter 8 describes the high capacity transmission network. Details the use ofWDM and optical networking. Chapter 9 gives a more detailed description of the Synchronous Digi-talHeirarchy. Chapter 10 discusses the need for synchronisation across the network anddescribes some of the methods of achieving this. Chapter 11 EoSDH: Discuss solutions for transmission of IP/Ethernet over SDH networks (GFP, LCAS) Chapter 12 is a list of abbreviations.
1 Introduction
5 Transport Networks Transport Networks
Synchronous Transmission Systems
Transport Networks
The aim of this chapter is to give the reader a basic introduction to transportnetworks; what they are, how they evolved and the technology advances thathave occurred including the emergence of the Synchronous Digital Hierarchy(SDH). The reader should refer to Chapter 3 for a basic introduction to the termsand concepts used in SDH.
Evolution of Transport Networks
How do individuals and organisations communicate with each other from farsides of the world? How do our telephone calls, fax messages, the internet andvideo conferences travel across the country or around the globe? A transportnetwork is the physical means by which this near instantaneous communicationis carried. So how has this global network evolved?
The global transport network began as a mechanism for transporting voiceconversations between telephone handsets. As telecommunications evolved thetype and volume of traffic carried has expanded and the requirements placed onthe transport network have increased. This in turn has driven the development oftransmission technology and the need for transport networks with a high degreeof intelligence. Reliable, flexible networks are required to transport a wide rangeof services and to deliver those services at the level of quality demanded by endusers.
Figure 2-1 Video conference across the world
Traffic carried today on networks may be Public Switched Telephone Network(PSTN) traffic such as circuit switched voice and data or packet switched datatraffic such as Internet Protocol (IP). Whatever the traffic type there is a need fora physical transport layer to manage traffic on the network infrastructure whichis typically optical fibre. SDH (along with the North American equivalent,SONET) is the dominant physical layer technology for optical fibre networks.
SDH is essentially a transport protocol; it is the physical bearer layer of thenetwork carrying different traffic types and applications between destinations.These traffic applications can vary from private circuits and PlesiochronousDigital Hierarchy (PDH) traffic through to Asynchronous Transfer Mode(ATM), Internet Protocol (IP) and data networks. SDH is a major networkdevelopment and enabling technology for services in a competitiveenvironment. Fundamentally new concepts are introduced into the transportnetwork by SDH, but to understand these we must first outline how transportnetworks have evolved.
2
5 62 Transport Networks
7 6 7 6Transport Networks Transport Networks
Synchronous Transmission Systems
Transport Networks
The aim of this chapter is to give the reader a basic introduction to transportnetworks; what they are, how they evolved and the technology advances thathave occurred including the emergence of the Synchronous Digital Hierarchy(SDH). The reader should refer to Chapter 3 for a basic introduction to the termsand concepts used in SDH.
Evolution of Transport Networks
How do individuals and organisations communicate with each other from farsides of the world? How do our telephone calls, fax messages, the internet andvideo conferences travel across the country or around the globe? A transportnetwork is the physical means by which this near instantaneous communicationis carried. So how has this global network evolved?
The global transport network began as a mechanism for transporting voiceconversations between telephone handsets. As telecommunications evolved thetype and volume of traffic carried has expanded and the requirements placed onthe transport network have increased. This in turn has driven the development oftransmission technology and the need for transport networks with a high degreeof intelligence. Reliable, flexible networks are required to transport a wide rangeof services and to deliver those services at the level of quality demanded by endusers.
Figure 2-1 Video conference across the world
Traffic carried today on networks may be Public Switched Telephone Network(PSTN) traffic such as circuit switched voice and data or packet switched datatraffic such as Internet Protocol (IP). Whatever the traffic type there is a need fora physical transport layer to manage traffic on the network infrastructure whichis typically optical fibre. SDH (along with the North American equivalent,SONET) is the dominant physical layer technology for optical fibre networks.
SDH is essentially a transport protocol; it is the physical bearer layer of thenetwork carrying different traffic types and applications between destinations.These traffic applications can vary from private circuits and PlesiochronousDigital Hierarchy (PDH) traffic through to Asynchronous Transfer Mode(ATM), Internet Protocol (IP) and data networks. SDH is a major networkdevelopment and enabling technology for services in a competitiveenvironment. Fundamentally new concepts are introduced into the transportnetwork by SDH, but to understand these we must first outline how transportnetworks have evolved.
2
5 62 Transport Networks
7 6 7 6Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Simple Telephone Networks
Communication systems began as permanent point to point connectionsbetween two subscribers. These connections carried telephone conversationsfrom one person to another. It is not difficult to imagine that other people wouldwant to be included in the system and that each subscriber would want to haveaccess to all other subscribers when necessary.
Initially each pair of subscribers were connected by a dedicated telephone lineas in Figure 2-2 (a).
Figure 2-2 Point to point and operator networks
As the number of subscribers grew the number of connections required rapidlyincreased so dedicated lines very quickly became impractical. A central pointemerged to which all subscribers were connected. At this central point any twosubscribers could be interconnected when it was demanded, but only when itwas demanded, as shown in Figure 2-2 (b). For example when A demands to
speak to C a semi-permanent connection was made between A and C for theduration of the call. This is the simple function of a switch and from this basicfunction telephone exchanges evolved.
Figure 2-3 Early telephone exchange
As more people wanted to be connected to the exchange and the geographicalscope of these telephone networks increased, it became necessary to havehierarchies of exchanges as shown in Figure 2-4.
These exchanges were typically interconnected by simple point to pointtransmission links; this was the transport network. The intelligence of thenetwork was in the exchange where switching took place, rather than thetransport links themselves. These networks were adequate, although not optimalfor the transport of voice, fax and modem.
A
B
CD
E
A
B
CD
E
Permanent connectionfrom A to C
Operator makes connectionfrom A to C for theduration of the call
(a) 10 transport links (b) 5 transport links
7 82 Transport Networks
9 8 9 8Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Simple Telephone Networks
Communication systems began as permanent point to point connectionsbetween two subscribers. These connections carried telephone conversationsfrom one person to another. It is not difficult to imagine that other people wouldwant to be included in the system and that each subscriber would want to haveaccess to all other subscribers when necessary.
Initially each pair of subscribers were connected by a dedicated telephone lineas in Figure 2-2 (a).
Figure 2-2 Point to point and operator networks
As the number of subscribers grew the number of connections required rapidlyincreased so dedicated lines very quickly became impractical. A central pointemerged to which all subscribers were connected. At this central point any twosubscribers could be interconnected when it was demanded, but only when itwas demanded, as shown in Figure 2-2 (b). For example when A demands to
speak to C a semi-permanent connection was made between A and C for theduration of the call. This is the simple function of a switch and from this basicfunction telephone exchanges evolved.
Figure 2-3 Early telephone exchange
As more people wanted to be connected to the exchange and the geographicalscope of these telephone networks increased, it became necessary to havehierarchies of exchanges as shown in Figure 2-4.
These exchanges were typically interconnected by simple point to pointtransmission links; this was the transport network. The intelligence of thenetwork was in the exchange where switching took place, rather than thetransport links themselves. These networks were adequate, although not optimalfor the transport of voice, fax and modem.
A
B
CD
E
A
B
CD
E
Permanent connectionfrom A to C
Operator makes connectionfrom A to C for theduration of the call
(a) 10 transport links (b) 5 transport links
7 82 Transport Networks
9 8 9 8Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 2-4 Exchange hierarchy
The Transport Network Evolution
Transport networks are no longer just the simple point to point links betweenexchanges. Operators and users now require intelligent, hierarchicalinternational networks.
The last few decades have seen an enormous growth in the amount of voicetraffic transported, increasing the load on networks. The most important driverfor the evolution of transport has, however, been the emergence of manydifferent types of traffic; new, non-voice, services mostly aimed at the business
customer. The personal computer and data explosion has placed demand onnetworks not only in terms of traffic volume, but also reliability. The transportof data files and internet traffic requires integrity so that software at the receivingend can interpret files.
Businesses rely on these alternative services to maintain a competitiveadvantage. They demand ever improved transmission quality, higher reliabilityof service and more flexible connection patterns. With these demands a newdimension of transport network has emerged.
The growth rates of voice services is now relatively low (less than 5% in mostcountries), while some new non-voice services are growing by as much as 50%each year. The higher margins associated with business customers means thatmany network operators gain a disproportionate share of their income fromthese alternative services. They are understandably keen to improve the qualityand flexibility of the transport network.
Characteristics of Modern Transport Networks
Some of the major developments which have enabled transport networks to meettoday�s requirements are:
Digital Multiplexing
This was introduced over twenty years ago and enabled analogue speech signalsto be carried in a digital form over networks. Digital traffic can be carried muchmore efficiently and enables performance monitoring for quality purposes.
Optical Fibre
This is commonly deployed in transport networks today. It has a much largertraffic carrying capacity than copper or coaxial links and has driven down thecosts associated with carrying traffic.
Protection Schemes
These have been standardised to ensure a reliable service. Should a fault or fibrebreak occur traffic can be switched to an alternative route, so the end userexperiences no disruption of service.
LE LE LE LE LE LE LE LE
GSC GSC GSC GSC
DSC DSC
MSC
Subscribers
MSCDSCGSCLE
Main Switching CentreDigital Switching CentreGroup Switching CentreLocal Exchange
KEY:
Point to point links
9 102 Transport Networks
11 10 11 10Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 2-4 Exchange hierarchy
The Transport Network Evolution
Transport networks are no longer just the simple point to point links betweenexchanges. Operators and users now require intelligent, hierarchicalinternational networks.
The last few decades have seen an enormous growth in the amount of voicetraffic transported, increasing the load on networks. The most important driverfor the evolution of transport has, however, been the emergence of manydifferent types of traffic; new, non-voice, services mostly aimed at the business
customer. The personal computer and data explosion has placed demand onnetworks not only in terms of traffic volume, but also reliability. The transportof data files and internet traffic requires integrity so that software at the receivingend can interpret files.
Businesses rely on these alternative services to maintain a competitiveadvantage. They demand ever improved transmission quality, higher reliabilityof service and more flexible connection patterns. With these demands a newdimension of transport network has emerged.
The growth rates of voice services is now relatively low (less than 5% in mostcountries), while some new non-voice services are growing by as much as 50%each year. The higher margins associated with business customers means thatmany network operators gain a disproportionate share of their income fromthese alternative services. They are understandably keen to improve the qualityand flexibility of the transport network.
Characteristics of Modern Transport Networks
Some of the major developments which have enabled transport networks to meettoday�s requirements are:
Digital Multiplexing
This was introduced over twenty years ago and enabled analogue speech signalsto be carried in a digital form over networks. Digital traffic can be carried muchmore efficiently and enables performance monitoring for quality purposes.
Optical Fibre
This is commonly deployed in transport networks today. It has a much largertraffic carrying capacity than copper or coaxial links and has driven down thecosts associated with carrying traffic.
Protection Schemes
These have been standardised to ensure a reliable service. Should a fault or fibrebreak occur traffic can be switched to an alternative route, so the end userexperiences no disruption of service.
LE LE LE LE LE LE LE LE
GSC GSC GSC GSC
DSC DSC
MSC
Subscribers
MSCDSCGSCLE
Main Switching CentreDigital Switching CentreGroup Switching CentreLocal Exchange
KEY:
Point to point links
9 102 Transport Networks
11 10 11 10Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Ring Topologies
These are being deployed in increasing numbers. This is because should a linkbe lost, there is an alternative traffic path the other way around the ring.Operators can minimise the number of links and optical fibre deployed in thenetwork. This is very important as the cost of putting new optical fibre cables inthe ground is so high.
Network Management
Management of these networks from a single remote site is an important featurefor operators. Software has been developed which allows all nodes and trafficpaths to be managed from a single site. An operator can now manage a varietyof functions such as provisioning capacity in response to customers demandsand monitoring the quality of the network.
Each of these topics will be discussed in their own right in subsequent chapters.The rest of this chapter will discuss these major developments in greater detail.
Transmission Technology
Analogue Transmission
Until about 1970 transportation of voice signals was achieved by carryinganalogue signals over copper twisted pairs. Frequency Division Multiplexing(FDM) was used on long-haul routes to combine multiple traffic signals on asingle coaxial cable.
Digital Transmission
In the early 1970s digital transmission systems began to appear, utilising PulseCode Modulation (PCM) - first proposed by Alec Reeves of STC in 1937. PCMenables analogue speech signals to be represented in a binary form. Using thismethod it is possible to convert the standard 300 to 3400 Hz analogue telephonebandwidth into a 64 kbit/s digital bit stream.
Figure 2-5 shows the principles of PCM. The analogue speech signals aresampled, quantised (rounded to the nearest integer value) and then encoded togive a binary pattern which faithfully represents the analogue speech signal from
which it was derived. This binary information can be passed through a digitaltransmission system, after which the original analogue speech signal can then bereconstituted.
Figure 2-5 Pulse Code Modulation
Time Division Multiplexing
Engineers saw the potential to produce more cost-effective transport systems bycombining several PCM channels and transmitting them down the same coppertwisted pair as had previously been occupied by a single analogue signal.
The method used to combine multiple 64 kbit/s digital bit streams into a singlehigh speed bit stream is known as Time Division Multiplexing (TDM). In simpleterms, a byte from each incoming channel is transmitted in turn down theoutgoing higher speed channel, (see Figure 2-6). This process is sometimesreferred to as �sequential byte interleaving�.
In Europe and many other parts of the world, a standard TDM scheme wasadopted whereby thirty 64 kbit/s channels were combined with two additionalchannels carrying control and signalling information, to produce a structurewith a bit rate of 2.048 Mbit/s (for simplicity, referred to as 2 Mbit/s). As thecost of digital electronics began to fall, major cost savings became possiblethrough the use of these techniques.
Sampler
0
20
15
10
5
011010010Quantiser Encoder
11 122 Transport Networks
13 12 13 12Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Ring Topologies
These are being deployed in increasing numbers. This is because should a linkbe lost, there is an alternative traffic path the other way around the ring.Operators can minimise the number of links and optical fibre deployed in thenetwork. This is very important as the cost of putting new optical fibre cables inthe ground is so high.
Network Management
Management of these networks from a single remote site is an important featurefor operators. Software has been developed which allows all nodes and trafficpaths to be managed from a single site. An operator can now manage a varietyof functions such as provisioning capacity in response to customers demandsand monitoring the quality of the network.
Each of these topics will be discussed in their own right in subsequent chapters.The rest of this chapter will discuss these major developments in greater detail.
Transmission Technology
Analogue Transmission
Until about 1970 transportation of voice signals was achieved by carryinganalogue signals over copper twisted pairs. Frequency Division Multiplexing(FDM) was used on long-haul routes to combine multiple traffic signals on asingle coaxial cable.
Digital Transmission
In the early 1970s digital transmission systems began to appear, utilising PulseCode Modulation (PCM) - first proposed by Alec Reeves of STC in 1937. PCMenables analogue speech signals to be represented in a binary form. Using thismethod it is possible to convert the standard 300 to 3400 Hz analogue telephonebandwidth into a 64 kbit/s digital bit stream.
Figure 2-5 shows the principles of PCM. The analogue speech signals aresampled, quantised (rounded to the nearest integer value) and then encoded togive a binary pattern which faithfully represents the analogue speech signal from
which it was derived. This binary information can be passed through a digitaltransmission system, after which the original analogue speech signal can then bereconstituted.
Figure 2-5 Pulse Code Modulation
Time Division Multiplexing
Engineers saw the potential to produce more cost-effective transport systems bycombining several PCM channels and transmitting them down the same coppertwisted pair as had previously been occupied by a single analogue signal.
The method used to combine multiple 64 kbit/s digital bit streams into a singlehigh speed bit stream is known as Time Division Multiplexing (TDM). In simpleterms, a byte from each incoming channel is transmitted in turn down theoutgoing higher speed channel, (see Figure 2-6). This process is sometimesreferred to as �sequential byte interleaving�.
In Europe and many other parts of the world, a standard TDM scheme wasadopted whereby thirty 64 kbit/s channels were combined with two additionalchannels carrying control and signalling information, to produce a structurewith a bit rate of 2.048 Mbit/s (for simplicity, referred to as 2 Mbit/s). As thecost of digital electronics began to fall, major cost savings became possiblethrough the use of these techniques.
Sampler
0
20
15
10
5
011010010Quantiser Encoder
11 122 Transport Networks
13 12 13 12Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 2-6 Time Division Multiplexing
Digital Hierarchy
As demand for voice telephony increased, and levels of traffic in the networkgrew ever higher. It became clear that the standard 2 Mbit/s signal was notsufficient to cope with the traffic loads occurring in the trunk network. In orderto avoid having to use excessively large numbers of 2 Mbit/s links, it wasdecided to create a further level of multiplexing. The standard adopted in Europeinvolved the combination of four 2 Mbit/s channels to produce a single 8 Mbit/schannel (more exactly, 8.448 Mbit/s). As the need arose, further levels ofmultiplexing were added to the standard at 34 Mbit/s (34.368 Mbit/s),140 Mbit/s (139.264 Mbit/s), and 565 Mbit/s (564.992 Mbit/s) to produce a fullhierarchy of bit rates.
While the European digital transmission hierarchy was being developed, similarwork was occurring in North America to develop their own hierarchy. Althoughthe same principles were used, the hierarchy which evolved differedconsiderably in that its bit rates were 1.5 Mbit/s (1.544 Mbit/s), 6 Mbit/s(6.312 Mbit/s), and 45 Mbit/s (44.736 Mbit/s). These differences were to makeinterworking between the two hierarchies expensive to achieve. Figure 2-7shows the comparison between the North American and European transmissionhierarchies.
Figure 2-7 European & North American Transmission Hierarchies
Plesiochronous Transmission
The multiplexing hierarchies described above appear simple enough inprinciple, but in practice there are complications. You may have noticed thatmultiplexing four 2.048 Mbit/s signals should result in an 8.192 Mbit/s signalnot 8.448 Mbit/s. There are also similar differences at the higher rate signals.This is because when multiplexing a number of 2 Mbit/s channels they are likelyto have been created by different pieces of equipment, each generating a slightly
1d
1c
1b
1a
2d
2c
2b
2a
3d
3c
3b
3a
4d
4c
4b
4a
125 � s
3d
2d
1d
4c
3c
2c
1c
4b
3b
2b
1b
4a
3a
2a
4d
1a
125 � 4 = 31.25 � s
To line term inationequipm ent (LTE) fortransm ission
Com bined dig ita l s ignalrunning at 256 kbit/s,show ing how bytesare interleaved
4 into 1 TDMm ultip lexer
Analogue todig ita l converter
Analogue voicefrequency inputs
D igitised voice channels,running at 64 kbit/s. Eachchannel has exactly the sam e bit rate. Four bytes(A, B , C , D) are shown foreach channel
Channel 1
Channel 2
Channel 3
Channel 4
64 kbit/s
x30
x4
x4
x3
x7 x6
x4 x4
x3
* not recognised by ITU-T
x24
x4
DS1 DS2 DS3
DS0
Zero Order
First Order Second Order Third Order Fourth Order
1,544 kbit/s 6,312 kbit/s 44,736 kbit/s *274,176 kbit/s
2,048 kbit/s 8,448 kbit/s 34,368 kbit/s 139,264 kbit/s *564,992 kbit/s
13 142 Transport Networks
15 14 15 14Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 2-6 Time Division Multiplexing
Digital Hierarchy
As demand for voice telephony increased, and levels of traffic in the networkgrew ever higher. It became clear that the standard 2 Mbit/s signal was notsufficient to cope with the traffic loads occurring in the trunk network. In orderto avoid having to use excessively large numbers of 2 Mbit/s links, it wasdecided to create a further level of multiplexing. The standard adopted in Europeinvolved the combination of four 2 Mbit/s channels to produce a single 8 Mbit/schannel (more exactly, 8.448 Mbit/s). As the need arose, further levels ofmultiplexing were added to the standard at 34 Mbit/s (34.368 Mbit/s),140 Mbit/s (139.264 Mbit/s), and 565 Mbit/s (564.992 Mbit/s) to produce a fullhierarchy of bit rates.
While the European digital transmission hierarchy was being developed, similarwork was occurring in North America to develop their own hierarchy. Althoughthe same principles were used, the hierarchy which evolved differedconsiderably in that its bit rates were 1.5 Mbit/s (1.544 Mbit/s), 6 Mbit/s(6.312 Mbit/s), and 45 Mbit/s (44.736 Mbit/s). These differences were to makeinterworking between the two hierarchies expensive to achieve. Figure 2-7shows the comparison between the North American and European transmissionhierarchies.
Figure 2-7 European & North American Transmission Hierarchies
Plesiochronous Transmission
The multiplexing hierarchies described above appear simple enough inprinciple, but in practice there are complications. You may have noticed thatmultiplexing four 2.048 Mbit/s signals should result in an 8.192 Mbit/s signalnot 8.448 Mbit/s. There are also similar differences at the higher rate signals.This is because when multiplexing a number of 2 Mbit/s channels they are likelyto have been created by different pieces of equipment, each generating a slightly
1d
1c
1b
1a
2d
2c
2b
2a
3d
3c
3b
3a
4d
4c
4b
4a
125 � s
3d
2d
1d
4c
3c
2c
1c
4b
3b
2b
1b
4a
3a
2a
4d
1a
125 � 4 = 31.25 � s
To line term inationequipm ent (LTE) fortransm ission
Com bined dig ita l s ignalrunning at 256 kbit/s,show ing how bytesare interleaved
4 into 1 TDMm ultip lexer
Analogue todig ita l converter
Analogue voicefrequency inputs
D igitised voice channels,running at 64 kbit/s. Eachchannel has exactly the sam e bit rate. Four bytes(A, B , C , D) are shown foreach channel
Channel 1
Channel 2
Channel 3
Channel 4
64 kbit/s
x30
x4
x4
x3
x7 x6
x4 x4
x3
* not recognised by ITU-T
x24
x4
DS1 DS2 DS3
DS0
Zero Order
First Order Second Order Third Order Fourth Order
1,544 kbit/s 6,312 kbit/s 44,736 kbit/s *274,176 kbit/s
2,048 kbit/s 8,448 kbit/s 34,368 kbit/s 139,264 kbit/s *564,992 kbit/s
13 142 Transport Networks
15 14 15 14Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
different bit rate. Before the 2 Mbit/s channels can be bit interleaved they mustall be brought up to the same bit rate by adding �dummy� information bits, or�justification bits�. The same problems with synchronisation as described aboveoccur at every level of the multiplexing hierarchy, so justification bits are addedat each stage. This process is known as �plesiochronous� operation, after theGreek meaning �almost synchronous� and has led to the name PlesiochronousDigital Hierarchy (PDH) being applied to this type of network. The operation ofa plesiochronous multiplexer is shown in Figure 2-8.
Figure 2-8 Plesiochronous Multiplexing
The problem of flexibility in a plesiochronous network is illustrated byconsidering what a network operator may need to do in order to be able toprovide a business customer with a 2 Mbit/s leased line. If a higher speedstructure passes near the customer, the operation of providing a single 2 Mbit/s
line from within the higher speed structure would seem straightforward enough.In practice, however, it is not so simple.
The use of justification bits at each level in the PDH means that identifying theexact location of the 32 channels that make up the 2 Mbit/s line within, forexample, a 140 Mbit/s structure is impossible. In order to access a single2 Mbit/s line, the 140 Mbit/s structure must be completely demultiplexed downto its 64 constituent 2 Mbit/s lines via 34 and 8 Mbit/s stages. Once the required2 Mbit/s line has been identified and extracted, the remaining 2 Mbit/s linesmust then be multiplexed again up to 140 Mbit/s. This is shown in Figure 2-9.
Obviously this problem with the �drop and insert� of channels does not make forvery flexible connection patterns or rapid provisioning of services, while the�multiplexer mountains� required are extremely expensive.
Figure 2-9 Plesiochronous Drop and Insert
This illustrates one of the main limitations of PDH; the inability to identifyindividual channels in a high speed bit stream. Another limitation of PDHrelates to management; the PDH frame structure has insufficient provision forcarrying network management information. These limitations are not critical ina network dominated by voice traffic, but as more sophisticated services becomepopular, PDH can no longer cope.
4 3 2 1
3 2 1
4 3 2 1J J
J 3 2 1J J
masteroscillator
not many justification bitsadded (two shown)
bit rate
adaptation
bit rate
adoption
many justification bitsadded (three shown)
�slow� incoming2 Mbit/s channel
bits
�fast� incoming2 Mbit/s channel
Bit Interleaver
timing
Customer
2.048 Mbit/s
8.448 Mbit/s
34.368 Mbit/s
140 Mbit/sLTE
140
34
34
8
8
2
140 Mbit/sLTE
140
34
34
8
8
2
15 162 Transport Networks
17 16 17 16Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
different bit rate. Before the 2 Mbit/s channels can be bit interleaved they mustall be brought up to the same bit rate by adding �dummy� information bits, or�justification bits�. The same problems with synchronisation as described aboveoccur at every level of the multiplexing hierarchy, so justification bits are addedat each stage. This process is known as �plesiochronous� operation, after theGreek meaning �almost synchronous� and has led to the name PlesiochronousDigital Hierarchy (PDH) being applied to this type of network. The operation ofa plesiochronous multiplexer is shown in Figure 2-8.
Figure 2-8 Plesiochronous Multiplexing
The problem of flexibility in a plesiochronous network is illustrated byconsidering what a network operator may need to do in order to be able toprovide a business customer with a 2 Mbit/s leased line. If a higher speedstructure passes near the customer, the operation of providing a single 2 Mbit/s
line from within the higher speed structure would seem straightforward enough.In practice, however, it is not so simple.
The use of justification bits at each level in the PDH means that identifying theexact location of the 32 channels that make up the 2 Mbit/s line within, forexample, a 140 Mbit/s structure is impossible. In order to access a single2 Mbit/s line, the 140 Mbit/s structure must be completely demultiplexed downto its 64 constituent 2 Mbit/s lines via 34 and 8 Mbit/s stages. Once the required2 Mbit/s line has been identified and extracted, the remaining 2 Mbit/s linesmust then be multiplexed again up to 140 Mbit/s. This is shown in Figure 2-9.
Obviously this problem with the �drop and insert� of channels does not make forvery flexible connection patterns or rapid provisioning of services, while the�multiplexer mountains� required are extremely expensive.
Figure 2-9 Plesiochronous Drop and Insert
This illustrates one of the main limitations of PDH; the inability to identifyindividual channels in a high speed bit stream. Another limitation of PDHrelates to management; the PDH frame structure has insufficient provision forcarrying network management information. These limitations are not critical ina network dominated by voice traffic, but as more sophisticated services becomepopular, PDH can no longer cope.
4 3 2 1
3 2 1
4 3 2 1J J
J 3 2 1J J
masteroscillator
not many justification bitsadded (two shown)
bit rate
adaptation
bit rate
adoption
many justification bitsadded (three shown)
�slow� incoming2 Mbit/s channel
bits
�fast� incoming2 Mbit/s channel
Bit Interleaver
timing
Customer
2.048 Mbit/s
8.448 Mbit/s
34.368 Mbit/s
140 Mbit/sLTE
140
34
34
8
8
2
140 Mbit/sLTE
140
34
34
8
8
2
15 162 Transport Networks
17 16 17 16Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
The Emergence of SDH Standards
Synchronous transmission systems have been developed so that operators candeploy flexible, resilient networks. The dropping and insertion of channels canbe achieved in a single multiplexer. Provision for a network managementcapability is defined in the standards. In fact, a concerted standards effort hasbeen involved in the development of SDH. The opportunity of defining this setof standards has been used to address a number of other problems. For examplethe need to define standard interfaces between equipment for multi-vendorinteroperability and the need to facilitate interworking between North Americanand European transmission hierarchies.
This standards work culminated in CCITT (now ITU-T) RecommendationsG.707, G.708, and G.709 covering the Synchronous Digital Hierarchy whichwere published in 1989. In North America, ANSI published its SONETstandards, which can now be thought of as a subset of the worldwide SDHstandards.
The ITU-T recommendations define a number of basic transmission rates withinthe SDH. The first of these is 155.52 Mbit/s, normally referred to as �STM-1�(where STM stands for �Synchronous Transport Module�). Higher transmissionrates of STM-4, STM-16 and STM-64 (622.08 Mbit/s, 2488.32 Mbit/s and9953.28 Mbit/s respectively) are also defined, with further levels proposed forstudy.
The recommendations also define a multiplexing structure whereby an STM-1signal can carry a number of lower rate signals as a payload. Existing PDHsignals can be carried over this SDH network as a payload. This process will beexplained in more detail in the next chapter.
The Open Systems Interconnect Model
The seven layer Open Systems Interconnect (OSI) model defined by theInternational Organisation for Standardisation (ISO), is widely used to representthe functions of different network systems and applications, interoperabilitybetween equipment and interactions between networks. It can be used torepresent how these applications and protocols interact as it progresses throughthe network.
Figure 2-10 The OSI model
Within this model SDH is essentially a transport protocol and is the dominanttechnology for the physical or bearer layer. SDH can act as the physical bearerlayer for applications in layers 2 to 4 such as ATM or IP or act as layers 1 to 4and transport higher level applications directly. Its role is to manage theutilisation of the fibre infrastructure efficiently, detect failures and recover fromthem transparently to the higher layers.
One of the most common network layer protocols is IP. Using this as anexample, IP routers are layer 3 network devices which work on an end to endbasis, routing and forwarding packets to their destination. Routers willautomatically discover network topology and select the best paths, however, toensure that these end to end paths are resilient to transient failures, such as theloss of packets due to congestion, restoration based on routers is inherently slow.The network therefore relies on the underlying layers, in particular the physicalSDH layer, to detect failures and restore traffic over a backup path in a very shorttime. In the case of SDH this is typically 50 ms.
Application Layer
Presentation Layer
Session Layer
Transport Layer
Network Layer
Data Link Layer
Physical Layer
(typically IP)
(typically Frame Relay
(SDH)
SolutionCommunications
The means ofestablishing/maintaininga connection
Infrastructure
The What the PC �sees�
or ATM)
1
2
3
4
5
6
7
17 182 Transport Networks
19 18 19 18Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
The Emergence of SDH Standards
Synchronous transmission systems have been developed so that operators candeploy flexible, resilient networks. The dropping and insertion of channels canbe achieved in a single multiplexer. Provision for a network managementcapability is defined in the standards. In fact, a concerted standards effort hasbeen involved in the development of SDH. The opportunity of defining this setof standards has been used to address a number of other problems. For examplethe need to define standard interfaces between equipment for multi-vendorinteroperability and the need to facilitate interworking between North Americanand European transmission hierarchies.
This standards work culminated in CCITT (now ITU-T) RecommendationsG.707, G.708, and G.709 covering the Synchronous Digital Hierarchy whichwere published in 1989. In North America, ANSI published its SONETstandards, which can now be thought of as a subset of the worldwide SDHstandards.
The ITU-T recommendations define a number of basic transmission rates withinthe SDH. The first of these is 155.52 Mbit/s, normally referred to as �STM-1�(where STM stands for �Synchronous Transport Module�). Higher transmissionrates of STM-4, STM-16 and STM-64 (622.08 Mbit/s, 2488.32 Mbit/s and9953.28 Mbit/s respectively) are also defined, with further levels proposed forstudy.
The recommendations also define a multiplexing structure whereby an STM-1signal can carry a number of lower rate signals as a payload. Existing PDHsignals can be carried over this SDH network as a payload. This process will beexplained in more detail in the next chapter.
The Open Systems Interconnect Model
The seven layer Open Systems Interconnect (OSI) model defined by theInternational Organisation for Standardisation (ISO), is widely used to representthe functions of different network systems and applications, interoperabilitybetween equipment and interactions between networks. It can be used torepresent how these applications and protocols interact as it progresses throughthe network.
Figure 2-10 The OSI model
Within this model SDH is essentially a transport protocol and is the dominanttechnology for the physical or bearer layer. SDH can act as the physical bearerlayer for applications in layers 2 to 4 such as ATM or IP or act as layers 1 to 4and transport higher level applications directly. Its role is to manage theutilisation of the fibre infrastructure efficiently, detect failures and recover fromthem transparently to the higher layers.
One of the most common network layer protocols is IP. Using this as anexample, IP routers are layer 3 network devices which work on an end to endbasis, routing and forwarding packets to their destination. Routers willautomatically discover network topology and select the best paths, however, toensure that these end to end paths are resilient to transient failures, such as theloss of packets due to congestion, restoration based on routers is inherently slow.The network therefore relies on the underlying layers, in particular the physicalSDH layer, to detect failures and restore traffic over a backup path in a very shorttime. In the case of SDH this is typically 50 ms.
Application Layer
Presentation Layer
Session Layer
Transport Layer
Network Layer
Data Link Layer
Physical Layer
(typically IP)
(typically Frame Relay
(SDH)
SolutionCommunications
The means ofestablishing/maintaininga connection
Infrastructure
The What the PC �sees�
or ATM)
1
2
3
4
5
6
7
17 182 Transport Networks
19 18 19 18Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
SDH to the Operator
SDH is not just the current technology most common in transmission systems,it has also brought about a new dimension in the transport network itself. SDHSystems and Standards have been developed to ensure that the needs of theoperator are met. So what are these requirements which other technologiescannot meet?
Reliability
Operators rely on the protection schemes of SDH to offer their customers highavailability of service and guarantee high levels of network performance.
Growth to High Capacity and Bandwidth Management
As the operators business grows the traffic carried by an operator will grow involume and type, so the network must adapt, have a flexible architecture andmanage bandwidth to ensure efficient utilisation of the fibre infrastructure. Thearchitecture of SDH networks commonly consists of tiers or layers and thefunctionality depends on the tier. For example the highest layer is often referredto as the backbone and will carry high speed traffic over long distances. SDHtechnology now offers very high capacity systems to accommodate this trafficsuch as STM-64 bit rate and integrated Wave Division Multiplex (WDM)solutions.
Flexibility
Operators need to offer their customers flexibility by adjusting the network toaccommodate changes in traffic patterns. SDH network elements must offer avariety of functions and be easily configured. Today SDH network elements areoften universal offering add/drop multiplexing with cross-connect capabilitiesto allow the routing and grooming of traffic.
Synchronisation
Network operators must deliver synchronised timing to all the nodes in thenetwork to ensure that information passed from one node to another is not lost.Synchronisation is of growing concern to operators with advances in technologywhich are increasingly sensitive to timing (Fax, data, ATM, video compression).Synchronisation is becoming more critical as SDH provides an ideal way ofnetworking.
Network Management
Software control operators must ensure a reliable service for their customers. Itis essential that the network can monitor the performance and quality of links ata central site so that it can be rapidly adapted in the event of a failure. Asoperators gain more customers they must be able to remotely provision trafficcircuits for these customers from a remote site. Provision of software channelswithin the SDH format, allows this software control to be implemented.
Each of the topics highlighted are considered in greater detail in subsequentchapters of this book.
19 202 Transport Networks
21 20 21 20Transport Networks Transport Networks
Synchronous Transmission Systems Synchronous Transmission Systems
SDH to the Operator
SDH is not just the current technology most common in transmission systems,it has also brought about a new dimension in the transport network itself. SDHSystems and Standards have been developed to ensure that the needs of theoperator are met. So what are these requirements which other technologiescannot meet?
Reliability
Operators rely on the protection schemes of SDH to offer their customers highavailability of service and guarantee high levels of network performance.
Growth to High Capacity and Bandwidth Management
As the operators business grows the traffic carried by an operator will grow involume and type, so the network must adapt, have a flexible architecture andmanage bandwidth to ensure efficient utilisation of the fibre infrastructure. Thearchitecture of SDH networks commonly consists of tiers or layers and thefunctionality depends on the tier. For example the highest layer is often referredto as the backbone and will carry high speed traffic over long distances. SDHtechnology now offers very high capacity systems to accommodate this trafficsuch as STM-64 bit rate and integrated Wave Division Multiplex (WDM)solutions.
Flexibility
Operators need to offer their customers flexibility by adjusting the network toaccommodate changes in traffic patterns. SDH network elements must offer avariety of functions and be easily configured. Today SDH network elements areoften universal offering add/drop multiplexing with cross-connect capabilitiesto allow the routing and grooming of traffic.
Synchronisation
Network operators must deliver synchronised timing to all the nodes in thenetwork to ensure that information passed from one node to another is not lost.Synchronisation is of growing concern to operators with advances in technologywhich are increasingly sensitive to timing (Fax, data, ATM, video compression).Synchronisation is becoming more critical as SDH provides an ideal way ofnetworking.
Network Management
Software control operators must ensure a reliable service for their customers. Itis essential that the network can monitor the performance and quality of links ata central site so that it can be rapidly adapted in the event of a failure. Asoperators gain more customers they must be able to remotely provision trafficcircuits for these customers from a remote site. Provision of software channelswithin the SDH format, allows this software control to be implemented.
Each of the topics highlighted are considered in greater detail in subsequentchapters of this book.
19 202 Transport Networks
21 20 21 20Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems
Basics of the SynchronousDigital Hierarchy
SDH and the North American equivalent, SONET are the dominanttechnologies in the physical transport layer of todays optical fibre networks.Their role is to transport and manage many traffic types over the physicalinfrastructure. The protocol for this is defined by the SDH standards and theterms �SDH transmission� and �SDH transport� are in common use to describetraffic conforming to these standards.
With reference to the seven layer OSI model outlined in Chapter 2, SDH iscommonly viewed as layer 1, the physical transport layer protocol. In this roleit acts as the physical bearer for applications in layers 2 to 4, that is, it is the wayin which traffic in higher layers such as ATM and IP is transported. In simpleterms consider SDH transmission as a �pipe� which carries traffic in the form of�packages of information�. These packages are the applications such as PDH,ATM or IP.
Figure 3-1 SDH Transmission ‘Pipe’
SDH allows for the transport of many different types of traffic; voice, video,multimedia and packet based data applications such as IP. Its role, however, isessentially the same, to manage the utilisation of the fibre infrastructure. Thismeans managing the bandwidth efficiently whilst carrying a variety of traffictypes, to detect failures and recover from them transparently to the higher layers.
The aim of the first part of this chapter is to provide those new to the subject ofSDH and transmission with an explanation of the terms and concepts commonlyused in SDH.
The second part of the chapter will explain the SDH multiplexing structure insimple terms, that is how and in what form information is conveyed over theSDH hierarchy.
Terms and concepts
Basic Terminology
A Transport Network can be viewed as the links and associated equipmentwhich enable traffic to be carried between customers or nodes in a network. Atthese nodes a higher layer function may be carried out, such as switching orrouting. Remember that many links may connect different nodes or customersand these may share the same particular physical transport link. Capacity on thistransport link can be reserved. For example, an end customer may lease a certainamount of dedicated capacity from the operator, termed a leased line or privatecircuit.
Time Division Multiplexing (TDM) has been defined in Chapter 2. In simpleterms several digital streams of information are combined by taking a byte ofinformation from each incoming channel in turn and transmitting these signalson the same channel at a higher speed. This method is used to combine 64 kbit/sdigital channels into a single high speed bit stream (usually 2 Mbit/s).
Network elements are equipments located at each node in the SDH transportnetwork, which perform some function on the traffic carried such asmultiplexing or routing. Chapter 4 describes network elements and theirfunctions in more detail.
3
ATMTraffic
PDHTraffic
PDHTraffic
IPTraffic
SDH PIPE
21 223 Basics of the Synchronous Digital Hierarchy
23 22 23 22Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems
Basics of the SynchronousDigital Hierarchy
SDH and the North American equivalent, SONET are the dominanttechnologies in the physical transport layer of todays optical fibre networks.Their role is to transport and manage many traffic types over the physicalinfrastructure. The protocol for this is defined by the SDH standards and theterms �SDH transmission� and �SDH transport� are in common use to describetraffic conforming to these standards.
With reference to the seven layer OSI model outlined in Chapter 2, SDH iscommonly viewed as layer 1, the physical transport layer protocol. In this roleit acts as the physical bearer for applications in layers 2 to 4, that is, it is the wayin which traffic in higher layers such as ATM and IP is transported. In simpleterms consider SDH transmission as a �pipe� which carries traffic in the form of�packages of information�. These packages are the applications such as PDH,ATM or IP.
Figure 3-1 SDH Transmission ‘Pipe’
SDH allows for the transport of many different types of traffic; voice, video,multimedia and packet based data applications such as IP. Its role, however, isessentially the same, to manage the utilisation of the fibre infrastructure. Thismeans managing the bandwidth efficiently whilst carrying a variety of traffictypes, to detect failures and recover from them transparently to the higher layers.
The aim of the first part of this chapter is to provide those new to the subject ofSDH and transmission with an explanation of the terms and concepts commonlyused in SDH.
The second part of the chapter will explain the SDH multiplexing structure insimple terms, that is how and in what form information is conveyed over theSDH hierarchy.
Terms and concepts
Basic Terminology
A Transport Network can be viewed as the links and associated equipmentwhich enable traffic to be carried between customers or nodes in a network. Atthese nodes a higher layer function may be carried out, such as switching orrouting. Remember that many links may connect different nodes or customersand these may share the same particular physical transport link. Capacity on thistransport link can be reserved. For example, an end customer may lease a certainamount of dedicated capacity from the operator, termed a leased line or privatecircuit.
Time Division Multiplexing (TDM) has been defined in Chapter 2. In simpleterms several digital streams of information are combined by taking a byte ofinformation from each incoming channel in turn and transmitting these signalson the same channel at a higher speed. This method is used to combine 64 kbit/sdigital channels into a single high speed bit stream (usually 2 Mbit/s).
Network elements are equipments located at each node in the SDH transportnetwork, which perform some function on the traffic carried such asmultiplexing or routing. Chapter 4 describes network elements and theirfunctions in more detail.
3
ATMTraffic
PDHTraffic
PDHTraffic
IPTraffic
SDH PIPE
21 223 Basics of the Synchronous Digital Hierarchy
23 22 23 22Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
A tributary is a stream of traffic which is combined along with other tributarystreams by the multiplexing function to give a smaller number of output trafficstreams, see Figure 3-2.
An aggregate signal is the term associated with the output traffic stream, seeFigure 3-2.
.
Figure 3-2 Multiplexing Function
The tributaries of an SDH network element are the traffic interfaces into theSDH network. SDH network elements support many non-SDH tributariesenabling the efficient transport of many traffic types. For example at the loweror access layer of the network, an SDH network element may offer some of thefollowing tributaries to carry traffic directly:
� PDH rate interfaces such as 2 Mbit/s, 34 Mbit/s and 140 Mbit/s
� Analogue voice interfaces
� Ethernet interfaces to take in direct IP and data from LANs
� ISDN/ HDSL (High-speed Digital Subscriber Line) interfaces
The Synchronous Transport Module
So how are these tributary signals conveyed in SDH? The following sectionshows how information is packaged into a synchronous transport module inorder that it can be transported and managed across the network.
A Container is the basic element of the SDH signal. It consists of the bytes ofinformation from a PDH signal which are packaged into a container. There areseveral types of container, each corresponding to a to a PDH signal rate.
The Path Overhead. Each container has some control information associatedwith it. This information is generated at the originating node of the path and isterminated at the termination node of the path. The information allows theoperator to label the traffic so as to trace the signal through the network andidentify it for protection and performance monitoring purposes.
Virtual Container refers to the package of the container and its associated pathoverhead. Returning to the �pipe� analogy, the virtual container can be viewed asthe PDH traffic package which is carried along the SDH �pipe� .
There are different types of Virtual Container or VC. A VC-12 is built up of aC-12 container which contains a 2 Mbit/s PDH signal, a VC-3 carries a C-3container which contains a 34 Mbit/s PDH signal and a VC-4 carries a140 Mbit/s PDH signal in a C-4 container. This is described in detail in thefollowing sections.
SDH(e.g. STM-1)
PDHsignals
(e.g. 2, 34 M)
IP LANtraffic
ATM
aggregate trafficSTM-n (e.g. STM-4)
aggregate trafficSTM-n (e.g. STM-4)
Tributary traffic
Container
Container
Path overhead
ContainerPathovhd
Virtual Container
23 243 Basics of the Synchronous Digital Hierarchy
25 24 25 24Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
A tributary is a stream of traffic which is combined along with other tributarystreams by the multiplexing function to give a smaller number of output trafficstreams, see Figure 3-2.
An aggregate signal is the term associated with the output traffic stream, seeFigure 3-2.
.
Figure 3-2 Multiplexing Function
The tributaries of an SDH network element are the traffic interfaces into theSDH network. SDH network elements support many non-SDH tributariesenabling the efficient transport of many traffic types. For example at the loweror access layer of the network, an SDH network element may offer some of thefollowing tributaries to carry traffic directly:
� PDH rate interfaces such as 2 Mbit/s, 34 Mbit/s and 140 Mbit/s
� Analogue voice interfaces
� Ethernet interfaces to take in direct IP and data from LANs
� ISDN/ HDSL (High-speed Digital Subscriber Line) interfaces
The Synchronous Transport Module
So how are these tributary signals conveyed in SDH? The following sectionshows how information is packaged into a synchronous transport module inorder that it can be transported and managed across the network.
A Container is the basic element of the SDH signal. It consists of the bytes ofinformation from a PDH signal which are packaged into a container. There areseveral types of container, each corresponding to a to a PDH signal rate.
The Path Overhead. Each container has some control information associatedwith it. This information is generated at the originating node of the path and isterminated at the termination node of the path. The information allows theoperator to label the traffic so as to trace the signal through the network andidentify it for protection and performance monitoring purposes.
Virtual Container refers to the package of the container and its associated pathoverhead. Returning to the �pipe� analogy, the virtual container can be viewed asthe PDH traffic package which is carried along the SDH �pipe� .
There are different types of Virtual Container or VC. A VC-12 is built up of aC-12 container which contains a 2 Mbit/s PDH signal, a VC-3 carries a C-3container which contains a 34 Mbit/s PDH signal and a VC-4 carries a140 Mbit/s PDH signal in a C-4 container. This is described in detail in thefollowing sections.
SDH(e.g. STM-1)
PDHsignals
(e.g. 2, 34 M)
IP LANtraffic
ATM
aggregate trafficSTM-n (e.g. STM-4)
aggregate trafficSTM-n (e.g. STM-4)
Tributary traffic
Container
Container
Path overhead
ContainerPathovhd
Virtual Container
23 243 Basics of the Synchronous Digital Hierarchy
25 24 25 24Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
Nesting: a virtual container can contain other virtual containers and this isreferred to as nesting. For example a VC-4 can be packaged with 63 VC-12s.This simplifies the transport and management of this signal across the network.
Synchronous Transport Modules: a signal is packaged in a virtual container,but how is it actually transported on the transport link? To actually carry thevirtual containers over the network several of them are placed in a SynchronousTransport Module or STM.
The virtual containers are placed in the payload area of the STM. Going backto the initial analogy, the STMs can be visualised as pipes which make up thenetwork and the virtual container as the packages which are carried through thepipes.
The basic unit of SDH is the STM-1 frame. Four STM-1 frames are interleavedor multiplexed to give an STM-4 frame which has a faster rate of transmission.STM-16 and STM-64 offer even higher rates and so transport a greater numberof signals in a payload. So STM-4, STM-16 and STM-64 can be visualised asincreasingly �fatter pipes� as shown in Figure 3-3.
Figure 3-3 STM Traffic ‘Pipes’
Section overhead: information bytes are added to the STM frame providing acommunication channel between adjacent nodes enabling control oftransmission over the link. It allows the two nodes to �talk� to each other so thatin the event of a section failure, for example, protection switching occurs.
VC-12VC-12VC-12
VC-12VC-12
VC-12
VC-4
Virtual Containers
Payload Area
Synchronous Transport Module
STM-1
STM-4
STM-64
Virtual Containers
Payload Area
SectionOverhead
Synchronous Transport Module
25 263 Basics of the Synchronous Digital Hierarchy
27 26 27 26Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
Nesting: a virtual container can contain other virtual containers and this isreferred to as nesting. For example a VC-4 can be packaged with 63 VC-12s.This simplifies the transport and management of this signal across the network.
Synchronous Transport Modules: a signal is packaged in a virtual container,but how is it actually transported on the transport link? To actually carry thevirtual containers over the network several of them are placed in a SynchronousTransport Module or STM.
The virtual containers are placed in the payload area of the STM. Going backto the initial analogy, the STMs can be visualised as pipes which make up thenetwork and the virtual container as the packages which are carried through thepipes.
The basic unit of SDH is the STM-1 frame. Four STM-1 frames are interleavedor multiplexed to give an STM-4 frame which has a faster rate of transmission.STM-16 and STM-64 offer even higher rates and so transport a greater numberof signals in a payload. So STM-4, STM-16 and STM-64 can be visualised asincreasingly �fatter pipes� as shown in Figure 3-3.
Figure 3-3 STM Traffic ‘Pipes’
Section overhead: information bytes are added to the STM frame providing acommunication channel between adjacent nodes enabling control oftransmission over the link. It allows the two nodes to �talk� to each other so thatin the event of a section failure, for example, protection switching occurs.
VC-12VC-12VC-12
VC-12VC-12
VC-12
VC-4
Virtual Containers
Payload Area
Synchronous Transport Module
STM-1
STM-4
STM-64
Virtual Containers
Payload Area
SectionOverhead
Synchronous Transport Module
25 263 Basics of the Synchronous Digital Hierarchy
27 26 27 26Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
Nesting: a virtual container can contain other virtual containers and this isreferred to as nesting. For example a VC-4 can be packaged with 63 VC-12s.This simplifies the transport and management of this signal across the network.
Synchronous Transport Modules: a signal is packaged in a virtual container,but how is it actually transported on the transport link? To actually carry thevirtual containers over the network several of them are placed in a SynchronousTransport Module or STM.
The virtual containers are placed in the payload area of the STM. Going backto the initial analogy, the STMs can be visualised as pipes which make up thenetwork and the virtual container as the packages which are carried through thepipes.
The basic unit of SDH is the STM-1 frame. Four STM-1 frames are interleavedor multiplexed to give an STM-4 frame which has a faster rate of transmission.STM-16 and STM-64 offer even higher rates and so transport a greater numberof signals in a payload. So STM-4, STM-16 and STM-64 can be visualised asincreasingly �fatter pipes� as shown in Figure 3-3.
Figure 3-3 STM Traffic ‘Pipes’
Section overhead: information bytes are added to the STM frame providing acommunication channel between adjacent nodes enabling control oftransmission over the link. It allows the two nodes to �talk� to each other so thatin the event of a section failure, for example, protection switching occurs.
VC-12VC-12VC-12
VC-12VC-12
VC-12
VC-4
Virtual Containers
Payload Area
Synchronous Transport Module
STM-1
STM-4
STM-64
Virtual Containers
Payload Area
SectionOverhead
Synchronous Transport Module
25 26 Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
A Path or Trail is used to refer to the end to end circuit for the traffic, this is theroute taken by a virtual container across the network.
A Section is defined as the transport link between two adjacent nodes. A pathis comprised of a number of sections.
Figure 3-4 shows the relationship between path and sections.
Figure 3-4 Paths and Sections
Going back to our initial analogy of a �pipe�, the sections can be visualised as thelengths of �pipe� between the network element nodes and the path as the routethe virtual container �packages� take over these sections of pipe. The end userstraffic will be transported in virtual container �packages� on a certain path, overseveral �pipe sections. (This is a simplistic definition, in fact paths and sectionsare different layers of the transport network, this is explained in Chapter 5.)
An STM is dedicated to a single section, hence the section overhead is processedin each node and a new STM with new overhead is constructed for the nextsection. The virtual container in contrast follows a path over several sections, sothe path overhead remains with the container on its end to end path.
Example of SDH Multiplexing
Summarising the previous section, information will enter the SDH network as adigital stream of information. Information from this signal is mapped into acontainer, each container then has some control information added, known asthe path overhead. The combination of this signal and the overhead is referred toas a virtual container. The virtual containers form the payload of theSynchronous Transport Module (STM) which also has control informationcalled the section overhead.
Figure 3-5 Example of SDH Multiplexing
Section
Section Section
Path
SDH LocalNetwork
SDH RegionalNetwork
SDH BackboneNetwork
London Edinburgh2 Mbit/s
STM-1
S T M -1
STM-4
STM-16/64
SDH RegionalNetworkSTM-4
SDH LocalNetworkSTM-1
VC-12
S T M -4
S TM -16/64
2 Mbit/s
27 283 Basics of the Synchronous Digital Hierarchy
29 28 29 28Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
A Path or Trail is used to refer to the end to end circuit for the traffic, this is theroute taken by a virtual container across the network.
A Section is defined as the transport link between two adjacent nodes. A pathis comprised of a number of sections.
Figure 3-4 shows the relationship between path and sections.
Figure 3-4 Paths and Sections
Going back to our initial analogy of a �pipe�, the sections can be visualised as thelengths of �pipe� between the network element nodes and the path as the routethe virtual container �packages� take over these sections of pipe. The end userstraffic will be transported in virtual container �packages� on a certain path, overseveral �pipe sections. (This is a simplistic definition, in fact paths and sectionsare different layers of the transport network, this is explained in Chapter 5.)
An STM is dedicated to a single section, hence the section overhead is processedin each node and a new STM with new overhead is constructed for the nextsection. The virtual container in contrast follows a path over several sections, sothe path overhead remains with the container on its end to end path.
Example of SDH Multiplexing
Summarising the previous section, information will enter the SDH network as adigital stream of information. Information from this signal is mapped into acontainer, each container then has some control information added, known asthe path overhead. The combination of this signal and the overhead is referred toas a virtual container. The virtual containers form the payload of theSynchronous Transport Module (STM) which also has control informationcalled the section overhead.
Figure 3-5 Example of SDH Multiplexing
Section
Section Section
Path
SDH LocalNetwork
SDH RegionalNetwork
SDH BackboneNetwork
London Edinburgh2 Mbit/s
STM-1
S T M -1
STM-4
STM-16/64
SDH RegionalNetworkSTM-4
SDH LocalNetworkSTM-1
VC-12
S T M -4
S TM -16/64
2 Mbit/s
27 283 Basics of the Synchronous Digital Hierarchy
29 28 29 28Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
Referring to Figure 3-5:
(1) Information enters the network as a 2 Mbit/s digital stream. It will beaccommodated in a VC-12 virtual container.
(2) An SDH network element will multiplex this signal along with othertributary signals into a faster rate aggregate signal. In this example this is anSTM-1 signal at 155 Mbit/s. This is in the SDH local network.
(3) This signal can then be further multiplexed to give an STM-4 signal at622 Mbit/s in the next layer and so on up to STM-64 when it is carried at10 GBit/s. At this fast rate many signals can be transported in a single fibre, thisis referred to as the backbone network and will transport the information toanywhere in the country.
(4) The 2 Mbit/s signal can be extracted and delivered to its destination or if thedestination is a terminating equipment, the aggregate signal is demultiplexed toretrieve the 2 Mbit/s signal.
The SDH multiplexing structure defines the standard routes for mapping thesignal containers into an STM, the basic unit of which is an STM-1 frame(155 Mbit/s). A number of other basic transmission rates are defined, related bya multiplication factor of four. These are 622 Mbit/s known as STM-4,2.5 Gbit/s known as STM-16 and 10 Gbit/s or STM-64, with further levelsproposed for study
But why increase the rate of transmission from STM-1 (155 Mbit/s) to STM-16(2.5 Gbit/s) or STM-64 (10 Gbit/s)? To transport information from one end ofthe country to another requires a fibre from one end to the other. Installing fibreis very expensive, so to limit the fibre deployed, it is important to carry as muchinformation on one fibre as possible, that is, transport at a faster rate.
Explanation of the Multiplexing Structure
The SDH multiplexing structure defines how information is packaged to buildan STM-1 frame. This route for mapping containers into an STM-N signal isdefined in the ITU-T recommendations. An overview follows, for a moredetailed description refer to Chapter 5.
We have said that virtual containers are packaged into STMs by networkelements. For the network element at the far end to extract a virtual container itmust know the exact location of the virtual container within the payload of theSTM. A pointer denotes this location. In a synchronous network all theequipment is synchronised to an overall network clock. The timing of theplesiochronous signals that make up the virtual container may vary in frequencyand/or phase from the network clock. As a result, the location of a virtualcontainer within the STM frame may not be fixed, so a pointer is associated witheach virtual container to indicate its position within the STM payload.
SDH Frame: STM-1, the basic element of SDH, comprises 2430 bytes ofinformation. This is arranged in an array of 270 columns by 9 rows. Within thisis contained the STM-1 payload, pointers and section overhead.
The make up of the STM payload is defined by the SDH mapping structure.Customers transmission rates are mapped into containers (C) and a pathoverhead (POH) added to give a virtual container (VC). These are formed intotributary units (TU) which consist of the virtual container plus pointer. Thepointer indicates the position of virtual container within the tributary unit.
1
2
3
4
5
6
7
8
9
AU Pointer(s)
STM-1 Payload 9 Rows
(RSOH)
Section
Overhead
(MSOH)
Section
Overhead
270 Columns (Bytes)
29 303 Basics of the Synchronous Digital Hierarchy
31 30 31 30Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
Referring to Figure 3-5:
(1) Information enters the network as a 2 Mbit/s digital stream. It will beaccommodated in a VC-12 virtual container.
(2) An SDH network element will multiplex this signal along with othertributary signals into a faster rate aggregate signal. In this example this is anSTM-1 signal at 155 Mbit/s. This is in the SDH local network.
(3) This signal can then be further multiplexed to give an STM-4 signal at622 Mbit/s in the next layer and so on up to STM-64 when it is carried at10 GBit/s. At this fast rate many signals can be transported in a single fibre, thisis referred to as the backbone network and will transport the information toanywhere in the country.
(4) The 2 Mbit/s signal can be extracted and delivered to its destination or if thedestination is a terminating equipment, the aggregate signal is demultiplexed toretrieve the 2 Mbit/s signal.
The SDH multiplexing structure defines the standard routes for mapping thesignal containers into an STM, the basic unit of which is an STM-1 frame(155 Mbit/s). A number of other basic transmission rates are defined, related bya multiplication factor of four. These are 622 Mbit/s known as STM-4,2.5 Gbit/s known as STM-16 and 10 Gbit/s or STM-64, with further levelsproposed for study
But why increase the rate of transmission from STM-1 (155 Mbit/s) to STM-16(2.5 Gbit/s) or STM-64 (10 Gbit/s)? To transport information from one end ofthe country to another requires a fibre from one end to the other. Installing fibreis very expensive, so to limit the fibre deployed, it is important to carry as muchinformation on one fibre as possible, that is, transport at a faster rate.
Explanation of the Multiplexing Structure
The SDH multiplexing structure defines how information is packaged to buildan STM-1 frame. This route for mapping containers into an STM-N signal isdefined in the ITU-T recommendations. An overview follows, for a moredetailed description refer to Chapter 5.
We have said that virtual containers are packaged into STMs by networkelements. For the network element at the far end to extract a virtual container itmust know the exact location of the virtual container within the payload of theSTM. A pointer denotes this location. In a synchronous network all theequipment is synchronised to an overall network clock. The timing of theplesiochronous signals that make up the virtual container may vary in frequencyand/or phase from the network clock. As a result, the location of a virtualcontainer within the STM frame may not be fixed, so a pointer is associated witheach virtual container to indicate its position within the STM payload.
SDH Frame: STM-1, the basic element of SDH, comprises 2430 bytes ofinformation. This is arranged in an array of 270 columns by 9 rows. Within thisis contained the STM-1 payload, pointers and section overhead.
The make up of the STM payload is defined by the SDH mapping structure.Customers transmission rates are mapped into containers (C) and a pathoverhead (POH) added to give a virtual container (VC). These are formed intotributary units (TU) which consist of the virtual container plus pointer. Thepointer indicates the position of virtual container within the tributary unit.
1
2
3
4
5
6
7
8
9
AU Pointer(s)
STM-1 Payload 9 Rows
(RSOH)
Section
Overhead
(MSOH)
Section
Overhead
270 Columns (Bytes)
29 303 Basics of the Synchronous Digital Hierarchy
31 30 31 30Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
The tributary units are packed into tributary unit groups (TUGs) andadministrative unit groups (AUGs) according to the SDH multiplexingstructure rules shown in Figure 3-6. Note that this sequential packaging canresult in the nesting of smaller virtual containers within larger ones.
Figure 3-6 SDH Multiplexing Structure
Network Simplification. The SDH multiplexing rules ensure that the exactposition of the virtual container within the payload can be identified at eachnode. This has the advantage that each node can directly access any virtualcontainer in the payload without the need to unpack and pack the whole payload.The �multiplexer mountains� which were a feature of PDH networks are nolonger required.
Using the Multiplex Structure. Following the multiplexing structure anSTM-1 signal can be formed in different ways. The VC-4 which will form the
payload of the STM frame could contain: one 140 Mbit/s signals, three34 Mbit/s signals, 63 2 Mbit/s signals or combinations of the above, providedthe overall capacity is not exceeded.
Higher order rates: When a transmission rate higher than STM-1 is required,it is achieved by using a simple byte-interleaved multiplexing scheme to achieverates of 622 Mbit/s (STM-4), 2.5 Gbit/s (STM-16) and 10 Gbit/s STM-64). Thestandardised rates are summarised in the following table.
Introduction to Network Management
Network elements perform the multiplexing and routing of traffic from endcustomers to carry it across the network, but how does the telecom operatorcontrol this? The SDH frame structure contains channels which can be used tofully control the network remotely by software. The network can be configuredand monitored from a central point, this is the function of network management.
An operator needs to be able to provide circuits rapidly for their customers astraffic patterns change. They also need to be able to monitor these circuits and
STM-N AUG AU-4x N x 1
x 3 TUG-3
AU-3 VC-3
x 3
TU-3 VC-3
TUG-2 TU-2 VC-2
TU-12
TU-11
VC-12
VC-11
C-3
C-2
C-12
C-11
C-4
x 1
x 7
x 7
x 1
x 3
x 4
139264
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
44736
34368
6312
2048
1544
pointer processing
multiplexing
aligning
mapping
SONET specific options
ETSI specific options
VC-4 Table 3-1 SDH Transmission Rates
Bit Rate PDH Europe SDH
Name Container Transport
10 Gbit/s STM-64
2.5 GBit/s STM-16
622 MBit/s STM-4
155 MBit/s STM-1
140 MBit/s E4 VC-4
34 MBit/s E3 VC-3
8 Mbit/s E2
2 MBit/s E1 VC-12
64 kBit/s E0
31 323 Basics of the Synchronous Digital Hierarchy
33 32 33 32Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
The tributary units are packed into tributary unit groups (TUGs) andadministrative unit groups (AUGs) according to the SDH multiplexingstructure rules shown in Figure 3-6. Note that this sequential packaging canresult in the nesting of smaller virtual containers within larger ones.
Figure 3-6 SDH Multiplexing Structure
Network Simplification. The SDH multiplexing rules ensure that the exactposition of the virtual container within the payload can be identified at eachnode. This has the advantage that each node can directly access any virtualcontainer in the payload without the need to unpack and pack the whole payload.The �multiplexer mountains� which were a feature of PDH networks are nolonger required.
Using the Multiplex Structure. Following the multiplexing structure anSTM-1 signal can be formed in different ways. The VC-4 which will form the
payload of the STM frame could contain: one 140 Mbit/s signals, three34 Mbit/s signals, 63 2 Mbit/s signals or combinations of the above, providedthe overall capacity is not exceeded.
Higher order rates: When a transmission rate higher than STM-1 is required,it is achieved by using a simple byte-interleaved multiplexing scheme to achieverates of 622 Mbit/s (STM-4), 2.5 Gbit/s (STM-16) and 10 Gbit/s STM-64). Thestandardised rates are summarised in the following table.
Introduction to Network Management
Network elements perform the multiplexing and routing of traffic from endcustomers to carry it across the network, but how does the telecom operatorcontrol this? The SDH frame structure contains channels which can be used tofully control the network remotely by software. The network can be configuredand monitored from a central point, this is the function of network management.
An operator needs to be able to provide circuits rapidly for their customers astraffic patterns change. They also need to be able to monitor these circuits and
STM-N AUG AU-4x N x 1
x 3 TUG-3
AU-3 VC-3
x 3
TU-3 VC-3
TUG-2 TU-2 VC-2
TU-12
TU-11
VC-12
VC-11
C-3
C-2
C-12
C-11
C-4
x 1
x 7
x 7
x 1
x 3
x 4
139264
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
44736
34368
6312
2048
1544
pointer processing
multiplexing
aligning
mapping
SONET specific options
ETSI specific options
VC-4 Table 3-1 SDH Transmission Rates
Bit Rate PDH Europe SDH
Name Container Transport
10 Gbit/s STM-64
2.5 GBit/s STM-16
622 MBit/s STM-4
155 MBit/s STM-1
140 MBit/s E4 VC-4
34 MBit/s E3 VC-3
8 Mbit/s E2
2 MBit/s E1 VC-12
64 kBit/s E0
31 323 Basics of the Synchronous Digital Hierarchy
33 32 33 32Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
react should a fault occur. A single network management screen located at acentral site enables operators to control and monitor the network. Thiscentralised control can provide a great saving in time spent by maintenancepersonnel in travelling to sites and can increase the reliability of the network.
Within the SDH section overhead, channels are dedicated to networkmanagement. These channels form a part of the communication networkbetween the management system and the network elements called the DataCommunications Network or DCN.
A typical network management system consists of a hierarchy of elementmanagers at a lower layer managing a domain of the network, and networkmanagers sitting above several element managers, offering one point of accessand a central view of the network. The functionality of these layers differs.
This is a simplified introduction to network management. A more detaileddescription and discussion of related issues is given in Chapter 7.
Introduction to Protection
With the increase in the amount of traffic which can be carried over a single fibreand the increasing proportion of high value business traffic, protection isbecoming a more important issue. An operator needs to demonstrate to theircustomers that they will be able to reliably carry traffic across their network.Failures in network transport mechanisms, though rare, do occur and thenetwork can be vulnerable to natural disasters and accidents such excavation andbreakage of the optical fibre cable.
Reliability and resilience of networks is crucial. Traffic must be protected so thatthe instant a failure occurs, the traffic can be re-routed onto an alternative pathwith no disruption to service. This has lead to the increasing requirement for fastprotection switching techniques. There are several mechanisms defined instandards and these are explained in the Chapter 6.
Introduction to Equipment Standardisation
A major standards effort has supported the development of synchronoustransmission. Network operators wanted to avoid being locked into vendors’proprietary solutions. In North America, a situation existed where fibre-optictransmission equipment manufacturers had each developed their ownproprietary method of encoding signals in their systems. As a result, networkplanning and OA&M (Operations, Administration, and Maintenance) becamevery complicated, while network restoration in an emergency was at bestdifficult.
In order to move away from proprietary interfaces and achieve trueinterconnectivity between vendors, subcommittee T1X1 of the AmericanNational Standards Institute (ANSI) began work in 1985 on developing aSynchronous Optical NETwork (SONET) based on a proposal by Bellcore. In1986, the CCITT (now the ITU-T) became interested in the work being carriedout on SONET and after much debate on how to incorporate both the USA andEuropean transmission hierarchies into a single scheme, the worldwide SDHstandards were agreed and published in 1989. Since then, an ongoing standardseffort has been maintained.
33 343 Basics of the Synchronous Digital Hierarchy
35 34 35 34Basics of the Synchronous Digital Hierarchy Basics of the Synchronous Digital Hierarchy
Synchronous Transmission Systems Synchronous Transmission Systems
react should a fault occur. A single network management screen located at acentral site enables operators to control and monitor the network. Thiscentralised control can provide a great saving in time spent by maintenancepersonnel in travelling to sites and can increase the reliability of the network.
Within the SDH section overhead, channels are dedicated to networkmanagement. These channels form a part of the communication networkbetween the management system and the network elements called the DataCommunications Network or DCN.
A typical network management system consists of a hierarchy of elementmanagers at a lower layer managing a domain of the network, and networkmanagers sitting above several element managers, offering one point of accessand a central view of the network. The functionality of these layers differs.
This is a simplified introduction to network management. A more detaileddescription and discussion of related issues is given in Chapter 7.
Introduction to Protection
With the increase in the amount of traffic which can be carried over a single fibreand the increasing proportion of high value business traffic, protection isbecoming a more important issue. An operator needs to demonstrate to theircustomers that they will be able to reliably carry traffic across their network.Failures in network transport mechanisms, though rare, do occur and thenetwork can be vulnerable to natural disasters and accidents such excavation andbreakage of the optical fibre cable.
Reliability and resilience of networks is crucial. Traffic must be protected so thatthe instant a failure occurs, the traffic can be re-routed onto an alternative pathwith no disruption to service. This has lead to the increasing requirement for fastprotection switching techniques. There are several mechanisms defined instandards and these are explained in the Chapter 6.
Introduction to Equipment Standardisation
A major standards effort has supported the development of synchronoustransmission. Network operators wanted to avoid being locked into vendors’proprietary solutions. In North America, a situation existed where fibre-optictransmission equipment manufacturers had each developed their ownproprietary method of encoding signals in their systems. As a result, networkplanning and OA&M (Operations, Administration, and Maintenance) becamevery complicated, while network restoration in an emergency was at bestdifficult.
In order to move away from proprietary interfaces and achieve trueinterconnectivity between vendors, subcommittee T1X1 of the AmericanNational Standards Institute (ANSI) began work in 1985 on developing aSynchronous Optical NETwork (SONET) based on a proposal by Bellcore. In1986, the CCITT (now the ITU-T) became interested in the work being carriedout on SONET and after much debate on how to incorporate both the USA andEuropean transmission hierarchies into a single scheme, the worldwide SDHstandards were agreed and published in 1989. Since then, an ongoing standardseffort has been maintained.
33 343 Basics of the Synchronous Digital Hierarchy
35 34 35 34Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems
Elements of a SynchronousTransmission System
There are three basic functions of SDH transmission equipment; linetermination, multiplexing and cross-connection. In the past, these functionswere provided by separate pieces of equipment, but with the introduction ofSDH it is possible to combine these functions in a single network element. Thischapter will first outline these functions and their role in an SDH network, thendiscuss the types of network element.
Functionality of a Network Element
Multiplexing
In Chapter 2 the concept of multiplexing was introduced, that is the combiningof several lower speed signals into a single higher speed signal, thus ensuringmaximum utilisation of the physical infrastructure. Synchronous transmissionsystems use Time Division Multiplexing (TDM).
Line Termination/ Transmission
In one direction the digital tributary signals are terminated, multiplexed and thetransmitted as a higher bit rate signal. In the opposite direction the higher bit ratesignal is terminated, demultiplexed and the digital tributary signals
reconstituted. These are the tasks of Line Terminals. Synchronous transmissionnetworks typically use optical fibre as the physical transport links so thisrequires the termination and transmission of optical signals.
In PDH systems the termination, multiplexing and transmission tasks requiredseparate pieces of equipment, but in SDH these functions can be combined in asingle network element.
Figure 4-1 STM-4 Multiplexing Function
Cross-connection
Cross-connection in a synchronous network involves setting up semi-permanentinterconnections between different channels in a network element. This enablestraffic to be routed down to the virtual container level. If the operator needs tochange traffic circuits on the network, routing can be achieved by changingconnections. The different connection types are detailed in the subsequentsections of this chapter.
This description would seem to suggest that cross-connection is similar toswitching, but there are fundamental differences between the two. The maindifference is that a switch operates as a temporary connection which is set upunder the control of the end user, while cross-connection is a transmissiontechnique used to set up a semi-permanent connection under the control of thenetwork operator via a Network Manager. The operator will change thesesemi-permanent connections as traffic patterns change.
4
STM-1 STM-1
STM-4 onregional
tier
Tributary trafficfrom local tier
STM-1 STM-1
STM-4 onregional
tier
35 364 Elements of a Synchronous Transmission System
37 36 37 36Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems
Elements of a SynchronousTransmission System
There are three basic functions of SDH transmission equipment; linetermination, multiplexing and cross-connection. In the past, these functionswere provided by separate pieces of equipment, but with the introduction ofSDH it is possible to combine these functions in a single network element. Thischapter will first outline these functions and their role in an SDH network, thendiscuss the types of network element.
Functionality of a Network Element
Multiplexing
In Chapter 2 the concept of multiplexing was introduced, that is the combiningof several lower speed signals into a single higher speed signal, thus ensuringmaximum utilisation of the physical infrastructure. Synchronous transmissionsystems use Time Division Multiplexing (TDM).
Line Termination/ Transmission
In one direction the digital tributary signals are terminated, multiplexed and thetransmitted as a higher bit rate signal. In the opposite direction the higher bit ratesignal is terminated, demultiplexed and the digital tributary signals
reconstituted. These are the tasks of Line Terminals. Synchronous transmissionnetworks typically use optical fibre as the physical transport links so thisrequires the termination and transmission of optical signals.
In PDH systems the termination, multiplexing and transmission tasks requiredseparate pieces of equipment, but in SDH these functions can be combined in asingle network element.
Figure 4-1 STM-4 Multiplexing Function
Cross-connection
Cross-connection in a synchronous network involves setting up semi-permanentinterconnections between different channels in a network element. This enablestraffic to be routed down to the virtual container level. If the operator needs tochange traffic circuits on the network, routing can be achieved by changingconnections. The different connection types are detailed in the subsequentsections of this chapter.
This description would seem to suggest that cross-connection is similar toswitching, but there are fundamental differences between the two. The maindifference is that a switch operates as a temporary connection which is set upunder the control of the end user, while cross-connection is a transmissiontechnique used to set up a semi-permanent connection under the control of thenetwork operator via a Network Manager. The operator will change thesesemi-permanent connections as traffic patterns change.
4
STM-1 STM-1
STM-4 onregional
tier
Tributary trafficfrom local tier
STM-1 STM-1
STM-4 onregional
tier
35 364 Elements of a Synchronous Transmission System
37 36 37 36Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 4-2 Cross-connect
For example a business customer may have traffic circuits transiting the networkfrom A to B. Business operations at site D may be expanded and in the future thecustomer wants these traffic circuits instead to be transported to site D.
The cross-connect function, does not necessarily mean the need for a separatepiece of equipment. SDH cross-connect functionality can reside in almost anynetwork element, the most obvious being an add-drop multiplexer which isdescribed later.
Other Terms used for SDH Network Element Functionality:
Consolidation: is when traffic on partially filled paths may be reorganised ontoa single more heavily loaded path.
Figure 4-3 Consolidation
Grooming is when incoming traffic, which is directed towards a variety ofdestinations is reorganised. Traffic for specific destinations is reorganised ontopaths with other traffic allocated for that destination. Or traffic of a specific typesuch as ATM or data traffic from several destinations can be separated fromPSTN traffic and transported on a different path.
Figure 4-4 Grooming
Types of Connection
� Uni-directional is a one way connection through the SDH networkelement, for example send traffic only
� Bi-directional is two way connection through the network element,send and receive
� Drop and Continue is a connection where the signal is dropped to atributary at the network element but it also continues on in theaggregate signal to other network elements. This type of connectioncan be used for broadcast and protection mechanisms, see Figure 4-5and Chapter 6.
� Broadcast Connection: is a connection whereby an incoming virtualcontainer is connected to more than one outgoing virtual container. Inessence one incoming signal to the network element can betransmitted to several sites from the virtual container. This type ofconnection can be used for the broadcast of video, see Figure 4-5.
VC-12
VC-12
VC-12
(a)
(b)
SDH NetworkElement
CustomerSite A
CustomerSite D
CustomerSite C
CustomerSite B
SDHMultiplexer
Partiallyfilledtrafficpaths
GroomerMixedtrafficpaths
Groomedtrafficpaths
NetworkElement
37 384 Elements of a Synchronous Transmission System
39 38 39 38Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 4-2 Cross-connect
For example a business customer may have traffic circuits transiting the networkfrom A to B. Business operations at site D may be expanded and in the future thecustomer wants these traffic circuits instead to be transported to site D.
The cross-connect function, does not necessarily mean the need for a separatepiece of equipment. SDH cross-connect functionality can reside in almost anynetwork element, the most obvious being an add-drop multiplexer which isdescribed later.
Other Terms used for SDH Network Element Functionality:
Consolidation: is when traffic on partially filled paths may be reorganised ontoa single more heavily loaded path.
Figure 4-3 Consolidation
Grooming is when incoming traffic, which is directed towards a variety ofdestinations is reorganised. Traffic for specific destinations is reorganised ontopaths with other traffic allocated for that destination. Or traffic of a specific typesuch as ATM or data traffic from several destinations can be separated fromPSTN traffic and transported on a different path.
Figure 4-4 Grooming
Types of Connection
� Uni-directional is a one way connection through the SDH networkelement, for example send traffic only
� Bi-directional is two way connection through the network element,send and receive
� Drop and Continue is a connection where the signal is dropped to atributary at the network element but it also continues on in theaggregate signal to other network elements. This type of connectioncan be used for broadcast and protection mechanisms, see Figure 4-5and Chapter 6.
� Broadcast Connection: is a connection whereby an incoming virtualcontainer is connected to more than one outgoing virtual container. Inessence one incoming signal to the network element can betransmitted to several sites from the virtual container. This type ofconnection can be used for the broadcast of video, see Figure 4-5.
VC-12
VC-12
VC-12
(a)
(b)
SDH NetworkElement
CustomerSite A
CustomerSite D
CustomerSite C
CustomerSite B
SDHMultiplexer
Partiallyfilledtrafficpaths
GroomerMixedtrafficpaths
Groomedtrafficpaths
NetworkElement
37 384 Elements of a Synchronous Transmission System
39 38 39 38Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 4-5 Broadcast with Drop and Continue
Types of Network Element
ITU-T Recommendation G.782 identifies examples of SDH equipmentproviding combinations of SDH functions. These are classified intomultiplexers (of which there are seven variants) and cross-connects (where thereare three variants). For simplicity, three types of SDH network elements will beconsidered: line systems, add-drop multiplexers and digital cross-connects.
Line terminals
The simplest type of SDH network element is a line terminal. This willimplement only the line termination and multiplexing functions, thus theirdeployment is typically in point to point configurations. Several tributarystreams will be combined at the line terminal to give an aggregate stream at ahigher speed and this will be transmitted on an optical link. Network elementsare required at the two end points of this link and a fixed connection for customercircuits is set up between these two termination points.
Figure 4-6 Terminal Multiplexer
Add-Drop Multiplexers
Add-Drop Multiplexers (ADM) offer cross-connection as well as the linetermination and multiplexer functionality. In SDH it is possible to extract(�drop�) a virtual container and in the reverse direction insert (�add�) a virtualcontainer into the STM signal directly without unpacking the signal as explainedin Chapter 3. This fundamental advantage of synchronous systems means that itis possible to flexibly connect signals between the network element interfaces(aggregate or tributary). This routing capability allows cross-connectfunctionality to be distributed around the network rather than beingconcentrated in large dedicated cross-connects.
In the line terminal case, the links set up were fixed point to point circuits. Theadded functionality of ADMs allows a more flexible network to be set up and theroute by which customer circuits transit the network can be easily changed.
This flexibility can be demonstrated by an ADM chain network. Consider thetransport link as a bus route, at each stop (ADM) the passenger (traffic circuit)on the bus network can choose to jump off or stay on the bus.
BroadcastCentre
City A
City B
City C
Drop & Continue
Drop & Continue
STM-1
STM-1
Tributary
STM-4 aggregate trafficSTM-1STM-1
STM-1
STM-1
STM-1STM-1
traffic
Tributarytraffic
39 404 Elements of a Synchronous Transmission System
41 40 41 40Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 4-5 Broadcast with Drop and Continue
Types of Network Element
ITU-T Recommendation G.782 identifies examples of SDH equipmentproviding combinations of SDH functions. These are classified intomultiplexers (of which there are seven variants) and cross-connects (where thereare three variants). For simplicity, three types of SDH network elements will beconsidered: line systems, add-drop multiplexers and digital cross-connects.
Line terminals
The simplest type of SDH network element is a line terminal. This willimplement only the line termination and multiplexing functions, thus theirdeployment is typically in point to point configurations. Several tributarystreams will be combined at the line terminal to give an aggregate stream at ahigher speed and this will be transmitted on an optical link. Network elementsare required at the two end points of this link and a fixed connection for customercircuits is set up between these two termination points.
Figure 4-6 Terminal Multiplexer
Add-Drop Multiplexers
Add-Drop Multiplexers (ADM) offer cross-connection as well as the linetermination and multiplexer functionality. In SDH it is possible to extract(�drop�) a virtual container and in the reverse direction insert (�add�) a virtualcontainer into the STM signal directly without unpacking the signal as explainedin Chapter 3. This fundamental advantage of synchronous systems means that itis possible to flexibly connect signals between the network element interfaces(aggregate or tributary). This routing capability allows cross-connectfunctionality to be distributed around the network rather than beingconcentrated in large dedicated cross-connects.
In the line terminal case, the links set up were fixed point to point circuits. Theadded functionality of ADMs allows a more flexible network to be set up and theroute by which customer circuits transit the network can be easily changed.
This flexibility can be demonstrated by an ADM chain network. Consider thetransport link as a bus route, at each stop (ADM) the passenger (traffic circuit)on the bus network can choose to jump off or stay on the bus.
BroadcastCentre
City A
City B
City C
Drop & Continue
Drop & Continue
STM-1
STM-1
Tributary
STM-4 aggregate trafficSTM-1STM-1
STM-1
STM-1
STM-1STM-1
traffic
Tributarytraffic
39 404 Elements of a Synchronous Transmission System
41 40 41 40Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 4-7 Add/Drop Chain
In an ADM individual traffic circuits can be taken out of the aggregate streamwhile the rest of the traffic is passed along the chain. This creates a bus structure,a signal can �jump off� or �stay on� the bus at each ADM stop. Many ADMs canbe connected by the bus and the �connectivity of each ADM, that is whether thetraffic circuit is dropped at the stop or passed on can be changed. Thus a flexibleconnection between several points is created, rather than a fixed line betweenevery two points. If a customer wants to route their traffic circuit to a differentnode, the request can be carried out remotely by reconfiguring the connectionsin the ADM.
Figure 4-8 Types of Connections
Different types of multiplexer offer different levels of cross-connectivity. AnADM as described will perform the simple add-drop function whereby somevirtual containers can be dropped, others can be inserted and the remainder arepassed through unchanged. ADMs can also offer time slot interchange: whichcross-connects one virtual container from one place on the East side to adifferent place on the West side. Connections can also be made betweentributary ports, thus providing cross-connect functionality between tributaries,also referred to as �hairpinning�.
Ring Deployment of an ADM
An ADM is particularly suitable for setting up a ring network. Signals are fedinto the ring via the tributary interfaces of an ADM, then coupled into the higherrate aggregate signal of the ring for transportation to the other nodes.
Rings are a common network configuration because they can increase thesurvivability of the network. Networks can be subject to node failure or linkbreakages so resilience is required to prevent the loss of traffic.
How is this achieved? In a point to point network each link must be duplicatedto provide an alternative path for the traffic should there be a failure. In a ringnetwork traffic can simply be diverted the other way round the ring as shown inFigure 4-9. In SDH this reconfiguration can occur by action at the networkelement without the intervention of external network management.
Figure 4-9 Add/Drop Chain
or
(b)
(a)
(c)
(d)
(a) Drop traffic: �get off bus�.Aggregate to tributaryconnection.
(b) Add traffic: �get on bus�.Tributary to aggregateconnection.
(c) Pass aggregate traffic on:�stay on bus�. Aggregate toaggregate connection.
(d) Local traffic connected todifferent local path. Tributaryto tributary connection.
Fibre Ring
Tributaries
Working trafficProtection route
MuxSDH
Tributaries
MuxSDH
MuxSDH
MuxSDH
41 424 Elements of a Synchronous Transmission System
43 42 43 42Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 4-7 Add/Drop Chain
In an ADM individual traffic circuits can be taken out of the aggregate streamwhile the rest of the traffic is passed along the chain. This creates a bus structure,a signal can �jump off� or �stay on� the bus at each ADM stop. Many ADMs canbe connected by the bus and the �connectivity of each ADM, that is whether thetraffic circuit is dropped at the stop or passed on can be changed. Thus a flexibleconnection between several points is created, rather than a fixed line betweenevery two points. If a customer wants to route their traffic circuit to a differentnode, the request can be carried out remotely by reconfiguring the connectionsin the ADM.
Figure 4-8 Types of Connections
Different types of multiplexer offer different levels of cross-connectivity. AnADM as described will perform the simple add-drop function whereby somevirtual containers can be dropped, others can be inserted and the remainder arepassed through unchanged. ADMs can also offer time slot interchange: whichcross-connects one virtual container from one place on the East side to adifferent place on the West side. Connections can also be made betweentributary ports, thus providing cross-connect functionality between tributaries,also referred to as �hairpinning�.
Ring Deployment of an ADM
An ADM is particularly suitable for setting up a ring network. Signals are fedinto the ring via the tributary interfaces of an ADM, then coupled into the higherrate aggregate signal of the ring for transportation to the other nodes.
Rings are a common network configuration because they can increase thesurvivability of the network. Networks can be subject to node failure or linkbreakages so resilience is required to prevent the loss of traffic.
How is this achieved? In a point to point network each link must be duplicatedto provide an alternative path for the traffic should there be a failure. In a ringnetwork traffic can simply be diverted the other way round the ring as shown inFigure 4-9. In SDH this reconfiguration can occur by action at the networkelement without the intervention of external network management.
Figure 4-9 Add/Drop Chain
or
(b)
(a)
(c)
(d)
(a) Drop traffic: �get off bus�.Aggregate to tributaryconnection.
(b) Add traffic: �get on bus�.Tributary to aggregateconnection.
(c) Pass aggregate traffic on:�stay on bus�. Aggregate toaggregate connection.
(d) Local traffic connected todifferent local path. Tributaryto tributary connection.
Fibre Ring
Tributaries
Working trafficProtection route
MuxSDH
Tributaries
MuxSDH
MuxSDH
MuxSDH
41 424 Elements of a Synchronous Transmission System
43 42 43 42Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Hubbing Deployment
An ADM may be configured as a hub for use in multi-site network applications.the purpose of the hub is to consolidate several spurs onto the higher capacityoptical aggregate. This arrangement eliminates the cost and complexity ofmulti-terminal configurations and redundant cross-connections.
Figure 4-10 Fibre Hub
Multiplexer Types
Multiplexers can be classified in various ways, for example by the type andflexibility of connections that can be made. Multiplexers are commonlyclassified by the bit rate of the aggregate signal supported. For example an�STM-4 multiplexer� will accept tributary signals at a variety of PDH and SDHrates (2 Mbit/s, 34 Mbit/s, 45 Mbit/s, 140 Mbit/s and STM-1) and multiplexthese into an STM-4 aggregate signal.
Multiplexers may also be classified as full and partial access systems. A fullaccess ADM can access any of the traffic in its STM-N aggregate payload. Thatis, all aggregate traffic can be connected internally and all passed to the tributaryports. In contrast a partial access multiplexer is only able to access and connectto its tributary ports a portion of its aggregate traffic, the remainder of the trafficis connected straight through the multiplexer on the aggregate signal.
Multiplexers can be upgraded. This typically refers to the replacement of theaggregate ports with aggregate ports which transmit at a faster rate. For examplean STM-1 multiplexer may have its aggregate cards replaced with STM-4
aggregate cards. The aggregate signal rate of the multiplexer is increased toSTM-4, but only a portion of the aggregate traffic can be connected to thetributaries of each multiplexer. In this case the multiplexer has become a partialaccess multiplexer, see Figure 4-11.
Figure 4-11 Partial Access
The ability to upgrade the multiplexer to higher aggregate bit rate allows thenetwork operator to upgrade links to a higher speed as the traffic capacitydemand increases. Flexibility is, however, constrained because only a portion ofthe aggregate traffic can be accessed by a multiplexer. Traffic connectionsbetween aggregates and tributaries are limited and so it is more difficult toaccommodate changing traffic patterns. Some multiplexers are designed so thatthe �effective cross-connect capacity� increases, that is the connections areincreased as the multiplexer aggregate rate is upgraded.
Dedicated Cross-connects
As described above the cross-connectivity of ADMs allows thecross-connection function to be distributed throughout the network, but it is alsopossible to have stand-alone cross-connect equipment. Digital Cross-connectsor DXCs are the most complex and expensive kind of SDH equipment.
It is not the inclusion of the cross-connect functional block(s) whichdistinguishes the cross-connects from the ADMs, but the presence of the higherorder and lower order connection supervision. That is, the distinguishing featureof a DXC is its ability to provide supervision of the connections.
SDHMultiplexer
OpticalAggregates
Optical
Electrical
Tributaries
Tributaries
STM-16
Only four of the 16 VC-4scan be accessed
43 444 Elements of a Synchronous Transmission System
45 44 45 44Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Hubbing Deployment
An ADM may be configured as a hub for use in multi-site network applications.the purpose of the hub is to consolidate several spurs onto the higher capacityoptical aggregate. This arrangement eliminates the cost and complexity ofmulti-terminal configurations and redundant cross-connections.
Figure 4-10 Fibre Hub
Multiplexer Types
Multiplexers can be classified in various ways, for example by the type andflexibility of connections that can be made. Multiplexers are commonlyclassified by the bit rate of the aggregate signal supported. For example an�STM-4 multiplexer� will accept tributary signals at a variety of PDH and SDHrates (2 Mbit/s, 34 Mbit/s, 45 Mbit/s, 140 Mbit/s and STM-1) and multiplexthese into an STM-4 aggregate signal.
Multiplexers may also be classified as full and partial access systems. A fullaccess ADM can access any of the traffic in its STM-N aggregate payload. Thatis, all aggregate traffic can be connected internally and all passed to the tributaryports. In contrast a partial access multiplexer is only able to access and connectto its tributary ports a portion of its aggregate traffic, the remainder of the trafficis connected straight through the multiplexer on the aggregate signal.
Multiplexers can be upgraded. This typically refers to the replacement of theaggregate ports with aggregate ports which transmit at a faster rate. For examplean STM-1 multiplexer may have its aggregate cards replaced with STM-4
aggregate cards. The aggregate signal rate of the multiplexer is increased toSTM-4, but only a portion of the aggregate traffic can be connected to thetributaries of each multiplexer. In this case the multiplexer has become a partialaccess multiplexer, see Figure 4-11.
Figure 4-11 Partial Access
The ability to upgrade the multiplexer to higher aggregate bit rate allows thenetwork operator to upgrade links to a higher speed as the traffic capacitydemand increases. Flexibility is, however, constrained because only a portion ofthe aggregate traffic can be accessed by a multiplexer. Traffic connectionsbetween aggregates and tributaries are limited and so it is more difficult toaccommodate changing traffic patterns. Some multiplexers are designed so thatthe �effective cross-connect capacity� increases, that is the connections areincreased as the multiplexer aggregate rate is upgraded.
Dedicated Cross-connects
As described above the cross-connectivity of ADMs allows thecross-connection function to be distributed throughout the network, but it is alsopossible to have stand-alone cross-connect equipment. Digital Cross-connectsor DXCs are the most complex and expensive kind of SDH equipment.
It is not the inclusion of the cross-connect functional block(s) whichdistinguishes the cross-connects from the ADMs, but the presence of the higherorder and lower order connection supervision. That is, the distinguishing featureof a DXC is its ability to provide supervision of the connections.
SDHMultiplexer
OpticalAggregates
Optical
Electrical
Tributaries
Tributaries
STM-16
Only four of the 16 VC-4scan be accessed
43 444 Elements of a Synchronous Transmission System
45 44 45 44Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
All DXCs provide cross-connect functionality and it would be unusual to deploya DXC without full cross-connection between all inputs and outputs, whereasmany ADMs provide a limited cross-connection function. DXCs alsoincorporate those multiplexing and line termination functions which areessential for interfacing between the cross-connect matrix and the rest of thenetwork.
There are two types of dedicated SDH cross-connects, generally referred to as4/1 DXCs and 4/4 DXCs.
4/1 DXCs can usually accept combinations of 2, 155 and 622 Mbit/s inputs andcan cross-connect VC-12s, and many will also be able to cross-connect VC-2s,VC-3s and VC-4s. These more complex pieces of equipment are often referredto as 4/3/1 DXCs. 4/1 DXCs are, therefore, installed at points in the networkwhere:
� major path and circuit reorganisation is necessary � for example,between the core and regional networks
� where connection supervision is required � for example, at thegateway to another network.
4/1 cross-connects extract virtual containers from a variety of SDH links(mainly STM-1, STM-4 and STM-16) and reroute them.
- 4/4 DXCs are usually designed to accept inputs at 140, 155 or 622 Mbit/s andare optimised to switch VC-4s only. 4/4 cross-connects are core networkdevices, providing such capabilities as high level path management and networkrestoration.
Three factors limit the traffic capacity of a DXC: the number and size of thetributary ports, and the size of the internal switching core. In practice, portcapacity tends to be exhausted before core switching capacity, and is the mainreason for cross-connect upgrading.
The flexibility of DXCs means that they can be deployed in any configuration.The provision of the supervision capability, however, makes the DXC acomplex, and expensive, network element and the addition of support for selfhealing ring protocols increases the complexity. It is, therefore, more usual tobuild self healing rings with ADMs where the addition of the ring protocols isless complex since the connection supervision functions are not present.
Regenerators and Repeaters
Network elements may also be configured to extend the length of the spanbetween nodes, rather than performing the traffic interchange functions.
As signals travel along a transmission link they will degrade and noise canaccumulate. Multiplexers can be configured to regenerate the aggregate signaland optical amplifiers can be used as repeaters to re-power the optical signal.
A multiplexer configured as a regenerator converts the optical signal to anelectrical signal which is regenerated (�cleaned up�). The regenerated signal isconverted back to an optical aggregate signal and transmitted. For example, abackbone STM-16 link between two cities where STM-16 ADMs are located atthe two cities, but the intermediate span is long and the signal may be degradedto the point where the receiving ADM can no longer faithfully reconstitute thetransmitted signal. An ADM configured as a regenerator is introduced at alocation midway between the two cities to regenerate the signal and remove thepossible introduction of errors, see Figure 4-12.
Figure 4-12 Regeneration
Optical amplifiers are an alternative option to extend the reach of the opticalsignal. These operate as repeaters, re-powering the signal. No conversion to theelectrical constituent signals is undergone. Although the span is extended bypowering the signal it is not �cleaned up�, so depending on the length of the linkand type of fibre, regenerators may also be required.
SDHMux
SDHMux
Regenerator
Signaldegradation
�Clean�Signal
45 464 Elements of a Synchronous Transmission System
47 46 47 46Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
All DXCs provide cross-connect functionality and it would be unusual to deploya DXC without full cross-connection between all inputs and outputs, whereasmany ADMs provide a limited cross-connection function. DXCs alsoincorporate those multiplexing and line termination functions which areessential for interfacing between the cross-connect matrix and the rest of thenetwork.
There are two types of dedicated SDH cross-connects, generally referred to as4/1 DXCs and 4/4 DXCs.
4/1 DXCs can usually accept combinations of 2, 155 and 622 Mbit/s inputs andcan cross-connect VC-12s, and many will also be able to cross-connect VC-2s,VC-3s and VC-4s. These more complex pieces of equipment are often referredto as 4/3/1 DXCs. 4/1 DXCs are, therefore, installed at points in the networkwhere:
� major path and circuit reorganisation is necessary � for example,between the core and regional networks
� where connection supervision is required � for example, at thegateway to another network.
4/1 cross-connects extract virtual containers from a variety of SDH links(mainly STM-1, STM-4 and STM-16) and reroute them.
- 4/4 DXCs are usually designed to accept inputs at 140, 155 or 622 Mbit/s andare optimised to switch VC-4s only. 4/4 cross-connects are core networkdevices, providing such capabilities as high level path management and networkrestoration.
Three factors limit the traffic capacity of a DXC: the number and size of thetributary ports, and the size of the internal switching core. In practice, portcapacity tends to be exhausted before core switching capacity, and is the mainreason for cross-connect upgrading.
The flexibility of DXCs means that they can be deployed in any configuration.The provision of the supervision capability, however, makes the DXC acomplex, and expensive, network element and the addition of support for selfhealing ring protocols increases the complexity. It is, therefore, more usual tobuild self healing rings with ADMs where the addition of the ring protocols isless complex since the connection supervision functions are not present.
Regenerators and Repeaters
Network elements may also be configured to extend the length of the spanbetween nodes, rather than performing the traffic interchange functions.
As signals travel along a transmission link they will degrade and noise canaccumulate. Multiplexers can be configured to regenerate the aggregate signaland optical amplifiers can be used as repeaters to re-power the optical signal.
A multiplexer configured as a regenerator converts the optical signal to anelectrical signal which is regenerated (�cleaned up�). The regenerated signal isconverted back to an optical aggregate signal and transmitted. For example, abackbone STM-16 link between two cities where STM-16 ADMs are located atthe two cities, but the intermediate span is long and the signal may be degradedto the point where the receiving ADM can no longer faithfully reconstitute thetransmitted signal. An ADM configured as a regenerator is introduced at alocation midway between the two cities to regenerate the signal and remove thepossible introduction of errors, see Figure 4-12.
Figure 4-12 Regeneration
Optical amplifiers are an alternative option to extend the reach of the opticalsignal. These operate as repeaters, re-powering the signal. No conversion to theelectrical constituent signals is undergone. Although the span is extended bypowering the signal it is not �cleaned up�, so depending on the length of the linkand type of fibre, regenerators may also be required.
SDHMux
SDHMux
Regenerator
Signaldegradation
�Clean�Signal
45 464 Elements of a Synchronous Transmission System
47 46 47 46Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Submarine Systems
Short-haul unrepeatered submarine links will employ the same standards as landbased systems. In the case of long haul repeatered submarine transmissionsystems special measures need to be taken in order to maintain the very highreliability levels required for submerged repeaters.
For synchronous transmission systems the SDH specifies a relatively high levelof processing required within a repeater. This is a problem in submarine systemsbut can be overcome by wrapping the STM-1 signal within a submarine basedsignal frame which uses very simple signalling systems to reduce processingwithin repeaters and hence increase reliability.
Figure 4-13 A Synchronous Transmission Network
OpticalAmplifier
Submarine
System
MicrowaveSystem
MicrowaveSystem
SDH Cross-connection
Submarine Line Equipment
Microwave Line Equipment Synchronous Multiplexer
MicrowaveSystem
OpticalAmplifier
47 484 Elements of a Synchronous Transmission System
49 48 49 48Elements of a Synchronous Transmission System Elements of a Synchronous Transmission System
Synchronous Transmission Systems Synchronous Transmission Systems
Submarine Systems
Short-haul unrepeatered submarine links will employ the same standards as landbased systems. In the case of long haul repeatered submarine transmissionsystems special measures need to be taken in order to maintain the very highreliability levels required for submerged repeaters.
For synchronous transmission systems the SDH specifies a relatively high levelof processing required within a repeater. This is a problem in submarine systemsbut can be overcome by wrapping the STM-1 signal within a submarine basedsignal frame which uses very simple signalling systems to reduce processingwithin repeaters and hence increase reliability.
Figure 4-13 A Synchronous Transmission Network
OpticalAmplifier
Submarine
System
MicrowaveSystem
MicrowaveSystem
SDH Cross-connection
Submarine Line Equipment
Microwave Line Equipment Synchronous Multiplexer
MicrowaveSystem
OpticalAmplifier
47 484 Elements of a Synchronous Transmission System
49 48 49 48Network Architecture Network Architecture
Synchronous Transmission Systems
SDH Network Architecture and Design
This chapter provides an overview of the common styles of SDH networkarchitecture and discusses their properties.
Mapping User Connection Demand to a Physical Network
Networks can be viewed as a set of layers in which the lower layers provide aconnection service to higher level �clients�. Figure 5-1 illustrates an example ofhow a �Circuit Demand� (expressed in terms of 2 Mbit/s connectionrequirements) is mapped on to �Fibre Links� in the network.
In this example there are three stages to the mapping process:
1 The 2 Mbit/s connection requirements are expressed as VC-12connection requirements (a trivial mapping in this case).
2 The VC-12 demand must be mapped on to the VC-4 layer, that is, aset of VC-4 connections and VC-12 flexibility nodes must be definedto carry this VC-12 demand.
3 The resulting VC-4 layer must be mapped on to the physical links,that is, a set of STM-N multiplex sections and VC-4 flexibility pointsmust be defined to carry the VC-4 demand.
In other applications there may be further mapping processes. For example, atthe client level it may be necessary to map 64 kbit/s demand on to 2 Mbit/strunks and at the physical link level the STM-N multiplex sections may bemapped on to a WDM optical layer.
Figure 5-1 Mapping Demand to Physical Network
The mapping stages are the essence of the network design process. They areusually complex tasks with many choices. The choices are of two forms
� Choice of architecture - How will the links and nodes be structured toimplement at layer, for example, Star, Mesh, Ring �
� Optimisation of Design - What links and nodes will be used in thechosen architecture and how will traffic be routed.
This chapter is primarily concerned with the choice of architecture andconcentrates on the most common architectures to illustrate keys points.
5
A
F
B
E
C
D
G
Circuit Demand
Fibre Links
AB
F
DE
G C
Circuit Demand
VC12 Layer
VC4 Layer
Sections/Links
NE
TW
OR
K D
ES
IGN
49 505 Network Architecture
51 50 51 50Network Architecture Network Architecture
Synchronous Transmission Systems
SDH Network Architecture and Design
This chapter provides an overview of the common styles of SDH networkarchitecture and discusses their properties.
Mapping User Connection Demand to a Physical Network
Networks can be viewed as a set of layers in which the lower layers provide aconnection service to higher level �clients�. Figure 5-1 illustrates an example ofhow a �Circuit Demand� (expressed in terms of 2 Mbit/s connectionrequirements) is mapped on to �Fibre Links� in the network.
In this example there are three stages to the mapping process:
1 The 2 Mbit/s connection requirements are expressed as VC-12connection requirements (a trivial mapping in this case).
2 The VC-12 demand must be mapped on to the VC-4 layer, that is, aset of VC-4 connections and VC-12 flexibility nodes must be definedto carry this VC-12 demand.
3 The resulting VC-4 layer must be mapped on to the physical links,that is, a set of STM-N multiplex sections and VC-4 flexibility pointsmust be defined to carry the VC-4 demand.
In other applications there may be further mapping processes. For example, atthe client level it may be necessary to map 64 kbit/s demand on to 2 Mbit/strunks and at the physical link level the STM-N multiplex sections may bemapped on to a WDM optical layer.
Figure 5-1 Mapping Demand to Physical Network
The mapping stages are the essence of the network design process. They areusually complex tasks with many choices. The choices are of two forms
� Choice of architecture - How will the links and nodes be structured toimplement at layer, for example, Star, Mesh, Ring �
� Optimisation of Design - What links and nodes will be used in thechosen architecture and how will traffic be routed.
This chapter is primarily concerned with the choice of architecture andconcentrates on the most common architectures to illustrate keys points.
5
A
F
B
E
C
D
G
Circuit Demand
Fibre Links
AB
F
DE
G C
Circuit Demand
VC12 Layer
VC4 Layer
Sections/Links
NE
TW
OR
K D
ES
IGN
49 505 Network Architecture
51 50 51 50Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
The most essential consideration when designing a network today, is the rapidtraffic growth that is taking place. It is vital that the networks being deployed arescaleable and can be upgraded without the need for costly re-design.
The mapping of the VC-4 layer on to fibre links is examined first since this ismost closely related to the physical network.
Types of VC4 Layer Architecture
This section presents a number of VC4 layer architectures.
Flat VC4 Mesh Architecture
The physical network comprises a mesh of fibre links between nodes. Thesimplest and most obvious architecture is a VC-4 routing mesh that correspondsexactly to the physical network, as shown in Figure 5-2. The links in the meshare multiplex sections corresponding to the fibre links. The nodes are shown ascircles and comprise a VC-4 cross-connect and line terminals that feed eachincident fibre. Nodes with only two incident links such as �B� and �H�, wheretraffic is dropped from and inserted into the signal, can be implemented by anAdd/Drop Multiplexer or back to back line terminals, rather than a full cross-connect and line system.
The properties of this architecture are:Figure 5-2 Flat VC-4 Mesh Architecture
Protection by simple end-to-end diversity presents several problems.
� Planning for diversity is difficult in a large network. The networkshown in Figure 5-2 can obviously support diverse connectionsbetween each node pair, but if the working path from A-M is vianodes I,J,K this prevents a alternate diverse path from A-M beingrouted. The solution, as shown in Figure 5-2, in this simple networkis obvious, but it becomes much harder to plan diversity in largenetworks and complex routing algorithms must be employed.Planning can be simplified if the network can be broken up intosubnetworks. The mesh architecture, however, does not exhibit anynatural subnetwork boundaries and imposing them limits routingflexibility eroding one of the strengths of a mesh.
� It is difficult to keep track of both working and protection paths andto ensure that diversity is maintained during any networkrearrangement.
Demand Mapping Simple and direct
Routing Flexible, but routes pass through every intermediate node.
Nodes Complex - Multiple line terminals and cross-connect
Subnetworking Difficult to segment into subnetworks, no �natural� boundaries
Protection method End to End path diversity, requires careful planning to ensure path diversity.
Protection overhead 2.5 x equipment & link usage (see text below)
Demand Growth In general growth in one demand requires several links and nodes to grow (this is generally true of all architectures).
AB
CD
E
FG
H
J K
LM
4/4
4/4
Cross-connect Node
ADM Node
Line Terminal
VC4 (4/4)Cross-connect
I
51 525 Network Architecture
53 52 53 52Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
The most essential consideration when designing a network today, is the rapidtraffic growth that is taking place. It is vital that the networks being deployed arescaleable and can be upgraded without the need for costly re-design.
The mapping of the VC-4 layer on to fibre links is examined first since this ismost closely related to the physical network.
Types of VC4 Layer Architecture
This section presents a number of VC4 layer architectures.
Flat VC4 Mesh Architecture
The physical network comprises a mesh of fibre links between nodes. Thesimplest and most obvious architecture is a VC-4 routing mesh that correspondsexactly to the physical network, as shown in Figure 5-2. The links in the meshare multiplex sections corresponding to the fibre links. The nodes are shown ascircles and comprise a VC-4 cross-connect and line terminals that feed eachincident fibre. Nodes with only two incident links such as �B� and �H�, wheretraffic is dropped from and inserted into the signal, can be implemented by anAdd/Drop Multiplexer or back to back line terminals, rather than a full cross-connect and line system.
The properties of this architecture are:Figure 5-2 Flat VC-4 Mesh Architecture
Protection by simple end-to-end diversity presents several problems.
� Planning for diversity is difficult in a large network. The networkshown in Figure 5-2 can obviously support diverse connectionsbetween each node pair, but if the working path from A-M is vianodes I,J,K this prevents a alternate diverse path from A-M beingrouted. The solution, as shown in Figure 5-2, in this simple networkis obvious, but it becomes much harder to plan diversity in largenetworks and complex routing algorithms must be employed.Planning can be simplified if the network can be broken up intosubnetworks. The mesh architecture, however, does not exhibit anynatural subnetwork boundaries and imposing them limits routingflexibility eroding one of the strengths of a mesh.
� It is difficult to keep track of both working and protection paths andto ensure that diversity is maintained during any networkrearrangement.
Demand Mapping Simple and direct
Routing Flexible, but routes pass through every intermediate node.
Nodes Complex - Multiple line terminals and cross-connect
Subnetworking Difficult to segment into subnetworks, no �natural� boundaries
Protection method End to End path diversity, requires careful planning to ensure path diversity.
Protection overhead 2.5 x equipment & link usage (see text below)
Demand Growth In general growth in one demand requires several links and nodes to grow (this is generally true of all architectures).
AB
CD
E
FG
H
J K
LM
4/4
4/4
Cross-connect Node
ADM Node
Line Terminal
VC4 (4/4)Cross-connect
I
51 525 Network Architecture
53 52 53 52Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
� Simple end-to-end path diversity yields poor availability in largenetworks; typically more than 300 km path length.
� The protection path is always longer than the working path.Protection by path diversity, therefore, requires more than twice theresources that are used by the working path. Typically the protectionand working paths together will need 2.5 times the resources of asimple working path.
Hierarchical or Express VC-4 Mesh Architecture
One of the problems of the flat mesh architecture is that routes pass throughevery intermediate node. A node that only contributes a small amount of trafficmight still have large amounts of equipment just to handle the traffic in transit.
This can be avoided by the use of �Express� routes that link the larger nodes andbypass intermediate nodes as shown by the blue lines in Figure 5-3. The resultis a more efficient mesh architecture and in general a more manageable one.
Figure 5-3 Express Links
The �ad hoc� solution of using express links can be extended to the point wherethe express links and nodes form an new layer in the mesh architecture as shownin Figure 5-4. The mesh has been divided into a �Core Network� layer of highcapacity links between a few key nodes and a lower capacity feeder mesh layerconnecting the remaining nodes. It may then also become natural to partition thefeeder mesh into different regional subnetworks which interconnect to the corenetwork at major nodes.
Figure 5-4 Express Link Core Network
The hierarchical mesh architecture is a more efficient implementation of a meshbut some of the intrinsic routing flexibility is lost. This may not be a such badthing, since routing becomes more structured and easier to manage. It alsoprovides natural boundaries for subnetworks. This allows scope for segmentedprotection schemes such as Subnetwork Connection Protection (SNCP) whichsimplify management of path diversity and improve availability, though dualnode SNCP interworking is difficult to arrange.
The remaining attributes of the hierarchical mesh are as described in theprevious section.
AB
F GH
J
K
L M
�Express� Link
Design Issues� Identify Major Nodes and Routes� Build �Express� or �Bypass� links
- Reduces traffic through intermediatecross-connects
� Define rules for interconnect intoExpress links- Where- Single Entry/Exit
Major Node
I
E
�Express� orRegional Network
Design Issues� Express Links form �Core� Network� Remaining Network can be divided
into Regions� Path protection difficult to segment
- Dual Head SNCP
�Core� NetworkA
B
CD
E
FG
H
J K
LM
I
53 545 Network Architecture
55 54 55 54Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
� Simple end-to-end path diversity yields poor availability in largenetworks; typically more than 300 km path length.
� The protection path is always longer than the working path.Protection by path diversity, therefore, requires more than twice theresources that are used by the working path. Typically the protectionand working paths together will need 2.5 times the resources of asimple working path.
Hierarchical or Express VC-4 Mesh Architecture
One of the problems of the flat mesh architecture is that routes pass throughevery intermediate node. A node that only contributes a small amount of trafficmight still have large amounts of equipment just to handle the traffic in transit.
This can be avoided by the use of �Express� routes that link the larger nodes andbypass intermediate nodes as shown by the blue lines in Figure 5-3. The resultis a more efficient mesh architecture and in general a more manageable one.
Figure 5-3 Express Links
The �ad hoc� solution of using express links can be extended to the point wherethe express links and nodes form an new layer in the mesh architecture as shownin Figure 5-4. The mesh has been divided into a �Core Network� layer of highcapacity links between a few key nodes and a lower capacity feeder mesh layerconnecting the remaining nodes. It may then also become natural to partition thefeeder mesh into different regional subnetworks which interconnect to the corenetwork at major nodes.
Figure 5-4 Express Link Core Network
The hierarchical mesh architecture is a more efficient implementation of a meshbut some of the intrinsic routing flexibility is lost. This may not be a such badthing, since routing becomes more structured and easier to manage. It alsoprovides natural boundaries for subnetworks. This allows scope for segmentedprotection schemes such as Subnetwork Connection Protection (SNCP) whichsimplify management of path diversity and improve availability, though dualnode SNCP interworking is difficult to arrange.
The remaining attributes of the hierarchical mesh are as described in theprevious section.
AB
F GH
J
K
L M
�Express� Link
Design Issues� Identify Major Nodes and Routes� Build �Express� or �Bypass� links
- Reduces traffic through intermediatecross-connects
� Define rules for interconnect intoExpress links- Where- Single Entry/Exit
Major Node
I
E
�Express� orRegional Network
Design Issues� Express Links form �Core� Network� Remaining Network can be divided
into Regions� Path protection difficult to segment
- Dual Head SNCP
�Core� NetworkA
B
CD
E
FG
H
J K
LM
I
53 545 Network Architecture
55 54 55 54Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
VC-4 Ring Architectures
The ring architecture is the other common way to implement the VC-4 layer. Anumber of intersecting or overlapping rings are constructed using the availablefibre resources as shown in Figure 5-5. Rings normally offer reduced physicalinterconnectivity but provide implicit diversity.
A connection from node A to node B can be confined to the top ring, while onefrom node A to node F passes across two rings. A connection across a ring isprotected by a complementary connection the other way round the ring. Wheretwo rings meet the connection can either pass through one node or a matchednode connection scheme can be employed whereby two nodes on each ring (forexample, nodes C and E in Figure 5-5) act as main and standby nodes for ringinterconnection.
This arrangement provides the higher availability of SNCP protection schemeswhilst avoiding vulnerability to single node failure.
Figure 5-5 VC-4 Ring Architectures
The choice of what nodes form each ring (that is, the �ring cover� of a network)is in theory a difficult task. In practice it is simplified since rings are usuallyconstructed to link either geographically close nodes or nodes that share a traffic�community of interest�.
The main properties of a ring architecture are as follows:
At nodes where traffic just enters or exits a ring (rather than passing betweenrings) the traffic routing can be managed by the ring ADM itself. At morecomplex nodes where rings interconnect (for example, node E), some form ofadditional routing flexibility may be required. If, however, the number of ringsto interconnect is small (for example, three or four rings) it is possible to usedirect connection and ADMs to achieve this routing (see the �Direct connection�inset in Figure 5-6). A larger number of rings will at some point require a 4/4cross-connect to route traffic between ADMs; a significant additional cost (seethe �Via Cross-connect� inset in Figure 5-6).
ADM
Design Issues� Reduced physical connectivity� Implicit diversity� Rings matched to traffic and/or
topological groups- Implicit subnetworks
� Routing via ADM� Range of Protection schemes
- Path- SNCP- MS-SPRings (effectively
SNCP)� Matched node interworking� Simple intermediate nodes
- Traffic growth requires oneof more rings to grow
AB
CD
F G
J K
LM
I
E
Demand Mapping Demand is mapped on to rings (rather than links), routing round the rings is simple.
Routing Less Flexible, routes pass through every intermediate node on a ring.
Nodes Simple - most nodes are just ADMs
Subnetworking Rings form natural subnetworks
Protection method Variety of methods possible:Path protection - End to End path diversitySNCP - Removes end to end diversity requirementMS-SPRing - effectively a more efficient SNCPMatched Node interworking - SNCP without
interconnecting node vulnerability
Protection overhead MS-SPRing systems allow more efficient protection than path or SNCP schemes, that is, less than 100% overhead.
Demand Growth In general growth in one demand requires entire ring (or rings) to grow.
55 565 Network Architecture
57 56 57 56Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
VC-4 Ring Architectures
The ring architecture is the other common way to implement the VC-4 layer. Anumber of intersecting or overlapping rings are constructed using the availablefibre resources as shown in Figure 5-5. Rings normally offer reduced physicalinterconnectivity but provide implicit diversity.
A connection from node A to node B can be confined to the top ring, while onefrom node A to node F passes across two rings. A connection across a ring isprotected by a complementary connection the other way round the ring. Wheretwo rings meet the connection can either pass through one node or a matchednode connection scheme can be employed whereby two nodes on each ring (forexample, nodes C and E in Figure 5-5) act as main and standby nodes for ringinterconnection.
This arrangement provides the higher availability of SNCP protection schemeswhilst avoiding vulnerability to single node failure.
Figure 5-5 VC-4 Ring Architectures
The choice of what nodes form each ring (that is, the �ring cover� of a network)is in theory a difficult task. In practice it is simplified since rings are usuallyconstructed to link either geographically close nodes or nodes that share a traffic�community of interest�.
The main properties of a ring architecture are as follows:
At nodes where traffic just enters or exits a ring (rather than passing betweenrings) the traffic routing can be managed by the ring ADM itself. At morecomplex nodes where rings interconnect (for example, node E), some form ofadditional routing flexibility may be required. If, however, the number of ringsto interconnect is small (for example, three or four rings) it is possible to usedirect connection and ADMs to achieve this routing (see the �Direct connection�inset in Figure 5-6). A larger number of rings will at some point require a 4/4cross-connect to route traffic between ADMs; a significant additional cost (seethe �Via Cross-connect� inset in Figure 5-6).
ADM
Design Issues� Reduced physical connectivity� Implicit diversity� Rings matched to traffic and/or
topological groups- Implicit subnetworks
� Routing via ADM� Range of Protection schemes
- Path- SNCP- MS-SPRings (effectively
SNCP)� Matched node interworking� Simple intermediate nodes
- Traffic growth requires oneof more rings to grow
AB
CD
F G
J K
LM
I
E
Demand Mapping Demand is mapped on to rings (rather than links), routing round the rings is simple.
Routing Less Flexible, routes pass through every intermediate node on a ring.
Nodes Simple - most nodes are just ADMs
Subnetworking Rings form natural subnetworks
Protection method Variety of methods possible:Path protection - End to End path diversitySNCP - Removes end to end diversity requirementMS-SPRing - effectively a more efficient SNCPMatched Node interworking - SNCP without
interconnecting node vulnerability
Protection overhead MS-SPRing systems allow more efficient protection than path or SNCP schemes, that is, less than 100% overhead.
Demand Growth In general growth in one demand requires entire ring (or rings) to grow.
55 565 Network Architecture
57 56 57 56Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 5-6 Node Interconnection Arrangements
One way to minimise the ring count and the consequent need for cross-connects,is to use a higher capacity ring system. For example, STM-64 rings can carryfour times the capacity of the equivalent STM-16 rings. Thus the use of STM-64systems can remove the need for many cross-connects in the network. The useof a large number of WDM systems to carry equivalent bandwidth, however,will require the use of more cross-connects.
Ring networks have been found to frequently offer cost savings over meshnetworks. This arises from the simplicity and lower cost of nodes and theefficiency of shared protection schemes such as MS-SPRings.
Hierarchical or Express VC 4 Ring Architecture
Just as the efficiency of a mesh network can be improved by the use of �express�links so the same principle can be applied to ring networks by constructing�express� rings. Express rings only enable traffic access at key nodes; anexample is shown in Figure 5-7.
The express ring connects nodes C,E,I,K. Access to other nodes is via the otherrings also connected to the key nodes. The express ring uses fewer ADMs thanthe other rings and focuses traffic on to these ADMs. This allows high capacityADMs to be cost effectively deployed at appropriate locations with smallerlocations served by other low capacity rings. As in the express mesh network theexpress ring can be regarded as forming a higher level in the transmissionhierarchy.
Figure 5-7 An Express Ring
ADM
Design Issues� Interconnect nodes� Ring capacity increase
- Higher rates STM-4/16/64- Stacking Rings
� High Cap (STM-64) ADMsminimise need for Cross-connect
AB
CD
F G
J K
LM
I
E
4/4
Direct connection Via Cross-connect
ADM
Design Issues� Identify Major Nodes and Routes� Define �Express� Ring
- Define interconnect nodes- Eliminates intermediate nodes
(and rings)� Express ring uses fewer and
larger ADMs
AB
CD
F G
J K
LM
I
E
57 585 Network Architecture
59 58 59 58Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 5-6 Node Interconnection Arrangements
One way to minimise the ring count and the consequent need for cross-connects,is to use a higher capacity ring system. For example, STM-64 rings can carryfour times the capacity of the equivalent STM-16 rings. Thus the use of STM-64systems can remove the need for many cross-connects in the network. The useof a large number of WDM systems to carry equivalent bandwidth, however,will require the use of more cross-connects.
Ring networks have been found to frequently offer cost savings over meshnetworks. This arises from the simplicity and lower cost of nodes and theefficiency of shared protection schemes such as MS-SPRings.
Hierarchical or Express VC 4 Ring Architecture
Just as the efficiency of a mesh network can be improved by the use of �express�links so the same principle can be applied to ring networks by constructing�express� rings. Express rings only enable traffic access at key nodes; anexample is shown in Figure 5-7.
The express ring connects nodes C,E,I,K. Access to other nodes is via the otherrings also connected to the key nodes. The express ring uses fewer ADMs thanthe other rings and focuses traffic on to these ADMs. This allows high capacityADMs to be cost effectively deployed at appropriate locations with smallerlocations served by other low capacity rings. As in the express mesh network theexpress ring can be regarded as forming a higher level in the transmissionhierarchy.
Figure 5-7 An Express Ring
ADM
Design Issues� Interconnect nodes� Ring capacity increase
- Higher rates STM-4/16/64- Stacking Rings
� High Cap (STM-64) ADMsminimise need for Cross-connect
AB
CD
F G
J K
LM
I
E
4/4
Direct connection Via Cross-connect
ADM
Design Issues� Identify Major Nodes and Routes� Define �Express� Ring
- Define interconnect nodes- Eliminates intermediate nodes
(and rings)� Express ring uses fewer and
larger ADMs
AB
CD
F G
J K
LM
I
E
57 585 Network Architecture
59 58 59 58Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Network Optimisation and Flexibility
When changing from a flat mesh to a hierarchical mesh and from a flat ring toa hierarchical ring, the transmission design usually becomes more efficient andmore closely linked to the traffic patterns in the network. As in most instancesthere is a compromise between flexibility and optimisation as illustrated byFigure 5-8. At one extreme a highly optimised design would incur minimumcapital cost, but changes in traffic patterns might require site visits andunscheduled introduction of new systems leading to high operational costs. Atthe other extreme a network that is heavily over-provided can absorb changes intraffic patterns with lower operational requirements, but is more costly to build.In practice the lowest effective annual cost is found somewhere between theseextremes, but it is a difficult design task to allow �enough� flexibility forreasonable traffic variation.
Figure 5-8 Optimisation Versus Flexibility
VC-12 Layer Design
Previous sections were concerned with the design of the VC-4 layer network,but in many networks there will be some large point-to-point demands and alsomany that require less than a few tens of 2 Mbit/s trunks. It would be far tooinefficient to route these latter demands in a complete VC-4, so there is anotherstage of the design process concerned with mapping the VC-12 demand on tothe VC-4 layer (see Figure 5-1).
Note: The VC-12 or Lower Order (LO) design process is moresignificant in ETSI SDH networks than in SONET networks. In SONETthe basic Higher Order VC is the VC-3, which at 45 Mbit/s capacity isone third of the size of a VC-4. Many more demands are sufficiently largeto be carried in a partially filled VC-3 than in a VC-4, consequently asmaller proportion of SONET traffic is carried in the LO VC layer.
In many network architectures it is possible to reduce the VC-12 routed trafficto a relatively small proportion of the network traffic. For example, it is commonfor traffic hubbing to two or three centres on a regional MS-SPRing, whichminimises the VC-12 traffic.
The first step is to segregate the traffic demand into large and small point-to-point demands, as shown in Figure 5-9. The large demands are carried by �DirectVC-4� from source to destination. The remaining small demands constitute thetraffic routed in the VC-12 layer.
Cost
Optimised Design Flexible Design
Total Annual Cost
Capital Cost(Amortised)
Operational Cost
• Excess Capacity• Non hierarchical
• Capacity matches traffic• Hierarchy matches traffic
59 605 Network Architecture
61 60 61 60Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Network Optimisation and Flexibility
When changing from a flat mesh to a hierarchical mesh and from a flat ring toa hierarchical ring, the transmission design usually becomes more efficient andmore closely linked to the traffic patterns in the network. As in most instancesthere is a compromise between flexibility and optimisation as illustrated byFigure 5-8. At one extreme a highly optimised design would incur minimumcapital cost, but changes in traffic patterns might require site visits andunscheduled introduction of new systems leading to high operational costs. Atthe other extreme a network that is heavily over-provided can absorb changes intraffic patterns with lower operational requirements, but is more costly to build.In practice the lowest effective annual cost is found somewhere between theseextremes, but it is a difficult design task to allow �enough� flexibility forreasonable traffic variation.
Figure 5-8 Optimisation Versus Flexibility
VC-12 Layer Design
Previous sections were concerned with the design of the VC-4 layer network,but in many networks there will be some large point-to-point demands and alsomany that require less than a few tens of 2 Mbit/s trunks. It would be far tooinefficient to route these latter demands in a complete VC-4, so there is anotherstage of the design process concerned with mapping the VC-12 demand on tothe VC-4 layer (see Figure 5-1).
Note: The VC-12 or Lower Order (LO) design process is moresignificant in ETSI SDH networks than in SONET networks. In SONETthe basic Higher Order VC is the VC-3, which at 45 Mbit/s capacity isone third of the size of a VC-4. Many more demands are sufficiently largeto be carried in a partially filled VC-3 than in a VC-4, consequently asmaller proportion of SONET traffic is carried in the LO VC layer.
In many network architectures it is possible to reduce the VC-12 routed trafficto a relatively small proportion of the network traffic. For example, it is commonfor traffic hubbing to two or three centres on a regional MS-SPRing, whichminimises the VC-12 traffic.
The first step is to segregate the traffic demand into large and small point-to-point demands, as shown in Figure 5-9. The large demands are carried by �DirectVC-4� from source to destination. The remaining small demands constitute thetraffic routed in the VC-12 layer.
Cost
Optimised Design Flexible Design
Total Annual Cost
Capital Cost(Amortised)
Operational Cost
• Excess Capacity• Non hierarchical
• Capacity matches traffic• Hierarchy matches traffic
59 605 Network Architecture
61 60 61 60Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 5-9 Traffic Segregation
The VC-12 layer is illustrated in Figure 5-10. It is implemented as a partial meshof VC-4 links carrying VC-12s that are routed in VC-12 flexibility nodes(shown as the circles in the VC-12 layer). Several points should be noted:
� The VC-12 layer is a logical (not physical) partial mesh comprisinglogical VC-4 links between nodes with VC-12 routing capability. TheVC-4 links will very likely pass through intermediate nodes withoutrouting.
� As the VC-12 layer is a logical layer it can have a completelydifferent topology from the underlying VC-4 layer. In Figure 5-10 theVC-4 layer is implemented as a ring, whereas the VC-12 layer is amesh.
� The VC-12 routing nodes can either be true 4/1 cross-connects, or inmany cases an STM-4 or STM-16 ADM with VC-12 routingcapability can be used with considerable cost savings.
Figure 5-10 Comparison of VC-4 and VC-12 Layers
Traffic Demand
e.g. VC-4 Rings
e.g. VC-12 Mesh
� Threshold < Threshold
4/1
Direct VC-4
LO Managed VC-4
Segregate
Direct VC-4
Small VC-12 Demand
Design VC-12 Layer
LO Managed VC-4
Total VC-4
Design VC-4 Layer
4/1
VC-12 Layer
Design Issues� Identify �Best� VC-12 Routing Nodes� Define VC-4 highways linking Nodes� Route VC-12 traffic, generate VC-4
demand on highways� Design VC-4 layer network� Possible iteration of VC-12 design� Link protection can be done in VC-4 level
AB
C
E
FG
H
J KI
VC-4 Highway
VC-4 Layer (Rings)C
4/1
61 625 Network Architecture
63 62 63 62Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 5-9 Traffic Segregation
The VC-12 layer is illustrated in Figure 5-10. It is implemented as a partial meshof VC-4 links carrying VC-12s that are routed in VC-12 flexibility nodes(shown as the circles in the VC-12 layer). Several points should be noted:
� The VC-12 layer is a logical (not physical) partial mesh comprisinglogical VC-4 links between nodes with VC-12 routing capability. TheVC-4 links will very likely pass through intermediate nodes withoutrouting.
� As the VC-12 layer is a logical layer it can have a completelydifferent topology from the underlying VC-4 layer. In Figure 5-10 theVC-4 layer is implemented as a ring, whereas the VC-12 layer is amesh.
� The VC-12 routing nodes can either be true 4/1 cross-connects, or inmany cases an STM-4 or STM-16 ADM with VC-12 routingcapability can be used with considerable cost savings.
Figure 5-10 Comparison of VC-4 and VC-12 Layers
Traffic Demand
e.g. VC-4 Rings
e.g. VC-12 Mesh
� Threshold < Threshold
4/1
Direct VC-4
LO Managed VC-4
Segregate
Direct VC-4
Small VC-12 Demand
Design VC-12 Layer
LO Managed VC-4
Total VC-4
Design VC-4 Layer
4/1
VC-12 Layer
Design Issues� Identify �Best� VC-12 Routing Nodes� Define VC-4 highways linking Nodes� Route VC-12 traffic, generate VC-4
demand on highways� Design VC-4 layer network� Possible iteration of VC-12 design� Link protection can be done in VC-4 level
AB
C
E
FG
H
J KI
VC-4 Highway
VC-4 Layer (Rings)C
4/1
61 625 Network Architecture
63 62 63 62Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
In a �green field� network the optimisation of a mesh VC-12 layer of anyreasonable size can be a complex task. It involves:
� Choosing which nodes have VC-12 routing capability
� What nodes to associate with routing hubs
� What links to provide between them.
To do this properly requires a powerful network optimisation tool. Furthermorethe resulting �optimal� design may be sensitive to changes in trafficdistributions.
In practice the optimisation problem is often simplified since either one isadding to an established network or the primary routing nodes have beenidentified. An established network will have a known VC-12 layer topology. Thelocation of primary routing nodes may have already been determined by theavailability of suitable buildings or interconnect locations.
This leads to the architecture in Figure 5-11 where two locations have beenpredetermined as sites for 4/1 cross-connects. Individual nodes are assigned toa cross-connect. The creation of VC-12 highways is then a relatively simpletask.
In the architecture shown in Figure 5-11 the VC-12 layer is now a star network.The VC-4 links between the nodes. however, can be protected using VC-4protection schemes, so the availability is still good. Also, the underlying VC-4network is a ring network so there is scope for further optimisation. Consider theconnections from A and B to E. In practice the connection from A to E may berouted round a ring that passes through node B, as shown in Figure 5-12. If thedemands from A to E and B to E are small (for example 10 to 20 VC-12s) thereis scope for combining the A to E demand with the B to E demand at node B andforming a �collector VC-4� that starts at A, collects additional traffic at B andfinishes at E as shown in the lower inset diagram in Figure 5-12. A similartreatment could also be applied to the connections from C to E and D to Eprovided that these also carry a small number of VC-12s.
Figure 5-11 Predetermined Cross-Connect Sites
AB
CD
F G
JK
LM
I
EVC-12 Layer
Design Issues� Identify 4/1 Hubs
Major NodesWell Connected
� Assign Nodes to Hubs� Home traffic and generate VC-4
highway requirements� If link protection in VC-4 level, can limit
number of nodes on path� Issues
SimpleNo guarantee of optimisationReflects real network structureVC-4 from each node, possible low fill
AB
CE
FG H
J
KI
VC-4 Highway
VC-4 Layer (Rings)
D
L M
4/1
4/1
63 645 Network Architecture
65 64 65 64Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
In a �green field� network the optimisation of a mesh VC-12 layer of anyreasonable size can be a complex task. It involves:
� Choosing which nodes have VC-12 routing capability
� What nodes to associate with routing hubs
� What links to provide between them.
To do this properly requires a powerful network optimisation tool. Furthermorethe resulting �optimal� design may be sensitive to changes in trafficdistributions.
In practice the optimisation problem is often simplified since either one isadding to an established network or the primary routing nodes have beenidentified. An established network will have a known VC-12 layer topology. Thelocation of primary routing nodes may have already been determined by theavailability of suitable buildings or interconnect locations.
This leads to the architecture in Figure 5-11 where two locations have beenpredetermined as sites for 4/1 cross-connects. Individual nodes are assigned toa cross-connect. The creation of VC-12 highways is then a relatively simpletask.
In the architecture shown in Figure 5-11 the VC-12 layer is now a star network.The VC-4 links between the nodes. however, can be protected using VC-4protection schemes, so the availability is still good. Also, the underlying VC-4network is a ring network so there is scope for further optimisation. Consider theconnections from A and B to E. In practice the connection from A to E may berouted round a ring that passes through node B, as shown in Figure 5-12. If thedemands from A to E and B to E are small (for example 10 to 20 VC-12s) thereis scope for combining the A to E demand with the B to E demand at node B andforming a �collector VC-4� that starts at A, collects additional traffic at B andfinishes at E as shown in the lower inset diagram in Figure 5-12. A similartreatment could also be applied to the connections from C to E and D to Eprovided that these also carry a small number of VC-12s.
Figure 5-11 Predetermined Cross-Connect Sites
AB
CD
F G
JK
LM
I
EVC-12 Layer
Design Issues� Identify 4/1 Hubs
Major NodesWell Connected
� Assign Nodes to Hubs� Home traffic and generate VC-4
highway requirements� If link protection in VC-4 level, can limit
number of nodes on path� Issues
SimpleNo guarantee of optimisationReflects real network structureVC-4 from each node, possible low fill
AB
CE
FG H
J
KI
VC-4 Highway
VC-4 Layer (Rings)
D
L M
4/1
4/1
63 645 Network Architecture
65 64 65 64Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 5-12 VC-4 Collectors
The use of collector VC-4s:
� Increases the fill of the VC-4s making more efficient use of transport
� Reduces the number of VC-4 ports on the hubbing cross-connect
� Allows full routing flexibility
A connection from node A to B is achieved via the collector alone. A connectionfrom node A to C is implemented via the A and C node collectors and the hubcross-connect at E. The ADMs used to manage the collector can logically bedivided into VC-12 and VC-4 management ADMs. They can be implementedeither as two physical LO and HO boxes or by a combined multiplexer thatperforms VC-12 management and has a sufficiently high bandwidth, that is,STM-4 or STM-16, aggregates.
A B
C D
F
E
VC-12 Layer
Design Issues� Group VC-4 highways into common
collectorsRing membership simplifies task
� May need to protect collectors in VC-12layer
AB
CE
VC-4 Layer (Rings)
D4/1
4/1
4/1
CED
A B
VC-4 Collectors
HO Mux
LO Mux
65 665 Network Architecture
67 66 67 66Network Architecture Network Architecture
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 5-12 VC-4 Collectors
The use of collector VC-4s:
� Increases the fill of the VC-4s making more efficient use of transport
� Reduces the number of VC-4 ports on the hubbing cross-connect
� Allows full routing flexibility
A connection from node A to B is achieved via the collector alone. A connectionfrom node A to C is implemented via the A and C node collectors and the hubcross-connect at E. The ADMs used to manage the collector can logically bedivided into VC-12 and VC-4 management ADMs. They can be implementedeither as two physical LO and HO boxes or by a combined multiplexer thatperforms VC-12 management and has a sufficiently high bandwidth, that is,STM-4 or STM-16, aggregates.
A B
C D
F
E
VC-12 Layer
Design Issues� Group VC-4 highways into common
collectorsRing membership simplifies task
� May need to protect collectors in VC-12layer
AB
CE
VC-4 Layer (Rings)
D4/1
4/1
4/1
CED
A B
VC-4 Collectors
HO Mux
LO Mux
65 665 Network Architecture
67 66 67 66Protection Protection
Synchronous Transmission Systems
Protection
Introduction
With the evolution of modern society we have become increasingly moredependent on the transmission network. Businesses have spread geographicallyand become global. They rely on transmission links for the transport of data andvoice traffic. Not only has the volume of traffic increased, but the end customerhas become more demanding. A highly reliable service from operators isrequired. For example, a loss of traffic for a few seconds can result in a banklosing large sums of money.
What happens if a link fails? For example a fibre break can occur if a fibre ductis severed by a mechanical excavator. Even with this type of failure the end userwill not tolerate a loss of service. Alternatively a power cut may causeequipment to fail or essential maintenance may lead to equipment down time.The very high capacity of SDH links means that a single link failure can have ahuge impact on the services supplied by the network unless it is adequatelyprotected. A resilient network ensuring that traffic can be restored automaticallyin the event of failure is, therefore, of prime importance. SDH transmissionsystems enable deployment of standard automatic protection schemes.
The subject of network protection has many terms and concepts. Moreover forsome terms there is no universally accepted meaning. This chapter defines theseterms and concepts and then discusses common protection schemes and howthese inter-work
Terms
Subnetwork: A single network may be viewed as the interconnection ofmultiple subnetworks. A ring is a simple example of a subnetwork. Thesesubnetworks may be organised into different geographical areas or subnetworksbelonging to different operators.
Survivability: A network may be described as survivable if there is no singlepoint of failure between any two nodes. A simple example is shown inFigure 6-1. The provision of main and alternative traffic paths between any twoend nodes means that the network is survivable in the presence of a single pointof failure.
Figure 6-1 Survivability
Availability: Availability is a measure of the proportion of the time that anetwork is able to provide service to the end customer. It indicates how often orconsistently the network can offer the transport functions which the servicerequires for its successful use by the end customer. Since this is of importanceto the customer, it contributes to the definition of service level agreements.
6
A A
B B
Diverseprotectionpath Line
break
Diverse protectionpath now carriestraffic
A - B is an end to end VC-12 path across network
(i) (ii)
67 686 Protection
69 68 69 68Protection Protection
Synchronous Transmission Systems
Protection
Introduction
With the evolution of modern society we have become increasingly moredependent on the transmission network. Businesses have spread geographicallyand become global. They rely on transmission links for the transport of data andvoice traffic. Not only has the volume of traffic increased, but the end customerhas become more demanding. A highly reliable service from operators isrequired. For example, a loss of traffic for a few seconds can result in a banklosing large sums of money.
What happens if a link fails? For example a fibre break can occur if a fibre ductis severed by a mechanical excavator. Even with this type of failure the end userwill not tolerate a loss of service. Alternatively a power cut may causeequipment to fail or essential maintenance may lead to equipment down time.The very high capacity of SDH links means that a single link failure can have ahuge impact on the services supplied by the network unless it is adequatelyprotected. A resilient network ensuring that traffic can be restored automaticallyin the event of failure is, therefore, of prime importance. SDH transmissionsystems enable deployment of standard automatic protection schemes.
The subject of network protection has many terms and concepts. Moreover forsome terms there is no universally accepted meaning. This chapter defines theseterms and concepts and then discusses common protection schemes and howthese inter-work
Terms
Subnetwork: A single network may be viewed as the interconnection ofmultiple subnetworks. A ring is a simple example of a subnetwork. Thesesubnetworks may be organised into different geographical areas or subnetworksbelonging to different operators.
Survivability: A network may be described as survivable if there is no singlepoint of failure between any two nodes. A simple example is shown inFigure 6-1. The provision of main and alternative traffic paths between any twoend nodes means that the network is survivable in the presence of a single pointof failure.
Figure 6-1 Survivability
Availability: Availability is a measure of the proportion of the time that anetwork is able to provide service to the end customer. It indicates how often orconsistently the network can offer the transport functions which the servicerequires for its successful use by the end customer. Since this is of importanceto the customer, it contributes to the definition of service level agreements.
6
A A
B B
Diverseprotectionpath Line
break
Diverse protectionpath now carriestraffic
A - B is an end to end VC-12 path across network
(i) (ii)
67 686 Protection
69 68 69 68Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
It is typically measured as a percentage of the time a connection is functioning.This takes into account the survivability of the network, failure rate of itscomponents and repair times. This term reflects the average quality of service anend customer can expect to experience on a day to day basis.
Figure 6-2 Availability
The following definitions are some of the ways that availability can beimproved.
Equipment protection: The availability of the equipment can be improved byapplying local protection in the network element itself. For example the powersupply, system clock or tributary units can be duplicated. A failed card isreplaced by a �protection� card. It will be replaced automatically whereAutomatic Equipment Protection is in operation.
Network Resilience: To increase the survivability of the network and so theoverall availability, network links can be protected. Procedures are applied toensure that a failed transport link is replaced by a working link and that there isan alternative path should there be a complete network element (node) failure.There are two types of mechanism utlised to ensure that the service can berecovered in this way:
� Restoration: This is a slow automatic or manual process which usesspare capacity between end nodes to recover traffic after a loss ofservice. On detection of failure, traffic has to be re-routed via an
alternative path. The alternative path is found according to predefinedalgorithms and usually uses digital cross-connects. The wholeprocess can take several minutes.
� Protection: In contrast, protection involves automatic mechanismswithin the network elements, which ensure that failures are detectedand compensated for before a loss of service occurs. Protectionmakes use of pre-assigned capacity between nodes and is preferableto restoration because the reserve capacity is certain to be availableand can be accessed more quickly.
Figure 6-3 Contributing Factors to Overall Availability
These concepts will be discussed in the following sections.
Causes of failure: The physical sources of failure in SDH transmissionnetworks can be classified into the following categories:
� Fibres and cables. The main cause of failure of fibres and cables isdamage by external agents such as civil engineering works andenvironmental effects such as lightning strikes or earthquakes.
� Equipment may fail due to aging effects, random componentstressing or the ingress of moisture. Rigorous tests are, however,normally assumed to have eliminated early life failures.
Survivabilityof networks
Repair timeof components
Failure ofcomponents
Customer
Availability
Improve Overall Availability
Improve Availabilityof Equipment
Equipment
Increase Network Availability
Restoration NetworkProtection Protection
69 706 Protection
71 70 71 70Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
It is typically measured as a percentage of the time a connection is functioning.This takes into account the survivability of the network, failure rate of itscomponents and repair times. This term reflects the average quality of service anend customer can expect to experience on a day to day basis.
Figure 6-2 Availability
The following definitions are some of the ways that availability can beimproved.
Equipment protection: The availability of the equipment can be improved byapplying local protection in the network element itself. For example the powersupply, system clock or tributary units can be duplicated. A failed card isreplaced by a �protection� card. It will be replaced automatically whereAutomatic Equipment Protection is in operation.
Network Resilience: To increase the survivability of the network and so theoverall availability, network links can be protected. Procedures are applied toensure that a failed transport link is replaced by a working link and that there isan alternative path should there be a complete network element (node) failure.There are two types of mechanism utlised to ensure that the service can berecovered in this way:
� Restoration: This is a slow automatic or manual process which usesspare capacity between end nodes to recover traffic after a loss ofservice. On detection of failure, traffic has to be re-routed via an
alternative path. The alternative path is found according to predefinedalgorithms and usually uses digital cross-connects. The wholeprocess can take several minutes.
� Protection: In contrast, protection involves automatic mechanismswithin the network elements, which ensure that failures are detectedand compensated for before a loss of service occurs. Protectionmakes use of pre-assigned capacity between nodes and is preferableto restoration because the reserve capacity is certain to be availableand can be accessed more quickly.
Figure 6-3 Contributing Factors to Overall Availability
These concepts will be discussed in the following sections.
Causes of failure: The physical sources of failure in SDH transmissionnetworks can be classified into the following categories:
� Fibres and cables. The main cause of failure of fibres and cables isdamage by external agents such as civil engineering works andenvironmental effects such as lightning strikes or earthquakes.
� Equipment may fail due to aging effects, random componentstressing or the ingress of moisture. Rigorous tests are, however,normally assumed to have eliminated early life failures.
Survivabilityof networks
Repair timeof components
Failure ofcomponents
Customer
Availability
Improve Overall Availability
Improve Availabilityof Equipment
Equipment
Increase Network Availability
Restoration NetworkProtection Protection
69 706 Protection
71 70 71 70Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
� Mains power failures do occur and are outside the control of thenetwork operator. Major systems are backed-up by secondary powersupplies, but transient effects on signals are possible while switchingto the secondary power supply.
� Maintenance. Unscheduled maintenance and errors made duringmaintenance can affect the availability of service.
� Routing mistakes by the operator can also affect the availability ofservice, however, these can be minimised by the inclusion ofprotection switch triggers such as path trace alarms within thenetwork elements.
� Disasters whether caused by environmental or human action, usuallyhave widespread and severe effects such as the destruction of majornetwork components.
Equipment Protection
Quality targets are set for the elements in an SDH transmission network andthese affect the availability measure of the network. To achieve the availabilityrequirements it is sometimes necessary to duplicate modules within the networkelement.
Every component within a network element has a failure rate associated with it.This is used with information regarding the interaction of components, tocalculate failure rates for the circuit cards. Similarly the circuit card failure ratesand interaction information are used to calculate a failure rate for the networkelement. Taking into account the repair time and software failure information anoverall availability measure for the network element is calculated.
The availability can be improved by provisioning a standby component to takeover on failure. This local protection is commonly applied to several units;power supply, clock generation, switch matrix and tributary cards.
For example a standby tributary card maybe provisioned in the network element.In the event of failure of the working tributary card, traffic is automaticallyswitched to the standby card with no interruption of service to the end user.
Failure of cards is not the only reason for tributary protection. Standby cards arealso used during routine maintenance. Traffic can be manually switched to the
standby card while the working card is maintained. This also enables theworking card to be upgraded while the network element is in service with nointerruption of service to the end user.
There are different standard schemes for equipment protection. For example ifone standby card is included for each working card, these cards are 1+1protected (see Figure 6-4).
Figure 6-4 1+1 and 1:n Protection
It is also common for one protection card to be provisioned for several workingcards. In the event of failure of any of the working cards, traffic is automaticallyswitched to the protection card. This is 1:n protection.
For example in an STM-16 multiplexer, 1:16 protection could be implementedon the STM-1 tributary cards. Sixteen STM-1e (electrical) cards can be installedin the shelf to support sixteen STM-1 tributaries. A 17th card is also included.In the event of failure of any one of the STM-1e cards, traffic may be switchedto the standby protection card.
Working
Standby
Working
Working
Working
Working
Standby
1 + 1 Protection
1 : n Protection
71 726 Protection
73 72 73 72Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
� Mains power failures do occur and are outside the control of thenetwork operator. Major systems are backed-up by secondary powersupplies, but transient effects on signals are possible while switchingto the secondary power supply.
� Maintenance. Unscheduled maintenance and errors made duringmaintenance can affect the availability of service.
� Routing mistakes by the operator can also affect the availability ofservice, however, these can be minimised by the inclusion ofprotection switch triggers such as path trace alarms within thenetwork elements.
� Disasters whether caused by environmental or human action, usuallyhave widespread and severe effects such as the destruction of majornetwork components.
Equipment Protection
Quality targets are set for the elements in an SDH transmission network andthese affect the availability measure of the network. To achieve the availabilityrequirements it is sometimes necessary to duplicate modules within the networkelement.
Every component within a network element has a failure rate associated with it.This is used with information regarding the interaction of components, tocalculate failure rates for the circuit cards. Similarly the circuit card failure ratesand interaction information are used to calculate a failure rate for the networkelement. Taking into account the repair time and software failure information anoverall availability measure for the network element is calculated.
The availability can be improved by provisioning a standby component to takeover on failure. This local protection is commonly applied to several units;power supply, clock generation, switch matrix and tributary cards.
For example a standby tributary card maybe provisioned in the network element.In the event of failure of the working tributary card, traffic is automaticallyswitched to the standby card with no interruption of service to the end user.
Failure of cards is not the only reason for tributary protection. Standby cards arealso used during routine maintenance. Traffic can be manually switched to the
standby card while the working card is maintained. This also enables theworking card to be upgraded while the network element is in service with nointerruption of service to the end user.
There are different standard schemes for equipment protection. For example ifone standby card is included for each working card, these cards are 1+1protected (see Figure 6-4).
Figure 6-4 1+1 and 1:n Protection
It is also common for one protection card to be provisioned for several workingcards. In the event of failure of any of the working cards, traffic is automaticallyswitched to the protection card. This is 1:n protection.
For example in an STM-16 multiplexer, 1:16 protection could be implementedon the STM-1 tributary cards. Sixteen STM-1e (electrical) cards can be installedin the shelf to support sixteen STM-1 tributaries. A 17th card is also included.In the event of failure of any one of the STM-1e cards, traffic may be switchedto the standby protection card.
Working
Standby
Working
Working
Working
Working
Standby
1 + 1 Protection
1 : n Protection
71 726 Protection
73 72 73 72Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Equipment protection increases the availability of the individual networkelements, but does not protect against the loss of an entire network element. Toensure that traffic can be re-routed if a network element is lost, protectionschemes must be implemented which increase the survivability of the network.Network resilience rather than local equipment protection is required to protectagainst the failure of a node or the loss of a link.
Restoration
Restoration is concerned with the availability of an end to end path of service.It works across the entire network and re-routes traffic to maintain service. Apercentage of network capacity is set aside for restoration. After a loss of signalhas been detected, traffic is re-routed via the spare capacity. Re-routingalgorithms are programmed into the network element software. The alternativepath may be found by dropping lower priority traffic or using spare capacitybetween nodes.
In contrast with network protection procedures, the capacity used for restorationneed not be preassigned. In some protection schemes a link is dedicated as aprotection link for the working link. This is not the case in restoration, wherespare capacity can be shared.
Although this strategy offers great flexibility, there will often be a considerablenumber of re-routing options so the algorithms are relatively complex. Thecomputation time required to find an alternative traffic route means that it isdifficult to have rapid restoration of affected traffic. Also restoration is initiatedonly after a loss of signal has been detected by the network management system,not when the failure actually occurs. This leads to restoration times which arerelatively slow, ranging from seconds or minutes to hours. This complexprocedure is outlined below:
In a protected network the elements detect a failure as soon as it occurs and takecorrective action according to predefined procedures, without instruction fromthe network management system. Restoration is a slower process and so thedisruption the end user experiences is greater. In contrast with Restoration, anautomatic protection scheme such as Multiplex Section Protection (MSP) orMS-SPRing, traffic is re-routed in less than 50 ms, so the end users experienceno disruption.
Restoration has not yet been standardised. The various products on offer havebeen developed to meet the internal specifications of a number of operators.
Network Protection
Network protection procedures are employed to self-heal network failuresshould a link or network element fail. What effectively happens is that a networkelement will detect a failure or loss of traffic and initiates corrective actionwithout involving the network management system.
There are many protection mechanisms defined by standards bodies, the nextsection will discuss the commonly deployed schemes. These schemes can besubdivided into those that protect at the section layer and those that protect at thepath layer or subnetwork. The difference will become more obvious as thecommon schemes are discussed, but as a simple introduction:
� Protecting at the section layer involves the switching of all the trafficon a section to an alternative fibre section.
� Protecting at the path or trail layer involves the protection of a virtualcontainer on an end to end path within the subnetwork. In the event offailure, only the virtual container in question is switched to analternative path.
The type of protection scheme employed is usually dictated by the networkarchitecture and this is discussed later in this Chapter.
Dedicated Path/ VC Trail Protection
This type of protection involves duplicating the traffic in the form of virtualcontainers as it enters the network and transmitting this signal simultaneously in
1 Alarms from network detected via management system
2 Alarms analysed to determine their cause
3 Alternative subnetwork connections to restore path connected
4 Path implemented by changing connections
5 Path validated
73 746 Protection
75 74 75 74Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Equipment protection increases the availability of the individual networkelements, but does not protect against the loss of an entire network element. Toensure that traffic can be re-routed if a network element is lost, protectionschemes must be implemented which increase the survivability of the network.Network resilience rather than local equipment protection is required to protectagainst the failure of a node or the loss of a link.
Restoration
Restoration is concerned with the availability of an end to end path of service.It works across the entire network and re-routes traffic to maintain service. Apercentage of network capacity is set aside for restoration. After a loss of signalhas been detected, traffic is re-routed via the spare capacity. Re-routingalgorithms are programmed into the network element software. The alternativepath may be found by dropping lower priority traffic or using spare capacitybetween nodes.
In contrast with network protection procedures, the capacity used for restorationneed not be preassigned. In some protection schemes a link is dedicated as aprotection link for the working link. This is not the case in restoration, wherespare capacity can be shared.
Although this strategy offers great flexibility, there will often be a considerablenumber of re-routing options so the algorithms are relatively complex. Thecomputation time required to find an alternative traffic route means that it isdifficult to have rapid restoration of affected traffic. Also restoration is initiatedonly after a loss of signal has been detected by the network management system,not when the failure actually occurs. This leads to restoration times which arerelatively slow, ranging from seconds or minutes to hours. This complexprocedure is outlined below:
In a protected network the elements detect a failure as soon as it occurs and takecorrective action according to predefined procedures, without instruction fromthe network management system. Restoration is a slower process and so thedisruption the end user experiences is greater. In contrast with Restoration, anautomatic protection scheme such as Multiplex Section Protection (MSP) orMS-SPRing, traffic is re-routed in less than 50 ms, so the end users experienceno disruption.
Restoration has not yet been standardised. The various products on offer havebeen developed to meet the internal specifications of a number of operators.
Network Protection
Network protection procedures are employed to self-heal network failuresshould a link or network element fail. What effectively happens is that a networkelement will detect a failure or loss of traffic and initiates corrective actionwithout involving the network management system.
There are many protection mechanisms defined by standards bodies, the nextsection will discuss the commonly deployed schemes. These schemes can besubdivided into those that protect at the section layer and those that protect at thepath layer or subnetwork. The difference will become more obvious as thecommon schemes are discussed, but as a simple introduction:
� Protecting at the section layer involves the switching of all the trafficon a section to an alternative fibre section.
� Protecting at the path or trail layer involves the protection of a virtualcontainer on an end to end path within the subnetwork. In the event offailure, only the virtual container in question is switched to analternative path.
The type of protection scheme employed is usually dictated by the networkarchitecture and this is discussed later in this Chapter.
Dedicated Path/ VC Trail Protection
This type of protection involves duplicating the traffic in the form of virtualcontainers as it enters the network and transmitting this signal simultaneously in
1 Alarms from network detected via management system
2 Alarms analysed to determine their cause
3 Alternative subnetwork connections to restore path connected
4 Path implemented by changing connections
5 Path validated
73 746 Protection
75 74 75 74Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
two directions across the network. A dedicated protection path carries the trafficin one direction and a working path carries the signal on a different route. At thereceiving network element the quality of the signal from the two paths iscompared and the higher quality signal selected. This is referred to as theworking path. In event of failure on this path the receiving end will switch to theother path, the protection path.
Figure 6-5 Dedicated Path
This will protect not only the links themselves but also protect against the failureof an intermediate node. A special example of this type of mechanism is thepath-protected ring. As the traffic enters the ring it is simultaneously transmittedin both directions around the ring. Selection is made at the exit node for the bestof the two connections.
Figure 6-6 Path-protected ring
The mechanism can be applied to rings and also end to end paths across mesh ormixed networks across many network elements and intermediate subnetworks.
Subnetwork Connection Protection
SNCP is similar to path protection, but whereas dedicated path protectioninvolves switching at the end of an end to end path, in SNCP switching can beinitiated at the end of a path or at an intermediate node, see Figure 6-7. Thenetwork can be decomposed into a number of interconnected subnetworks.Within each subnetwork protection is provided at the path level and automaticprotection switching between the two paths is provided at the subnetworkboundaries.
The selection of the best quality signal is made, not just by the network elementat the end of the path, but also by intermediate nodes at the exit of eachsubnetwork through which the path traverses. The virtual container is notterminated at the intermediate network element it instead compares the qualityof the signal on the two incoming ports and selects the better quality signal. Thedifference between this and path protection is best illustrated by the followingexample.
A
B
C
DE
F
G
H
I
J
Node F receives both workingand protected traffic and
selects the highest quality
A
B
w1
w2
p1
p2
75 766 Protection
77 76 77 76Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
two directions across the network. A dedicated protection path carries the trafficin one direction and a working path carries the signal on a different route. At thereceiving network element the quality of the signal from the two paths iscompared and the higher quality signal selected. This is referred to as theworking path. In event of failure on this path the receiving end will switch to theother path, the protection path.
Figure 6-5 Dedicated Path
This will protect not only the links themselves but also protect against the failureof an intermediate node. A special example of this type of mechanism is thepath-protected ring. As the traffic enters the ring it is simultaneously transmittedin both directions around the ring. Selection is made at the exit node for the bestof the two connections.
Figure 6-6 Path-protected ring
The mechanism can be applied to rings and also end to end paths across mesh ormixed networks across many network elements and intermediate subnetworks.
Subnetwork Connection Protection
SNCP is similar to path protection, but whereas dedicated path protectioninvolves switching at the end of an end to end path, in SNCP switching can beinitiated at the end of a path or at an intermediate node, see Figure 6-7. Thenetwork can be decomposed into a number of interconnected subnetworks.Within each subnetwork protection is provided at the path level and automaticprotection switching between the two paths is provided at the subnetworkboundaries.
The selection of the best quality signal is made, not just by the network elementat the end of the path, but also by intermediate nodes at the exit of eachsubnetwork through which the path traverses. The virtual container is notterminated at the intermediate network element it instead compares the qualityof the signal on the two incoming ports and selects the better quality signal. Thedifference between this and path protection is best illustrated by the followingexample.
A
B
C
DE
F
G
H
I
J
Node F receives both workingand protected traffic and
selects the highest quality
A
B
w1
w2
p1
p2
75 766 Protection
77 76 77 76Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 6-7 Subnetwork Connection Protection
In the event of the two simultaneous failures, protection switching must occur atthe intermediate nodes �A� (see Figure 6-7) for the traffic to reach the far end.SNCP results in a higher availablity for the connection than dedicated pathbecause SNCP allows the network to survive two simultaneous failures whereasdedicated path protection cannot.
In principle path protection end to end appears to have many attractions; end toend network wide protection is possible and individual paths can be selectivelyprotected. Complex control, however, is required to ensure truly diverse routes.A large amount of capacity is used and it is very difficult to co-ordinatescheduled maintenance activities across the whole network. Path protection,therefore, becomes more acceptable when limited to the subnetwork level, thatis, SNCP. SNCP works particularly well over rings, because diversity of fibreroutes is ensured.
Resilience can be offered at a number of SDH levels including end to end path(trail), subnetwork level and multiplex section level. The mechanisms describedabove offer protection at the end to end path and subnetwork level. Theseinvolve the protection of individual virtual containers across an end to end path.In the event of failure only the virtual container in question is switched to analternative path, so selective protection of individual VC�s is possible. Forexample, a business customer may require protection of a leased line, the path ofthis can be protected across the whole network without protecting the othertraffic on the network. In contrast the next mechanism is a section layerprotection mechanism and involves the switching of all the virtual containertraffic on a section onto an alternative fibre section.
It should also be noted that both end to end path and subnetwork path protectioncan be applied to both higher order and lower order paths as described inChapter 5.
The previous examples describe 1+1 single-ended configurations. It should bementioned that 1:1 dual-ended configurations are possible, allowing low prioritytraffic to use the protection path. These are mentioned in ITU-Trecommendation G.841 but at the time of writing have not been defined.
Traffic switched toprotection path
WorkingProtection
pathpath
Tributaries
Tributaries
A A
Failure
Traffic switchedat end node
A
Failure
Failure
Traffic lost on bothworking and protected paths
(i) Dedicated Path
1 2 3
Traffic switched toprotection path
WorkingProtection
pathpath
Tributaries
Tributaries
A A
Failure
Traffic switchedat end node
A
Failure
Failure
Traffic switchedat intermediate node
4 5 6
(ii) Subnetwork Connection Protection
77 786 Protection
79 78 79 78Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 6-7 Subnetwork Connection Protection
In the event of the two simultaneous failures, protection switching must occur atthe intermediate nodes �A� (see Figure 6-7) for the traffic to reach the far end.SNCP results in a higher availablity for the connection than dedicated pathbecause SNCP allows the network to survive two simultaneous failures whereasdedicated path protection cannot.
In principle path protection end to end appears to have many attractions; end toend network wide protection is possible and individual paths can be selectivelyprotected. Complex control, however, is required to ensure truly diverse routes.A large amount of capacity is used and it is very difficult to co-ordinatescheduled maintenance activities across the whole network. Path protection,therefore, becomes more acceptable when limited to the subnetwork level, thatis, SNCP. SNCP works particularly well over rings, because diversity of fibreroutes is ensured.
Resilience can be offered at a number of SDH levels including end to end path(trail), subnetwork level and multiplex section level. The mechanisms describedabove offer protection at the end to end path and subnetwork level. Theseinvolve the protection of individual virtual containers across an end to end path.In the event of failure only the virtual container in question is switched to analternative path, so selective protection of individual VC�s is possible. Forexample, a business customer may require protection of a leased line, the path ofthis can be protected across the whole network without protecting the othertraffic on the network. In contrast the next mechanism is a section layerprotection mechanism and involves the switching of all the virtual containertraffic on a section onto an alternative fibre section.
It should also be noted that both end to end path and subnetwork path protectioncan be applied to both higher order and lower order paths as described inChapter 5.
The previous examples describe 1+1 single-ended configurations. It should bementioned that 1:1 dual-ended configurations are possible, allowing low prioritytraffic to use the protection path. These are mentioned in ITU-Trecommendation G.841 but at the time of writing have not been defined.
Traffic switched toprotection path
WorkingProtection
pathpath
Tributaries
Tributaries
A A
Failure
Traffic switchedat end node
A
Failure
Failure
Traffic lost on bothworking and protected paths
(i) Dedicated Path
1 2 3
Traffic switched toprotection path
WorkingProtection
pathpath
Tributaries
Tributaries
A A
Failure
Traffic switchedat end node
A
Failure
Failure
Traffic switchedat intermediate node
4 5 6
(ii) Subnetwork Connection Protection
77 786 Protection
79 78 79 78Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Multiplex Section Linear Protection
This procedure operates on a traffic section that is between two adjacent nodes.Between these two nodes there are two separate links on two different fibres;working and protection. In the event of failure of the link, the entire signal willbe switched from the working to the protection fibre.
Figure 6-8 Multiplex Section Protection Schemes
There are different types of Multiplex Section Protection (MSP):
1:1 protection is dual ended. Traffic is initially sent via the working link only.A failure is detected at the far end when it no longer receives traffic. A signal issent to the transmitting end this triggers protection switching to the standby pathat both ends. This means that low priority traffic can be carried on the protectionchannel, while traffic is on the working channel, this traffic will be lost ifprotection switching is initiated.
1:n is similar to the above except several working channels can be protected byone protection channel
1+1 MSP: traffic is initially transmitted on both working and protection paths,if loss of traffic is detected at the receiving end it will switch to the protectionpath. There is no need for backward signalling, however, the standby sectioncannot be utlised for other traffic and so fibre capacity requirements are high.
MSP protects traffic between adjacent network elements, but only the linkbetween these two nodes is protected, there is no protection against a completenode failure. Another limitation is that physically diverse paths for working andprotection fibres are required. If the working and protection fibres were in thesame physical duct and this was damaged, both working and protection pathswould be lost.
Figure 6-9 Diverse Path Protection
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
Switch when failure
(i) 1 : 1 Multiplex Section Protection
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
Switch
(ii) 1 + 1 Multiplex Section Protection
whenfailure
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
(i) Working and standby fibres in same cable
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
(ii) Working and standby fibres in diverse paths
Cable break affects both.
Cable break only affects working fibres.Traffic switched to standby.
Diverse path
79 806 Protection
81 80 81 80Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Multiplex Section Linear Protection
This procedure operates on a traffic section that is between two adjacent nodes.Between these two nodes there are two separate links on two different fibres;working and protection. In the event of failure of the link, the entire signal willbe switched from the working to the protection fibre.
Figure 6-8 Multiplex Section Protection Schemes
There are different types of Multiplex Section Protection (MSP):
1:1 protection is dual ended. Traffic is initially sent via the working link only.A failure is detected at the far end when it no longer receives traffic. A signal issent to the transmitting end this triggers protection switching to the standby pathat both ends. This means that low priority traffic can be carried on the protectionchannel, while traffic is on the working channel, this traffic will be lost ifprotection switching is initiated.
1:n is similar to the above except several working channels can be protected byone protection channel
1+1 MSP: traffic is initially transmitted on both working and protection paths,if loss of traffic is detected at the receiving end it will switch to the protectionpath. There is no need for backward signalling, however, the standby sectioncannot be utlised for other traffic and so fibre capacity requirements are high.
MSP protects traffic between adjacent network elements, but only the linkbetween these two nodes is protected, there is no protection against a completenode failure. Another limitation is that physically diverse paths for working andprotection fibres are required. If the working and protection fibres were in thesame physical duct and this was damaged, both working and protection pathswould be lost.
Figure 6-9 Diverse Path Protection
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
Switch when failure
(i) 1 : 1 Multiplex Section Protection
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
Switch
(ii) 1 + 1 Multiplex Section Protection
whenfailure
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
(i) Working and standby fibres in same cable
SDHMultiplexer
Working fibres
SDHMultiplexer
Standby fibres
(ii) Working and standby fibres in diverse paths
Cable break affects both.
Cable break only affects working fibres.Traffic switched to standby.
Diverse path
79 806 Protection
81 80 81 80Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Two alternative routes must, therefore, be laid between pairs of adjacent nodes.This needs consideration when deploying this type of protection scheme.
Multiplex section linear protection is typically used for linear meshed networks.Physically diverse paths are, however, required and the mesh becomesincreasingly more complex as it grows. With the scarcity of fibre becoming acritical issue many operators are beginning to favour the deployment of rings.Rings ensure the lowest possible connectivity between a set of nodes whilstensuring that between each pair of nodes there is a physically diverse path whichcan be used for protection.
Figure 6-10 Ring Topology Enabling Diverse Protection Paths
Self Healing Rings
Self healing Ring protection procedures are becoming increasingly common,because they provide a diverse route for protection and so efficient use of fibre.There are several types of ring protection schemes. These can be split into thosethat protect at the section layer and those at the path layer. These can then befurther subdivided into Uni-directional and Bi-directional schemes. Two types
of self healing ring mechanisms will be considered here, as these are the twocommonly deployed in the ETSI market:
� Path protected bi-directional rings (dedicated protection rings orpath-protected rings)
� Bi-directional shared protection rings (SPRings).
Figure 6-11 Self-healing Rings
Dedicated Protection Rings
This is a type of dedicated path protection, applied to a ring. As traffic enters thering at node A (see Figure 6-12) it is simultaneously sent in both directionsaround the ring. One direction can be considered a working path �w� and theother direction protection path, �p�. The receiving node will select the bestquality connection. For example assume that the best quality signal is �w�; in theevent of a fibre break between A and B on �w� , B will select traffic from path �p�.
A B
Working traffic A - B
Standby traffic A - B
Self-healing Rings
Section Layer Protected Path Layer Protected
Uni MS-DPRing Bi MS-SPRing
2F 4FETSI Markets
Uni Path/SNC Bi Path/SNCProtection Protection
HOP LOP(High order
Path)(Low order
path)Only common in ANSI markets
HOP LOPETSI Markets
81 826 Protection
83 82 83 82Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Two alternative routes must, therefore, be laid between pairs of adjacent nodes.This needs consideration when deploying this type of protection scheme.
Multiplex section linear protection is typically used for linear meshed networks.Physically diverse paths are, however, required and the mesh becomesincreasingly more complex as it grows. With the scarcity of fibre becoming acritical issue many operators are beginning to favour the deployment of rings.Rings ensure the lowest possible connectivity between a set of nodes whilstensuring that between each pair of nodes there is a physically diverse path whichcan be used for protection.
Figure 6-10 Ring Topology Enabling Diverse Protection Paths
Self Healing Rings
Self healing Ring protection procedures are becoming increasingly common,because they provide a diverse route for protection and so efficient use of fibre.There are several types of ring protection schemes. These can be split into thosethat protect at the section layer and those at the path layer. These can then befurther subdivided into Uni-directional and Bi-directional schemes. Two types
of self healing ring mechanisms will be considered here, as these are the twocommonly deployed in the ETSI market:
� Path protected bi-directional rings (dedicated protection rings orpath-protected rings)
� Bi-directional shared protection rings (SPRings).
Figure 6-11 Self-healing Rings
Dedicated Protection Rings
This is a type of dedicated path protection, applied to a ring. As traffic enters thering at node A (see Figure 6-12) it is simultaneously sent in both directionsaround the ring. One direction can be considered a working path �w� and theother direction protection path, �p�. The receiving node will select the bestquality connection. For example assume that the best quality signal is �w�; in theevent of a fibre break between A and B on �w� , B will select traffic from path �p�.
A B
Working traffic A - B
Standby traffic A - B
Self-healing Rings
Section Layer Protected Path Layer Protected
Uni MS-DPRing Bi MS-SPRing
2F 4FETSI Markets
Uni Path/SNC Bi Path/SNCProtection Protection
HOP LOP(High order
Path)(Low order
path)Only common in ANSI markets
HOP LOPETSI Markets
81 826 Protection
83 82 83 82Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 6-12 D-PRing
Multiplex Section Shared Protection Rings (MS-SPRing)
Multiplex Section Shared Protection Rings, commonly called �MS-SPRings�are a ring protection mechanism. In contrast to the dedicated protection ring,traffic is sent on only one route around the ring. A protection path is notdedicated for the protection of each working path, instead capacity on the ringis reserved for protection and this can be shared for the protection of severalworking paths. Protection switching is initiated at the section level in a similarway to multiplex section linear protection; in the event of failure all traffic on asection is switched. This mechanism can achieve significant capacity savingsover a dedicated ring protection mechanism, allowing the operator to increasethe number of working paths on a ring.
So what is an MS-SPRing? In normal mode from A (see Figure 6-13 (i)) avirtual container with traffic destined for B is sent from A to B on the sectionbetween A and B only. In the event of a failure between A and B (seeFigure 6-13 (ii)), all the traffic on this section will be re-routed onto the capacityreserved for protection. This is in a similar way to MSP, all the traffic on asection is switched to a protection section the instant there is a fibre break. Figure 6-13 MS-SPRing
A
B
wp
B selects traffic from w
A
B
wp
B selects traffic from p
(i) Normal operation (ii) Fibre break
A
B
Workingtraffic
Standbytraffic
16 VC-4s:8 Working8 Standby
C
D
EF
A
B
Workingtraffic
Standbytraffic
C
D
EF
(i) Normal operation
(ii) Fibre break
83 846 Protection
85 84 85 84Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 6-12 D-PRing
Multiplex Section Shared Protection Rings (MS-SPRing)
Multiplex Section Shared Protection Rings, commonly called �MS-SPRings�are a ring protection mechanism. In contrast to the dedicated protection ring,traffic is sent on only one route around the ring. A protection path is notdedicated for the protection of each working path, instead capacity on the ringis reserved for protection and this can be shared for the protection of severalworking paths. Protection switching is initiated at the section level in a similarway to multiplex section linear protection; in the event of failure all traffic on asection is switched. This mechanism can achieve significant capacity savingsover a dedicated ring protection mechanism, allowing the operator to increasethe number of working paths on a ring.
So what is an MS-SPRing? In normal mode from A (see Figure 6-13 (i)) avirtual container with traffic destined for B is sent from A to B on the sectionbetween A and B only. In the event of a failure between A and B (seeFigure 6-13 (ii)), all the traffic on this section will be re-routed onto the capacityreserved for protection. This is in a similar way to MSP, all the traffic on asection is switched to a protection section the instant there is a fibre break. Figure 6-13 MS-SPRing
A
B
wp
B selects traffic from w
A
B
wp
B selects traffic from p
(i) Normal operation (ii) Fibre break
A
B
Workingtraffic
Standbytraffic
16 VC-4s:8 Working8 Standby
C
D
EF
A
B
Workingtraffic
Standbytraffic
C
D
EF
(i) Normal operation
(ii) Fibre break
83 846 Protection
85 84 85 84Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Each section on the ring has reserved capacity for protection and the virtualcontainers in question are re-routed the other way around the ring to B on thisreserved �shared� capacity. For example assume that the fibre is carryingSTM-16 traffic, so on each section there are the equivalent of 16 STM-1channels or 16 virtual containers of traffic; VC-4s. On each section eight STM-1channels will carry working traffic and eight will be reserved for protection.
The capacity advantage that can be achieved with MS-SPRing over a dedicatedpath protection ring is not obvious until we extend this simple example to manypaths transiting the ring.
Take an example of a ring with six nodes with a capacity of STM-16, that is 16STM-1 equivalents. Consider a uniform traffic pattern whereby traffic enteringthe ring exits at the adjacent node. In the dedicated protection ring exampleshown in Figure 6-14 (i), between A and B eight STM-1 equivalents arerequired to route the traffic on route w1 and eight STM-1 equivalents arerequired to route this traffic the other way around the ring on route p1. Thisleaves only eight STM-1 equivalents on each section of the ring. If traffic isrouted in a similar manner between D and E on w2 and p2, then all the capacityon the STM-16 ring is used. So a maximum of 16 working paths can be set up(8 on A-B and 8 on D-E).
Now if we take the same ring, but use an MS-SPRing, Figure 6-14 (ii), thetraffic between A and B uses eight STM-1 equivalents on w1, but the eightSTM-1 equivalents required for protection in the other direction around the ring,on p2 can be shared. If all traffic exits and enters at adjacent nodes it is possibleto have working paths between all adjacent nodes, that is, eight STM-1equivalents are used for working traffic all around the ring and on every sectioneight STM-1 equivalents are still available for the �shared� protection of theseworking paths.
Figure 6-14 Shared Protection Ring Capacity Advantage
A
B
Workingtra ffic
Standbytraffic
16 VC-4s:8 Working8 Standby
C
D
EF
8 protected working paths ineach section. All traffic dropped
48 protected paths.
and added at each multiplexer.6 sections giving up to
A
B C
D
EF
8 workingpaths
8 workingpaths
Standby pathsfor A-B traffic
Standby pathsfor D-E traffic
8 protected working paths insections A-B and D-E giving
16 protected paths.
w1
w2
w3
w4
w5
w6
p
p
p
p
p
p
w1
w2
p1
p1
p2
p2
p1
p1
p1
p2
p2
p2
(ii) MS-SPRing
(i) Dedicated protection ring
85 866 Protection
87 86 87 86Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Each section on the ring has reserved capacity for protection and the virtualcontainers in question are re-routed the other way around the ring to B on thisreserved �shared� capacity. For example assume that the fibre is carryingSTM-16 traffic, so on each section there are the equivalent of 16 STM-1channels or 16 virtual containers of traffic; VC-4s. On each section eight STM-1channels will carry working traffic and eight will be reserved for protection.
The capacity advantage that can be achieved with MS-SPRing over a dedicatedpath protection ring is not obvious until we extend this simple example to manypaths transiting the ring.
Take an example of a ring with six nodes with a capacity of STM-16, that is 16STM-1 equivalents. Consider a uniform traffic pattern whereby traffic enteringthe ring exits at the adjacent node. In the dedicated protection ring exampleshown in Figure 6-14 (i), between A and B eight STM-1 equivalents arerequired to route the traffic on route w1 and eight STM-1 equivalents arerequired to route this traffic the other way around the ring on route p1. Thisleaves only eight STM-1 equivalents on each section of the ring. If traffic isrouted in a similar manner between D and E on w2 and p2, then all the capacityon the STM-16 ring is used. So a maximum of 16 working paths can be set up(8 on A-B and 8 on D-E).
Now if we take the same ring, but use an MS-SPRing, Figure 6-14 (ii), thetraffic between A and B uses eight STM-1 equivalents on w1, but the eightSTM-1 equivalents required for protection in the other direction around the ring,on p2 can be shared. If all traffic exits and enters at adjacent nodes it is possibleto have working paths between all adjacent nodes, that is, eight STM-1equivalents are used for working traffic all around the ring and on every sectioneight STM-1 equivalents are still available for the �shared� protection of theseworking paths.
Figure 6-14 Shared Protection Ring Capacity Advantage
A
B
Workingtra ffic
Standbytraffic
16 VC-4s:8 Working8 Standby
C
D
EF
8 protected working paths ineach section. All traffic dropped
48 protected paths.
and added at each multiplexer.6 sections giving up to
A
B C
D
EF
8 workingpaths
8 workingpaths
Standby pathsfor A-B traffic
Standby pathsfor D-E traffic
8 protected working paths insections A-B and D-E giving
16 protected paths.
w1
w2
w3
w4
w5
w6
p
p
p
p
p
p
w1
w2
p1
p1
p2
p2
p1
p1
p1
p2
p2
p2
(ii) MS-SPRing
(i) Dedicated protection ring
85 866 Protection
87 86 87 86Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
As it is possible to have working paths on every section (w1-w6) and there areeight STM-1 channels on each section, a total of 48 (8 x 6) paths could be set up,compared to the 16 for the dedicated protection ring.
This traffic pattern is not typical, but if calculations are carried out for a uniformtraffic pattern, which is typical for traffic on a trunk/ backbone network betweenmajor cities or a metropolitan data network, then SPRings can double thecapacity over a dedicated protection ring.
SPRings can also increase the capacity on fibres by reusing the channelsreserved for protection. In many networks there is a demand for low cost highbandwidth services where cost, not availability, is the priority, for example, IPtraffic. In a SPRing the protection bandwidth is allocated dynamically at theinstant of a fibre break. This means that no unnecessarily large amount ofbandwidth is permanently used for protection and is hence available for suchtraffic in addition to the full protected payload. This provides an easy way tointegrate SPRings with end to end path protection schemes where the protectionfor the path protected traffic is carried in the extra traffic channels effectively�sharing� protection bandwidth between the SPRing and the path protectednetwork.
The examples have described a 2-fibre SPRing where a single fibre is splitbetween working and protection with eight VC-4s carrying working traffic andeight dedicated for shared protection. A 4-fibre version of the SPRing ispossible. In this arrangement, there is no splitting of a single fibre’s capacitybetween working and protection channels; separate fibres are used for each. Thebasic availability is the same for 2 and 4-fibre rings, although the 4-fibre ring canhandle two concurrent span failures. 4-fibre rings can, therefore, providesignificant improvements in availability, especially if the fibre on the spans isdiversely routed. Even when this is not the case, the span switching still providesadditional protection against the failure of the ring optical aggregates. 4-fibrerings do of course require more fibre, however, they can carry twice as muchtraffic as 2-fibre rings.
As well as protecting against the failure of a link, SPRings protect against thefailure of any node in the ring, this not the case with linear MSP.
Comparison of Protection schemes
Table 6-1 Comparison of Schemes to Achieve Network Resilience
Typ
ical
swit
chin
gti
me
<50
ms
<50
ms
<50
ms
<50
ms
>1 m
in
Is p
rote
ctio
nb
and
wid
thre
-use
d?
Yes
No
No
No
Yes
Typ
ical
top
olo
gy
Rin
g
Line
ar/
Mes
h
Mix
ed
Mix
ed
Mes
h
Is s
chem
est
and
ard
ised
?
Yes
Yes
Yes
Yes No
Is s
chem
eV
Cse
lect
ive?
No
No
Yes
Yes
Yes
Wh
ere
can
pro
tect
ion
occ
ur
At a
ny n
ode
in th
e rin
g
Adj
acen
tno
de
Nod
e at
end
of p
ath
End
or
inte
rmed
iate
node
in p
ath
No
prot
ectio
nsw
itchi
ng
Pro
tect
wh
at?
All
traf
ficon
sec
tion
All
traf
ficon
sec
tion
Indi
vidu
alV
C
Indi
vidu
alV
C
Indi
vidu
alV
C
Net
wo
rk
MS
SP
Rin
g
1+1
MS
P
VC
-Tra
il/D
edic
ated
path
SN
CP
Res
tora
tion
87 886 Protection
89 88 89 88Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
As it is possible to have working paths on every section (w1-w6) and there areeight STM-1 channels on each section, a total of 48 (8 x 6) paths could be set up,compared to the 16 for the dedicated protection ring.
This traffic pattern is not typical, but if calculations are carried out for a uniformtraffic pattern, which is typical for traffic on a trunk/ backbone network betweenmajor cities or a metropolitan data network, then SPRings can double thecapacity over a dedicated protection ring.
SPRings can also increase the capacity on fibres by reusing the channelsreserved for protection. In many networks there is a demand for low cost highbandwidth services where cost, not availability, is the priority, for example, IPtraffic. In a SPRing the protection bandwidth is allocated dynamically at theinstant of a fibre break. This means that no unnecessarily large amount ofbandwidth is permanently used for protection and is hence available for suchtraffic in addition to the full protected payload. This provides an easy way tointegrate SPRings with end to end path protection schemes where the protectionfor the path protected traffic is carried in the extra traffic channels effectively�sharing� protection bandwidth between the SPRing and the path protectednetwork.
The examples have described a 2-fibre SPRing where a single fibre is splitbetween working and protection with eight VC-4s carrying working traffic andeight dedicated for shared protection. A 4-fibre version of the SPRing ispossible. In this arrangement, there is no splitting of a single fibre’s capacitybetween working and protection channels; separate fibres are used for each. Thebasic availability is the same for 2 and 4-fibre rings, although the 4-fibre ring canhandle two concurrent span failures. 4-fibre rings can, therefore, providesignificant improvements in availability, especially if the fibre on the spans isdiversely routed. Even when this is not the case, the span switching still providesadditional protection against the failure of the ring optical aggregates. 4-fibrerings do of course require more fibre, however, they can carry twice as muchtraffic as 2-fibre rings.
As well as protecting against the failure of a link, SPRings protect against thefailure of any node in the ring, this not the case with linear MSP.
Comparison of Protection schemes
Table 6-1 Comparison of Schemes to Achieve Network Resilience
Typ
ical
swit
chin
gti
me
<50
ms
<50
ms
<50
ms
<50
ms
>1 m
in
Is p
rote
ctio
nb
and
wid
thre
-use
d?
Yes
No
No
No
Yes
Typ
ical
top
olo
gy
Rin
g
Line
ar/
Mes
h
Mix
ed
Mix
ed
Mes
h
Is s
chem
est
and
ard
ised
?
Yes
Yes
Yes
Yes No
Is s
chem
eV
Cse
lect
ive?
No
No
Yes
Yes
Yes
Wh
ere
can
pro
tect
ion
occ
ur
At a
ny n
ode
in th
e rin
g
Adj
acen
tno
de
Nod
e at
end
of p
ath
End
or
inte
rmed
iate
node
in p
ath
No
prot
ectio
nsw
itchi
ng
Pro
tect
wh
at?
All
traf
ficon
sec
tion
All
traf
ficon
sec
tion
Indi
vidu
alV
C
Indi
vidu
alV
C
Indi
vidu
alV
C
Net
wo
rk
MS
SP
Rin
g
1+1
MS
P
VC
-Tra
il/D
edic
ated
path
SN
CP
Res
tora
tion
87 886 Protection
89 88 89 88Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
As is evident in the Table 6-1, the common protection schemes vary significantlyin their characteristics. For example:
� The time taken to switch traffic to a protection route and subsequentimpact of service to the end customer.
� The level at which protection resources appear, that is, protecting atthe path or section level can have a considerable effect on suchmatters as robustness, costs and restore times. For example sectionlevel protection involves the switching of all traffic across the sectionwhereas with path or SNCP it is possible to selectively protect asingle traffic path.
� If the bandwidth reserved for protection can be re-used to transportother traffic when not utilised for protection. For example with theMS-SPRing the �shared� protection capacity can be utilised for pathprotected traffic.
There is no one best protection scheme. Choice may be determined by the designof the network, for example SPRings tend to be used in a ring topology whereasrestoration tends to be used in a very highly meshed architecture with manycross-connects.
The choice of protection scheme can also be determined by the network tier atwhich the traffic is carried. At the backbone layers the rate of transmission ishigher, for example, STM-16 or STM-64, so a much larger amount of traffic iscarried on each fibre than a lower layer link. A break in this fibre would have amore severe impact than the loss of a lower layer fibre. Backbone traffic can,therefore, justify a full protection scheme such as MS-SPRing or 1+1 MSP.
Traffic patterns vary depending on the layer of the network (see Figure 6-15). Atthe backbone or trunk layer, traffic is typically uniform, being carried betweenmajor cities. The same is true for a metropolitan or data network. In this situationa SPRing can provide a capacity advantage over path protection. The re-use ofcapacity reserved for protection bandwidth is also an important consideration,such as �ring extra traffic�. At the backbone layer, fibre may be scarce andmaking the best use of available bandwidth is critical.
Figure 6-15 Network Hierarchy Traffic Patterns
City
AC
ity B
City
CC
ity D
Tru
nk
Reg
ion
alM
etro
Lo
cal
Uni
form
traf
fic
Tra
ffic
path
Hub
bing
MS
-SP
Rin
g
SN
CP
Pat
hP
rote
ctio
n
Mix
ed
89 906 Protection
91 90 91 90Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
As is evident in the Table 6-1, the common protection schemes vary significantlyin their characteristics. For example:
� The time taken to switch traffic to a protection route and subsequentimpact of service to the end customer.
� The level at which protection resources appear, that is, protecting atthe path or section level can have a considerable effect on suchmatters as robustness, costs and restore times. For example sectionlevel protection involves the switching of all traffic across the sectionwhereas with path or SNCP it is possible to selectively protect asingle traffic path.
� If the bandwidth reserved for protection can be re-used to transportother traffic when not utilised for protection. For example with theMS-SPRing the �shared� protection capacity can be utilised for pathprotected traffic.
There is no one best protection scheme. Choice may be determined by the designof the network, for example SPRings tend to be used in a ring topology whereasrestoration tends to be used in a very highly meshed architecture with manycross-connects.
The choice of protection scheme can also be determined by the network tier atwhich the traffic is carried. At the backbone layers the rate of transmission ishigher, for example, STM-16 or STM-64, so a much larger amount of traffic iscarried on each fibre than a lower layer link. A break in this fibre would have amore severe impact than the loss of a lower layer fibre. Backbone traffic can,therefore, justify a full protection scheme such as MS-SPRing or 1+1 MSP.
Traffic patterns vary depending on the layer of the network (see Figure 6-15). Atthe backbone or trunk layer, traffic is typically uniform, being carried betweenmajor cities. The same is true for a metropolitan or data network. In this situationa SPRing can provide a capacity advantage over path protection. The re-use ofcapacity reserved for protection bandwidth is also an important consideration,such as �ring extra traffic�. At the backbone layer, fibre may be scarce andmaking the best use of available bandwidth is critical.
Figure 6-15 Network Hierarchy Traffic Patterns
City
AC
ity B
City
CC
ity D
Tru
nk
Reg
ion
alM
etro
Lo
cal
Uni
form
traf
fic
Tra
ffic
path
Hub
bing
MS
-SP
Rin
g
SN
CP
Pat
hP
rote
ctio
n
Mix
ed
89 906 Protection
91 90 91 90Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
At the lower regional layer of the network, traffic is typically carried to onecentral point so it can be collected and transported to the next tier or taken to acentral exchange. This is described as �hubbing� traffic. In this situation theadvantages of SPRings are not so great and the need to protect every fibre is notas critical. Selective path protection schemes such as VC-Trail and SNCPprotection are more common in this situation. For example an end customer maydemand protection of their 2 Mbit/s leased line, this VC-12 path can beselectively protected by path protection.
Figure 6-16 Detection of Loss of VC-12
This path is protected at the VC-12 level across the entire network. If this pathwas only protected at a HO trunk layer, that is at the VC-4 level, by MSP or
MS-SPRing and there was a break in a lower layer fibre, this VC-12 path wouldbe lost. A full VC-4 circuit, however, would not be lost, so the protectionmechanism at the VC-4 layer would not detect the failure. An operator musttherefore, not only consider which protection scheme to employ but how thesewill interwork.
An effective deployment of subnetworks is interconnecting SNCP protectedsubnetworks and MS-SPRing protected subnetworks. For example anMS-SPRing subnetwork is ideal for the core or backbone network. This could beconnected to a regional or local network where subnetwork path protection wasused to apply selective protection to traffic.
Difficulties can arise at the network boundaries of the subnetworks. If singlepoints are used to interconnect the subnetworks, there is a vulnerable link in thechain so the interconnection between subnetworks becomes an issue. This isdiscussed in the following section.
At the upper regional layer of the network traffic hubbing can be used at anumber of points. Traffic is transported to multiple regional centres containing,for example, voice switches and routers, which are then connected via thebackbone. In these cases the traffic patterns can give capacity savings which arebetter than those for uniform traffic.
Interworking of Protection Schemes
As the size and demand placed on networks increases so their complexityincreases. Single ring or chain networks will rarely be implemented. Networkswill be made up of a number of sub-networks and each may have its ownindependent protection scheme. With the increasing number of operators, theinterconnection of networks belonging to different operators becomes an issue.These factors along with the drive for more reliable networks, means that theissue of interconnecting various individually protected subnetworks is ofincreasing importance.
Protection interworking is where more than one protection scheme operates ona single connection across a network. A single protection scheme may notprovide the optimum performance, it can be better to implement protection on asubnetwork basis, but then the interworking of these schemes must beconsidered
Trunk
RegionalMetro
LocalLocal
VC-12
VC-4
Fibrebreak
VC-4 protectiondoes not detectloss of VC-12 traffic
91 926 Protection
93 92 93 92Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
At the lower regional layer of the network, traffic is typically carried to onecentral point so it can be collected and transported to the next tier or taken to acentral exchange. This is described as �hubbing� traffic. In this situation theadvantages of SPRings are not so great and the need to protect every fibre is notas critical. Selective path protection schemes such as VC-Trail and SNCPprotection are more common in this situation. For example an end customer maydemand protection of their 2 Mbit/s leased line, this VC-12 path can beselectively protected by path protection.
Figure 6-16 Detection of Loss of VC-12
This path is protected at the VC-12 level across the entire network. If this pathwas only protected at a HO trunk layer, that is at the VC-4 level, by MSP or
MS-SPRing and there was a break in a lower layer fibre, this VC-12 path wouldbe lost. A full VC-4 circuit, however, would not be lost, so the protectionmechanism at the VC-4 layer would not detect the failure. An operator musttherefore, not only consider which protection scheme to employ but how thesewill interwork.
An effective deployment of subnetworks is interconnecting SNCP protectedsubnetworks and MS-SPRing protected subnetworks. For example anMS-SPRing subnetwork is ideal for the core or backbone network. This could beconnected to a regional or local network where subnetwork path protection wasused to apply selective protection to traffic.
Difficulties can arise at the network boundaries of the subnetworks. If singlepoints are used to interconnect the subnetworks, there is a vulnerable link in thechain so the interconnection between subnetworks becomes an issue. This isdiscussed in the following section.
At the upper regional layer of the network traffic hubbing can be used at anumber of points. Traffic is transported to multiple regional centres containing,for example, voice switches and routers, which are then connected via thebackbone. In these cases the traffic patterns can give capacity savings which arebetter than those for uniform traffic.
Interworking of Protection Schemes
As the size and demand placed on networks increases so their complexityincreases. Single ring or chain networks will rarely be implemented. Networkswill be made up of a number of sub-networks and each may have its ownindependent protection scheme. With the increasing number of operators, theinterconnection of networks belonging to different operators becomes an issue.These factors along with the drive for more reliable networks, means that theissue of interconnecting various individually protected subnetworks is ofincreasing importance.
Protection interworking is where more than one protection scheme operates ona single connection across a network. A single protection scheme may notprovide the optimum performance, it can be better to implement protection on asubnetwork basis, but then the interworking of these schemes must beconsidered
Trunk
RegionalMetro
LocalLocal
VC-12
VC-4
Fibrebreak
VC-4 protectiondoes not detectloss of VC-12 traffic
91 926 Protection
93 92 93 92Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Drivers for a Protection Interworking Scheme
Maximise traffic availability: Availability was defined previously as theprobability that an end to end connection is functioning. In a network of severalinterconnected subnetworks it should be ensured that the network can survivenot only failures in a single subnetwork, but also concurrent failures, that isfailures in several interconnecting subnetworks. Another consideration is thelink or links where these subnetworks interconnect. These must be as robust asthe subnetworks on either side.
Maintain protection Independence: The boundaries between subnetworksmay also represent administration or maintenance boundaries. It is desirable thatfailure in one subnetwork does not influence protection switching in aninterconnecting subnetwork. For example a planned outage for maintenancepurposes in one subnetwork should not effect switching in an interconnectingsubnetwork, particularly if this is managed by a different operator.
Interconnect subnetworks protecting at different layers: As described inprevious chapters an operator may adopt a multi-tier approach whereby there areseparate tiers for backbone, regional and local traffic, each of these will consistof different subnetworks. Administration in tiers may differ, for example trafficin the trunk tier administered at the VC-4 level, and in the regional tier at theVC-12 level. The interconnection of traffic and interworking of protectionschemes must be considered.
Interconnect networks using different protection schemes: An end to endpath is likely to transit several subnetworks and each may have a differentprotection scheme. To ensure that an end to end path is protected these schemesmust interwork. This is particularly important in countries where there areseveral operators and paths cross boundaries between operators.
Types of Protection Interworking
When a single end to end connection passes across several interconnectedsubnetworks, there are two types of protection interworking that can operate:concatenation (�chaining�) or nesting.
Figure 6-17 Concatenation and Nesting
Concatenation: The end to end connection is protected by �chaining� severalindependently protected subnetworks, that is, connecting the subnetworks inseries as shown in Figure 6-17. In each subnetwork a different protectionmechanism can operate on the connection. This is easier to understand andmanage than nesting as the protection is modular. Switching mechanisms do notinteract and so management is simpler.
Nesting: In nesting protection domains are overlaid, so that two mechanisms actsimultaneously on a single portion of the connection.
The example in Figure 6-17 illustrates nesting and concatenation for connectionX to Y. In the concatenated example the network can survive simultaneous faultsin each of the subnetworks, whilst a dedicated protection path from X to Y, inwhich only two end to end paths can be chosen would not survive the same fault.
SNCP MS-SPRingConcatenation
Subnetwork A Subnetwork B Subnetwork C
MS-SPRing
X Y
SNCP
MSP
A
BX Y
Nesting
93 946 Protection
95 94 95 94Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
Drivers for a Protection Interworking Scheme
Maximise traffic availability: Availability was defined previously as theprobability that an end to end connection is functioning. In a network of severalinterconnected subnetworks it should be ensured that the network can survivenot only failures in a single subnetwork, but also concurrent failures, that isfailures in several interconnecting subnetworks. Another consideration is thelink or links where these subnetworks interconnect. These must be as robust asthe subnetworks on either side.
Maintain protection Independence: The boundaries between subnetworksmay also represent administration or maintenance boundaries. It is desirable thatfailure in one subnetwork does not influence protection switching in aninterconnecting subnetwork. For example a planned outage for maintenancepurposes in one subnetwork should not effect switching in an interconnectingsubnetwork, particularly if this is managed by a different operator.
Interconnect subnetworks protecting at different layers: As described inprevious chapters an operator may adopt a multi-tier approach whereby there areseparate tiers for backbone, regional and local traffic, each of these will consistof different subnetworks. Administration in tiers may differ, for example trafficin the trunk tier administered at the VC-4 level, and in the regional tier at theVC-12 level. The interconnection of traffic and interworking of protectionschemes must be considered.
Interconnect networks using different protection schemes: An end to endpath is likely to transit several subnetworks and each may have a differentprotection scheme. To ensure that an end to end path is protected these schemesmust interwork. This is particularly important in countries where there areseveral operators and paths cross boundaries between operators.
Types of Protection Interworking
When a single end to end connection passes across several interconnectedsubnetworks, there are two types of protection interworking that can operate:concatenation (�chaining�) or nesting.
Figure 6-17 Concatenation and Nesting
Concatenation: The end to end connection is protected by �chaining� severalindependently protected subnetworks, that is, connecting the subnetworks inseries as shown in Figure 6-17. In each subnetwork a different protectionmechanism can operate on the connection. This is easier to understand andmanage than nesting as the protection is modular. Switching mechanisms do notinteract and so management is simpler.
Nesting: In nesting protection domains are overlaid, so that two mechanisms actsimultaneously on a single portion of the connection.
The example in Figure 6-17 illustrates nesting and concatenation for connectionX to Y. In the concatenated example the network can survive simultaneous faultsin each of the subnetworks, whilst a dedicated protection path from X to Y, inwhich only two end to end paths can be chosen would not survive the same fault.
SNCP MS-SPRingConcatenation
Subnetwork A Subnetwork B Subnetwork C
MS-SPRing
X Y
SNCP
MSP
A
BX Y
Nesting
93 946 Protection
95 94 95 94Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
The nesting example is generally less desirable as it is more complicated. It maynot be clear which mechanism switches first in the event of failure. Theimprovement in connection availability is likely to be small because althoughwhere the mechanisms are overlaid the connection is well protected, the pooreravailability of the rest of the connection will dominate the availability of theentire connection. Nesting although less desirable is often necessary as networktopology may dictate it.
Types of Interconnection
Figure 6-18 Subnetwork Interconnection
There are different ways of interconnecting subnetworks. Consider aconcatenated scheme. Figure 6-18 shows the single and dual node schemes thatcan be used. Dual node interconnection, means two nodes in each subnetworkare connected. Two paths are set up between each two subnetworks and so anend to end connection is protected against a failure in one of the subnetworks. Asingle failure at an interconnecting node is protected against or a loss of one ofthe interconnecting links. A single node interconnection scheme introduces asingle point of failure into the network. If the interconnecting link fails or one ofthe interconnecting nodes fails traffic will be lost. Even if 1+1 MSP is used on
the interconnecting link, the two nodes could be single points of failure.
There are two types of dual node interconnection:
Virtual Ring: Working and protection paths are physically different. There maybe two interconnecting nodes on each subnetwork or the interconnecting nodesmay be shared across the subnetworks as in Figure 6-18. This mechanism is asrobust as a single subnetwork, because there are two paths between thesubnetworks, and no single point of failure. If there were to be failures insubnetworks on each side of the interconnection, however, traffic would be lost.
Drop and Continue (Matched Nodes): This is a more robust form of dual nodeinterconnection. Traffic at the first node A is passed to the second subnetworkvia node B, but also continues to C and is passed to D, so two copies of the trafficare passed to the second subnetwork. In the event of concurrent failure in eachof the subnetworks traffic is not lost.
Drop and continue is also desirable because independence is maintainedbetween subnetworks, this is not the case with virtual rings. For subnetworkswhere boundaries represent boundaries between administrative regions ordifferent operator�s networks, this interconnection prevents failure and plannedoutage in one subnetwork affecting protection switching in the neighboringsubnetwork.
Drop and continue dual node interconnect can be used for the followingcombinations of subnetworks:
� SNCP subnetwork to SNCP subnetwork
� MS-SPRing to MS-SPRing
� MS-SPRing to SNCP subnetwork
� MS-SPRing to MS-Dedicated Protection Ring
� MS-Dedicated Protection Ring to SNCP subnetwork
Of these combinations, the last two are not required in ETSI markets.
Figure 6-19 illustrates matched node interconnection between SPRing andSNCP subnetworks. The figure shows both paths in the SNCP subnetwork toillustrate the use of drop and continue.
A
B
C
D
Single Node Dual NodeVirtual Ring
Dual NodeDrop and Continue
95 966 Protection
97 96 97 96Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
The nesting example is generally less desirable as it is more complicated. It maynot be clear which mechanism switches first in the event of failure. Theimprovement in connection availability is likely to be small because althoughwhere the mechanisms are overlaid the connection is well protected, the pooreravailability of the rest of the connection will dominate the availability of theentire connection. Nesting although less desirable is often necessary as networktopology may dictate it.
Types of Interconnection
Figure 6-18 Subnetwork Interconnection
There are different ways of interconnecting subnetworks. Consider aconcatenated scheme. Figure 6-18 shows the single and dual node schemes thatcan be used. Dual node interconnection, means two nodes in each subnetworkare connected. Two paths are set up between each two subnetworks and so anend to end connection is protected against a failure in one of the subnetworks. Asingle failure at an interconnecting node is protected against or a loss of one ofthe interconnecting links. A single node interconnection scheme introduces asingle point of failure into the network. If the interconnecting link fails or one ofthe interconnecting nodes fails traffic will be lost. Even if 1+1 MSP is used on
the interconnecting link, the two nodes could be single points of failure.
There are two types of dual node interconnection:
Virtual Ring: Working and protection paths are physically different. There maybe two interconnecting nodes on each subnetwork or the interconnecting nodesmay be shared across the subnetworks as in Figure 6-18. This mechanism is asrobust as a single subnetwork, because there are two paths between thesubnetworks, and no single point of failure. If there were to be failures insubnetworks on each side of the interconnection, however, traffic would be lost.
Drop and Continue (Matched Nodes): This is a more robust form of dual nodeinterconnection. Traffic at the first node A is passed to the second subnetworkvia node B, but also continues to C and is passed to D, so two copies of the trafficare passed to the second subnetwork. In the event of concurrent failure in eachof the subnetworks traffic is not lost.
Drop and continue is also desirable because independence is maintainedbetween subnetworks, this is not the case with virtual rings. For subnetworkswhere boundaries represent boundaries between administrative regions ordifferent operator�s networks, this interconnection prevents failure and plannedoutage in one subnetwork affecting protection switching in the neighboringsubnetwork.
Drop and continue dual node interconnect can be used for the followingcombinations of subnetworks:
� SNCP subnetwork to SNCP subnetwork
� MS-SPRing to MS-SPRing
� MS-SPRing to SNCP subnetwork
� MS-SPRing to MS-Dedicated Protection Ring
� MS-Dedicated Protection Ring to SNCP subnetwork
Of these combinations, the last two are not required in ETSI markets.
Figure 6-19 illustrates matched node interconnection between SPRing andSNCP subnetworks. The figure shows both paths in the SNCP subnetwork toillustrate the use of drop and continue.
A
B
C
D
Single Node Dual NodeVirtual Ring
Dual NodeDrop and Continue
95 966 Protection
97 96 97 96Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
C and F are primary interconnection nodes. D and J secondary interconnectionnodes. Traffic entering the SPRing at A and destined for H (blue line) passes viaB to the primary interconnection node C where the signal is dropped andcontinued to secondary interconnection node D. Each interconnection nodesends the signal to the SNCP subnetwork C to F and D to J. From nodes F andJ the signal is passed in opposite directions around the network to the destinationnode H. Node H selects between the two signals received.
Traffic entering the SNCP subnetwork at H (magenta line) is passed in bothdirections around the network to the interconnection node F and J. At eachinterconnection node the signal is dropped and continued to the otherinterconnection node. Each interconnection node selects from the two signalsand sends the selected signal to the SPRing. Secondary interconnection node Dpasses the signal to Primary interconnection node C. Node C selects between thesignals from F and D and passes the selected signal around the network to thedestination node A.
From an overall network perspective, matched node interconnection betweenSPRings and SNCP subnetworks enables an operator to apply the mosteffective subnetwork protection in each part of the network. SPRings are idealfor the core and much of the regional, metropolitan and dense urban networks,while subnetwork path protection is well suited to applying selective protectionin the regional, metropolitan, urban and, in particular, rural subnetworks.
Figure 6-19 SPRing to SNCP Subnetwork Matched Node Protection
SPRing
Subnetwork
A
B
C D
E
F
G
H
I
J
SNCPHO
97 986 Protection
99 98 99 98Protection Protection
Synchronous Transmission Systems Synchronous Transmission Systems
C and F are primary interconnection nodes. D and J secondary interconnectionnodes. Traffic entering the SPRing at A and destined for H (blue line) passes viaB to the primary interconnection node C where the signal is dropped andcontinued to secondary interconnection node D. Each interconnection nodesends the signal to the SNCP subnetwork C to F and D to J. From nodes F andJ the signal is passed in opposite directions around the network to the destinationnode H. Node H selects between the two signals received.
Traffic entering the SNCP subnetwork at H (magenta line) is passed in bothdirections around the network to the interconnection node F and J. At eachinterconnection node the signal is dropped and continued to the otherinterconnection node. Each interconnection node selects from the two signalsand sends the selected signal to the SPRing. Secondary interconnection node Dpasses the signal to Primary interconnection node C. Node C selects between thesignals from F and D and passes the selected signal around the network to thedestination node A.
From an overall network perspective, matched node interconnection betweenSPRings and SNCP subnetworks enables an operator to apply the mosteffective subnetwork protection in each part of the network. SPRings are idealfor the core and much of the regional, metropolitan and dense urban networks,while subnetwork path protection is well suited to applying selective protectionin the regional, metropolitan, urban and, in particular, rural subnetworks.
Figure 6-19 SPRing to SNCP Subnetwork Matched Node Protection
SPRing
Subnetwork
A
B
C D
E
F
G
H
I
J
SNCPHO
97 986 Protection
99 98 99 98Network Management Network management
Synchronous Transmission Systems
Network Management
The topic of network management is a complex one due to its application in anumber of different areas. The Post, Telephone and Telegraph companies(PTTs) initially deployed Operation Support Systems (OSS) for the telephoneservice. The development was primarily ’in house’ meeting the PTT�s businessrequirements for operation, administration, maintenance and provisioning(OAM&P). Cost reduction through automation was the key driver for thedevelopment of these OSSs.
A number of factors (advances in technology, deregulation, competition etc.)have resulted in new telecommunication services. Initially separate networkswere generated geared towards delivering particular services such as data, ATMand IP, resulting in an associated but separate management system.
The current trend is towards convergence and integration of these differenttechnologies. The network operator�s and the service provider�s requirementsare now changing.
Network management requirements for network operator’s and Service providers.
� Open framework/application �plug and play�, enabling the serviceprovider to interface ’applications’ simply and easily in order to gaina competitive edge or meet business objectives, for example,interfacing to a service level agreement (SLA) system.
� Interoperability. Interfacing with other service provider/networkoperator management systems to access data relating to servicesbeing carried between different network operators. For example, aleased line may span a number of different networks, operators mustprovide open access with the appropriate level of security to enabledata to be retrieved for the associated leased line.
� Multi-vendor. The ability to manage network elements supplied by avariety of manufacturers. Network operators do not want singlesupplier ’lock in’, they want freedom of choice enabling them to getthe best price for the functionality required.
� Multi-technology. Integration of other technologies on to onemanagement system such ATM, switching and IP. Convergingtechnologies with effectively one network providing a number ofdifferent services requires consolidation and integration of themanagement system. It also enables network operators to reduce theiroperating costs and to automate tasks.
� End-to-end trail management. Provisioning, surveillance andperformance monitoring on an end to end service basis. Enabling thetask of provisioning to be simplified with considerable reduction inthe time taken for the activation of new services. Also, if services aredisrupted, enabling the network operator to rapidly identify faults onthe network.
� Further automation of processes to reduce operational costs anderrors. Ideally network operators are striving for one touch flow, forexample, taking an order from a customer at the network operatorsfront office and providing inputs automatically for workforcemanagement, service level agreement, circuit provisioning andbilling etc.
With the emergence of the ISO OSI seven layer reference model, the opportunitywas taken to bring about interoperability in the management of transmissionnetworks. This opportunity was not missed in the definition of the SDHstandards which provide a method and a transmission format for networkmanagement.
7
99 1007 Network Management
101 100 101 100Network Management Network management
Synchronous Transmission Systems
Network Management
The topic of network management is a complex one due to its application in anumber of different areas. The Post, Telephone and Telegraph companies(PTTs) initially deployed Operation Support Systems (OSS) for the telephoneservice. The development was primarily ’in house’ meeting the PTT�s businessrequirements for operation, administration, maintenance and provisioning(OAM&P). Cost reduction through automation was the key driver for thedevelopment of these OSSs.
A number of factors (advances in technology, deregulation, competition etc.)have resulted in new telecommunication services. Initially separate networkswere generated geared towards delivering particular services such as data, ATMand IP, resulting in an associated but separate management system.
The current trend is towards convergence and integration of these differenttechnologies. The network operator�s and the service provider�s requirementsare now changing.
Network management requirements for network operator’s and Service providers.
� Open framework/application �plug and play�, enabling the serviceprovider to interface ’applications’ simply and easily in order to gaina competitive edge or meet business objectives, for example,interfacing to a service level agreement (SLA) system.
� Interoperability. Interfacing with other service provider/networkoperator management systems to access data relating to servicesbeing carried between different network operators. For example, aleased line may span a number of different networks, operators mustprovide open access with the appropriate level of security to enabledata to be retrieved for the associated leased line.
� Multi-vendor. The ability to manage network elements supplied by avariety of manufacturers. Network operators do not want singlesupplier ’lock in’, they want freedom of choice enabling them to getthe best price for the functionality required.
� Multi-technology. Integration of other technologies on to onemanagement system such ATM, switching and IP. Convergingtechnologies with effectively one network providing a number ofdifferent services requires consolidation and integration of themanagement system. It also enables network operators to reduce theiroperating costs and to automate tasks.
� End-to-end trail management. Provisioning, surveillance andperformance monitoring on an end to end service basis. Enabling thetask of provisioning to be simplified with considerable reduction inthe time taken for the activation of new services. Also, if services aredisrupted, enabling the network operator to rapidly identify faults onthe network.
� Further automation of processes to reduce operational costs anderrors. Ideally network operators are striving for one touch flow, forexample, taking an order from a customer at the network operatorsfront office and providing inputs automatically for workforcemanagement, service level agreement, circuit provisioning andbilling etc.
With the emergence of the ISO OSI seven layer reference model, the opportunitywas taken to bring about interoperability in the management of transmissionnetworks. This opportunity was not missed in the definition of the SDHstandards which provide a method and a transmission format for networkmanagement.
7
99 1007 Network Management
101 100 101 100Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
This chapter describes the management communications with the networkelement followed by a number of general concepts regarding the TMN hierarchyand functionality, and finally open network management platforms will bebriefly discussed.
The Physical Management Path
Within existing plesiochronous networks, no provision was made for a standardmanagement path within ITU-T Recommendations. To overcome the lack of amanagement channel, many manufacturers developed proprietary systemsbased on either the use of spare bits within the signal frame or via line codingmethods similar to those used in submarine systems.
Despite the restrictions on transmission rates, some of the methods employedare capable of supervision and monitoring of equipment and, to a limited extent,even remote configuration. The major drawbacks with these systems revolvearound the fact that management is restricted to a channel which can only beaccessed at a specific transmission rate, for example, 2 Mbit/s, 8 M/bit/s,34 Mbit/s, hence requiring multiplexing to gain access. This restriction limitsmanagement to a section by section facility. More importantly, it is impossiblefor interworking to exist between different systems. It is quite probable that onemanufacturer’s equipment will not even support a management channel betweentwo pieces of equipment from a third party.
SDH Management Communication Channel
With the introduction of SDH, the opportunity was taken to implement the idealsembodied in the OSI 7-layer reference model to define a management channel.This began with the definition of overhead capacity in the STM-1 frame, thusoffering a defined management channel for section by section communication.This was extended further to define overheads at the AU level and the TU levelthus providing management capacity over the core of the transmission networkand also a path management channel associated with a path extending from endto end across an entire network. This is illustrated in Figure 7-1.
Figure 7-1 Section, AU and TU Path Management
It is this ability to provide path management to the VC-12 (2 Mbit/s) level whichhighlights a significant advance in telecommunications management with theintroduction of the SDH. The move towards management standards, however,did not stop at this point. ITU-T Recommendations G.783 and G.784 go on topropose how the management channel should be used and further proposes theprotocols which should be employed for the remaining six layers of the OSImodel. The management channel is referred to as the Data CommunicationChannel (DCC) or sometimes the Embedded Control Channel (ECC).
In these clear definitions, and in the continuing work of the SDH Study Groups,the first major steps have been taken towards the implementation of OpenNetwork Management Systems in the telecommunications industry.
AU Port Associations
Section
Section Section
Section Section
AU Path
TU Path
TU AP
AU AP
TU PathTermination
TU AP
TU PathTermination
AU AP
AU AP
DXC Node
AU AP
MTN Structure
AU AP - AU Access PointTU AP - TU Access Point
101 1027 Network Management
103 102 103 102Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
This chapter describes the management communications with the networkelement followed by a number of general concepts regarding the TMN hierarchyand functionality, and finally open network management platforms will bebriefly discussed.
The Physical Management Path
Within existing plesiochronous networks, no provision was made for a standardmanagement path within ITU-T Recommendations. To overcome the lack of amanagement channel, many manufacturers developed proprietary systemsbased on either the use of spare bits within the signal frame or via line codingmethods similar to those used in submarine systems.
Despite the restrictions on transmission rates, some of the methods employedare capable of supervision and monitoring of equipment and, to a limited extent,even remote configuration. The major drawbacks with these systems revolvearound the fact that management is restricted to a channel which can only beaccessed at a specific transmission rate, for example, 2 Mbit/s, 8 M/bit/s,34 Mbit/s, hence requiring multiplexing to gain access. This restriction limitsmanagement to a section by section facility. More importantly, it is impossiblefor interworking to exist between different systems. It is quite probable that onemanufacturer’s equipment will not even support a management channel betweentwo pieces of equipment from a third party.
SDH Management Communication Channel
With the introduction of SDH, the opportunity was taken to implement the idealsembodied in the OSI 7-layer reference model to define a management channel.This began with the definition of overhead capacity in the STM-1 frame, thusoffering a defined management channel for section by section communication.This was extended further to define overheads at the AU level and the TU levelthus providing management capacity over the core of the transmission networkand also a path management channel associated with a path extending from endto end across an entire network. This is illustrated in Figure 7-1.
Figure 7-1 Section, AU and TU Path Management
It is this ability to provide path management to the VC-12 (2 Mbit/s) level whichhighlights a significant advance in telecommunications management with theintroduction of the SDH. The move towards management standards, however,did not stop at this point. ITU-T Recommendations G.783 and G.784 go on topropose how the management channel should be used and further proposes theprotocols which should be employed for the remaining six layers of the OSImodel. The management channel is referred to as the Data CommunicationChannel (DCC) or sometimes the Embedded Control Channel (ECC).
In these clear definitions, and in the continuing work of the SDH Study Groups,the first major steps have been taken towards the implementation of OpenNetwork Management Systems in the telecommunications industry.
AU Port Associations
Section
Section Section
Section Section
AU Path
TU Path
TU AP
AU AP
TU PathTermination
TU AP
TU PathTermination
AU AP
AU AP
DXC Node
AU AP
MTN Structure
AU AP - AU Access PointTU AP - TU Access Point
101 1027 Network Management
103 102 103 102Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
The TMN Layered Hierarchy
Figure 7-2 shows a simplified model of the TMN layered hierarchy. Thedefinition of distinct levels may vary due to system size and the managementstrategy, however, the principle structure remains the same.
Figure 7-2 The TMN Layered Hierarchy
Business Management Layer
This layer supports the high-end business functions such as resource planning,financial planning, marketing database management etc.
Service Management Layer
This layer supports the service delivery functions such as customer care, serviceconfiguration, order management, work force administration and management.
Network Management layer
This layer supports the monitoring and control of the whole network, which mayconsist of different network element types, for example SDH/SONET,switching, ATM etc., with the possibility of network elements from differentsuppliers.
Tasks on a network wide basis such as network monitoring, surveillance,network planning, network/system administration and OAM&P are included inthis layer.
The management system may also be required to perform more analyticalprocessing, such as performance monitoring and cost analysis.
The degree of functionality within the NML, as with the other layers, may vary.A comparison is shown in Table 7-1.
Element Management Layer
The element management layer would provide many of the facilities describedin the following section. Also, it would be expected to support additionalmanagement packages to provide the functions of financial, resource andmaintenance analysis on the information it collects.
Although the element management may reside within a network element, it ismore likely that it will be a software package implemented on some operatingsystem/hardware platform. The size of the platform and its capabilities may varydue to the need for the element manager to monitor and control various sizeddomains. The management systems must, however, offer the capability ofmigrating from smaller to larger systems as a network expands.
BML
SML
NML
EML
NEL
NM
EM
NE NE
NE
NE
NE
NE NE
NE
NE
NE
NENE
NENE
NE
EM EM EM
Business and service applications
BMLSMLNMLEMLNEL
-----
Business Management LayerService Management LayerNetwork Management LayerElement Management LayerNetwork Element Later
NMEMNE
Network ManagerElement ManagerNetwork Element
---
103 1047 Network Management
105 104 105 104Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
The TMN Layered Hierarchy
Figure 7-2 shows a simplified model of the TMN layered hierarchy. Thedefinition of distinct levels may vary due to system size and the managementstrategy, however, the principle structure remains the same.
Figure 7-2 The TMN Layered Hierarchy
Business Management Layer
This layer supports the high-end business functions such as resource planning,financial planning, marketing database management etc.
Service Management Layer
This layer supports the service delivery functions such as customer care, serviceconfiguration, order management, work force administration and management.
Network Management layer
This layer supports the monitoring and control of the whole network, which mayconsist of different network element types, for example SDH/SONET,switching, ATM etc., with the possibility of network elements from differentsuppliers.
Tasks on a network wide basis such as network monitoring, surveillance,network planning, network/system administration and OAM&P are included inthis layer.
The management system may also be required to perform more analyticalprocessing, such as performance monitoring and cost analysis.
The degree of functionality within the NML, as with the other layers, may vary.A comparison is shown in Table 7-1.
Element Management Layer
The element management layer would provide many of the facilities describedin the following section. Also, it would be expected to support additionalmanagement packages to provide the functions of financial, resource andmaintenance analysis on the information it collects.
Although the element management may reside within a network element, it ismore likely that it will be a software package implemented on some operatingsystem/hardware platform. The size of the platform and its capabilities may varydue to the need for the element manager to monitor and control various sizeddomains. The management systems must, however, offer the capability ofmigrating from smaller to larger systems as a network expands.
BML
SML
NML
EML
NEL
NM
EM
NE NE
NE
NE
NE
NE NE
NE
NE
NE
NENE
NENE
NE
EM EM EM
Business and service applications
BMLSMLNMLEMLNEL
-----
Business Management LayerService Management LayerNetwork Management LayerElement Management LayerNetwork Element Later
NMEMNE
Network ManagerElement ManagerNetwork Element
---
103 1047 Network Management
105 104 105 104Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
Network Element Layer
There is a degree of management that resides within the network elementsthemselves, and it is feasible that the element manager for a particularmanagement domain may physically reside within a network element.
Basic functionality within the element should include the facilities listed in thesubsequent section applied to the single element. In some circumstances thedecision may be taken to implement a distributed management system wherebyindividual elements perform a high degree of the functionality described. Suchan implementation has a number of advantages with regards to the speed atwhich the network as a whole can react to various events, in particular the caseof path restoration for protection purposes.
The alternative is an element with a minimum functionality, allowingmanagement functions to be performed at the Element Management Layer. Acomparison of the benefits of each strategy is shown in Table 7-1.
Functionality of a Network Management System
The classification of network management functions is described in ISO, asbelow. The functionality of the management system should include thesefeatures via the initial systems with a provision for additional feature packagesor modules.
� Configuration management
� Fault management
� Performance management
� Security management
� Accounting management
The network manager is not restricted in its capabilities to the management ofSDH equipment only. Managed objects defined in line with Open Networkstandards could extend to include the following items within its managementdomain.
� Network Elements
� Test Equipment
Table 7-1 Processing implementation options for network management layer, element management layer and network management layer
Hig
h le
vel o
f p
roce
ssin
g in
N
E O
pti
on
c
Net
wor
k da
taba
se m
ust
rely
on
upda
tes
from
NE
vi
a E
MR
educ
ed p
roce
ssin
g of
ev
ents
and
adm
inis
trat
ion
Pro
cess
ing
can
conc
entr
ate
on N
etw
ork
Ana
lysi
s an
d
Red
uced
pro
cess
ing
sinc
e in
form
atio
n is
filte
red
at th
e N
E le
vel
Dom
ain
data
base
pro
babl
y st
ill h
eld
at th
is le
vel
Incr
ease
d nu
mbe
r of
el
emen
ts c
an b
e ef
ficie
ntly
m
anag
ed
Req
uire
s hi
gh p
roce
ssin
g po
wer
in N
EV
ery
fast
loca
l con
figur
atio
n C
onne
ctio
n to
adj
acen
t NE
s al
low
s fa
st in
tra-
dom
ain
reco
nfig
urat
ion
Inte
r-do
mai
n re
conf
igur
atio
n m
ust b
e co
nfirm
ed a
t the
EM
le
vel o
r po
ssib
ly N
ML
Hig
h le
vel o
f p
roce
ssin
g a
t E
M O
pti
on
b
Mus
t rel
y on
upd
ates
from
EM
fo
r th
e ne
twor
k da
taba
se
(tha
t is,
pic
ture
of n
etw
ork)
Red
uced
pro
cess
ing
resp
onsi
bilit
y al
low
s a
redu
ctio
n in
the
requ
ired
relia
bilit
y
Req
uire
s a
rang
e of
EM
S to
de
al w
ith v
ario
us s
izes
of
man
agem
ent d
omai
nP
rovi
des
fast
rec
onfig
urat
ion
with
in a
dom
ain
Req
uire
s co
mm
unic
atio
n w
ith
othe
r E
MS
for
inte
r-do
mai
n tr
affic
Som
e pr
oces
sing
req
uire
d to
pr
even
t exc
essi
ve lo
ad o
n E
MR
elat
ivel
y fa
st r
econ
figur
atio
n fr
om E
M s
ome
dela
y in
co
nfirm
ing
conf
igur
atio
n w
ith
EM
dat
abas
e
Hig
h le
vel o
f p
roce
ssin
g a
t N
ML
Op
tio
n a
Acc
urat
e pi
ctur
e of
net
wor
k at
a c
entr
al s
iteR
equi
res
very
pow
erfu
l m
anag
er to
cop
e w
ith h
igh
lle
vel o
f pro
cess
ing
Sys
tem
mus
t be
very
rel
iabl
eS
low
con
trol
and
con
figur
atio
n
Low
leve
l of p
roce
ssin
g re
quire
d at
EM
leve
lA
sin
gle
EM
can
con
trol
a
larg
e nu
mbe
r of
Nes
All
reco
nfig
urat
ion
mus
t be
conf
irmed
with
the
NM
LR
equi
res
high
rel
iabi
lity
in E
ML
Req
uire
s ve
ry lo
w p
roce
ssin
g po
wer
All
inte
llige
nce
at a
hig
her
leve
l hen
ce n
o de
cisi
on
taki
ngS
low
rec
onfig
urat
ion
NE
TW
OR
K
MA
NA
GE
ME
NT
LA
YE
R (
NM
L)
ELE
ME
NT
M
AN
AG
EM
EN
T
LAY
ER
(E
ML)
NE
TW
OR
K
ELE
ME
NT
LA
YE
R (
NE
L)
105 1067 Network Management
107 106 107 106Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
Network Element Layer
There is a degree of management that resides within the network elementsthemselves, and it is feasible that the element manager for a particularmanagement domain may physically reside within a network element.
Basic functionality within the element should include the facilities listed in thesubsequent section applied to the single element. In some circumstances thedecision may be taken to implement a distributed management system wherebyindividual elements perform a high degree of the functionality described. Suchan implementation has a number of advantages with regards to the speed atwhich the network as a whole can react to various events, in particular the caseof path restoration for protection purposes.
The alternative is an element with a minimum functionality, allowingmanagement functions to be performed at the Element Management Layer. Acomparison of the benefits of each strategy is shown in Table 7-1.
Functionality of a Network Management System
The classification of network management functions is described in ISO, asbelow. The functionality of the management system should include thesefeatures via the initial systems with a provision for additional feature packagesor modules.
� Configuration management
� Fault management
� Performance management
� Security management
� Accounting management
The network manager is not restricted in its capabilities to the management ofSDH equipment only. Managed objects defined in line with Open Networkstandards could extend to include the following items within its managementdomain.
� Network Elements
� Test Equipment
Table 7-1 Processing implementation options for network management layer, element management layer and network management layer
Hig
h le
vel o
f p
roce
ssin
g in
N
E O
pti
on
c
Net
wor
k da
taba
se m
ust
rely
on
upda
tes
from
NE
vi
a E
MR
educ
ed p
roce
ssin
g of
ev
ents
and
adm
inis
trat
ion
Pro
cess
ing
can
conc
entr
ate
on N
etw
ork
Ana
lysi
s an
d
Red
uced
pro
cess
ing
sinc
e in
form
atio
n is
filte
red
at th
e N
E le
vel
Dom
ain
data
base
pro
babl
y st
ill h
eld
at th
is le
vel
Incr
ease
d nu
mbe
r of
el
emen
ts c
an b
e ef
ficie
ntly
m
anag
ed
Req
uire
s hi
gh p
roce
ssin
g po
wer
in N
EV
ery
fast
loca
l con
figur
atio
n C
onne
ctio
n to
adj
acen
t NE
s al
low
s fa
st in
tra-
dom
ain
reco
nfig
urat
ion
Inte
r-do
mai
n re
conf
igur
atio
n m
ust b
e co
nfirm
ed a
t the
EM
le
vel o
r po
ssib
ly N
ML
Hig
h le
vel o
f p
roce
ssin
g a
t E
M O
pti
on
b
Mus
t rel
y on
upd
ates
from
EM
fo
r th
e ne
twor
k da
taba
se
(tha
t is,
pic
ture
of n
etw
ork)
Red
uced
pro
cess
ing
resp
onsi
bilit
y al
low
s a
redu
ctio
n in
the
requ
ired
relia
bilit
y
Req
uire
s a
rang
e of
EM
S to
de
al w
ith v
ario
us s
izes
of
man
agem
ent d
omai
nP
rovi
des
fast
rec
onfig
urat
ion
with
in a
dom
ain
Req
uire
s co
mm
unic
atio
n w
ith
othe
r E
MS
for
inte
r-do
mai
n tr
affic
Som
e pr
oces
sing
req
uire
d to
pr
even
t exc
essi
ve lo
ad o
n E
MR
elat
ivel
y fa
st r
econ
figur
atio
n fr
om E
M s
ome
dela
y in
co
nfirm
ing
conf
igur
atio
n w
ith
EM
dat
abas
e
Hig
h le
vel o
f p
roce
ssin
g a
t N
ML
Op
tio
n a
Acc
urat
e pi
ctur
e of
net
wor
k at
a c
entr
al s
iteR
equi
res
very
pow
erfu
l m
anag
er to
cop
e w
ith h
igh
lle
vel o
f pro
cess
ing
Sys
tem
mus
t be
very
rel
iabl
eS
low
con
trol
and
con
figur
atio
n
Low
leve
l of p
roce
ssin
g re
quire
d at
EM
leve
lA
sin
gle
EM
can
con
trol
a
larg
e nu
mbe
r of
Nes
All
reco
nfig
urat
ion
mus
t be
conf
irmed
with
the
NM
LR
equi
res
high
rel
iabi
lity
in E
ML
Req
uire
s ve
ry lo
w p
roce
ssin
g po
wer
All
inte
llige
nce
at a
hig
her
leve
l hen
ce n
o de
cisi
on
taki
ngS
low
rec
onfig
urat
ion
NE
TW
OR
K
MA
NA
GE
ME
NT
LA
YE
R (
NM
L)
ELE
ME
NT
M
AN
AG
EM
EN
T
LAY
ER
(E
ML)
NE
TW
OR
K
ELE
ME
NT
LA
YE
R (
NE
L)
105 1067 Network Management
107 106 107 106Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
� Manpower
� Other Management Systems
Such a management system would be expected not only to manage synchronousnetwork elements, but also possess the ability to manage additional equipmentin the network via direct communications or another management system. Sincemany existing management systems are proprietary in nature, this infers the useof some type of mediation device between the two systems. In addition to therequired functionality described above, a management system would beexpected to offer the ability to operate enhanced packages offering features forTraffic analysis, Maintenance costing, Failure analysis etc.
Configuration Management
Configuration management provides a mechanism for managing the networkelements, often called objects, which are under the control of the managementsystem. The system should have the facility for:
� Changing the configuration
� Initialing objects, shutting them down and removing them fromservice.
� Collecting state information on a regular and on a demand basis.
� Provisioning services and resources to meet demand.
Examples of such activities are:
� Connecting end-to-end service.
� Setting up alternative routing options under fault conditions.
� Configuring gains on cards
� Providing alternative configurations depending on time of day
� Downline loading of information
An important aspect of configuration management is name management. Thisallows the user to symbolically name and refer to resources on the network.Several techniques exist for this, the most popular, and the ones for which ISOis developing standards, are called the ’white pages’ and ’yellow pages’ protocols,which are analogous with the white and yellow pages of a telephone directory.
Distributed facilities are often used for name management in which each area isresponsible for managing the names within the territory.
Fault Management
This area of management is often referred to as Event management, however, aclear distinction should be made between Faults and Events. An event is achange of status occurring somewhere within the managed domain. A fault oralarm is an indication of an actual or potential failure which may occur as theresult of an event.
This distinction highlights one of the most important functions of the eventmanager and that is in the filtering and management of alarms resulting fromvarious events. A single event may result in a multitude of alarms (for example,a break in a fibre will result in a very large number of alarms), but themanagement system must be capable of translating the multitude into a singlealarm which identifies the problem.
The management system should be capable of allowing these Alarm/Eventrelationships to be configured and re-configured to a network Operator’spreference, or as the result of changes in the network structure. In addition to thisthe Event Manager would possess the ability to record all events, but presentonly meaningful messages to the various levels of management.
Performance Management
The aim of performance management is to monitor and improve on theperformance of the network. It gathers statistical information to enable bothlong term planning and prediction of short term trends. A large network wouldrequire many monitor points to collect information and some method of localanalysis and filtering to prevent the network and the user from being swampedwith data. Performance monitoring requires the maintenance of logs and objectsand states, so as to show trends, and the adjustment of network resources inresponse to trends.
Examples of functions made available could include statistical analysis ofevents and alarms and a measurement of their effect and time to correct. Fromthis information estimates may be extracted as to the Quality of Serviceprovided on individual paths. Another function could be the use of historicaldata to predict trends in equipment deterioration or failure. These functions
107 1087 Network Management
109 108 109 108Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
� Manpower
� Other Management Systems
Such a management system would be expected not only to manage synchronousnetwork elements, but also possess the ability to manage additional equipmentin the network via direct communications or another management system. Sincemany existing management systems are proprietary in nature, this infers the useof some type of mediation device between the two systems. In addition to therequired functionality described above, a management system would beexpected to offer the ability to operate enhanced packages offering features forTraffic analysis, Maintenance costing, Failure analysis etc.
Configuration Management
Configuration management provides a mechanism for managing the networkelements, often called objects, which are under the control of the managementsystem. The system should have the facility for:
� Changing the configuration
� Initialing objects, shutting them down and removing them fromservice.
� Collecting state information on a regular and on a demand basis.
� Provisioning services and resources to meet demand.
Examples of such activities are:
� Connecting end-to-end service.
� Setting up alternative routing options under fault conditions.
� Configuring gains on cards
� Providing alternative configurations depending on time of day
� Downline loading of information
An important aspect of configuration management is name management. Thisallows the user to symbolically name and refer to resources on the network.Several techniques exist for this, the most popular, and the ones for which ISOis developing standards, are called the ’white pages’ and ’yellow pages’ protocols,which are analogous with the white and yellow pages of a telephone directory.
Distributed facilities are often used for name management in which each area isresponsible for managing the names within the territory.
Fault Management
This area of management is often referred to as Event management, however, aclear distinction should be made between Faults and Events. An event is achange of status occurring somewhere within the managed domain. A fault oralarm is an indication of an actual or potential failure which may occur as theresult of an event.
This distinction highlights one of the most important functions of the eventmanager and that is in the filtering and management of alarms resulting fromvarious events. A single event may result in a multitude of alarms (for example,a break in a fibre will result in a very large number of alarms), but themanagement system must be capable of translating the multitude into a singlealarm which identifies the problem.
The management system should be capable of allowing these Alarm/Eventrelationships to be configured and re-configured to a network Operator’spreference, or as the result of changes in the network structure. In addition to thisthe Event Manager would possess the ability to record all events, but presentonly meaningful messages to the various levels of management.
Performance Management
The aim of performance management is to monitor and improve on theperformance of the network. It gathers statistical information to enable bothlong term planning and prediction of short term trends. A large network wouldrequire many monitor points to collect information and some method of localanalysis and filtering to prevent the network and the user from being swampedwith data. Performance monitoring requires the maintenance of logs and objectsand states, so as to show trends, and the adjustment of network resources inresponse to trends.
Examples of functions made available could include statistical analysis ofevents and alarms and a measurement of their effect and time to correct. Fromthis information estimates may be extracted as to the Quality of Serviceprovided on individual paths. Another function could be the use of historicaldata to predict trends in equipment deterioration or failure. These functions
107 1087 Network Management
109 108 109 108Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
could provide valuable information in order to reduce the cost ofrepair/maintenance whilst increasing the availability of equipment throughpreventative maintenance.
Security Management
Security management can be divided into three areas:
� Physical Security, which is primarily the responsibility of the systemadministrator, since this relates to the security of the building wherethe manager resides.
� Access Security, which relates to the availability of the manager todifferent users and the maintaining of access during failures incommunications. Security management, therefore, primarilydetermines who may do what in controlling a network. It is usual toprovide security access in a layered structure, the upper layers beingable to perform all the functions available to the lower layers.
� Data Security, which relates to the storage of data by the NetworkOperator. In most cases all data needs some form of back up forsecurity against system faults. In addition to this there is a need forsecure storage of sensitive data to prevent unauthorised access. Insome countries the storage of such data is covered by legislation andthe encryption techniques used must adhere to such laws.
Although shown as a separate function, security management cuts across severalother functional areas, for example the need to limit access to users for certainconfiguration management functions.
New services such as customer network management introduced by the rapidgrowth in the internet, has now provided a focus on security. The end customerusing a PC can view performance reports relating to service level agreements oraccess network information.
Obviously the network operator would have ultimate control of the access andcontrol enabled for that particular end customer.
The appropriate steps have to be taken to ensure access is tightly controlled,there may be information on the network that may be of a sensitive nature thatunder any circumstances should not be accessible by the end customer.
Accounting Management
Accounting management aids in the preparation of bills for network users andfor tracking their payment. It also helps in the sale of network resources. It is theset of facilities which enables charges to be determined for the user of thenetwork resources and for costs to be identified and allocated to each resource.
This management function depends on statistics provided by the objects on thenetwork. Once again there is interaction between the various functions, forexample accounting management and configuration management. The lowesttransmission cost is often determined by the cheapest route or the cheapest timezone, both of which are controlled by configuration management.
Accounting management is frequently considered to include inventorymanagement. This function keeps track of the individual elements beingmanaged on the network, their characteristics, asset values and ownership andcontractual information.
Additional functionality
Customer network management(CNM)
The ubiquity of the internet together with the end customer’s need forinformation relating to their services, for example, reports relating to servicelevel agreements, has enabled network operators to offer customer networkmanagement. A PC located in the end customer’s premises for instance with theassociated security and access provides a possible window into the networkunder the full control of the network operator.
Process automation
Current work under study by the various standard bodies and forums in this areaare focusing on application plug and play with automation of processes,providing guidance and recommendations for an open framework using the besttechnology available. The goal is to reduce significantly the manual interventionrequired for performing the tasks.
Interconnection between the operators management systems is a requirementstemming from the deregulation and the subsequent competition.
109 1107 Network Management
111 110 111 110Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
could provide valuable information in order to reduce the cost ofrepair/maintenance whilst increasing the availability of equipment throughpreventative maintenance.
Security Management
Security management can be divided into three areas:
� Physical Security, which is primarily the responsibility of the systemadministrator, since this relates to the security of the building wherethe manager resides.
� Access Security, which relates to the availability of the manager todifferent users and the maintaining of access during failures incommunications. Security management, therefore, primarilydetermines who may do what in controlling a network. It is usual toprovide security access in a layered structure, the upper layers beingable to perform all the functions available to the lower layers.
� Data Security, which relates to the storage of data by the NetworkOperator. In most cases all data needs some form of back up forsecurity against system faults. In addition to this there is a need forsecure storage of sensitive data to prevent unauthorised access. Insome countries the storage of such data is covered by legislation andthe encryption techniques used must adhere to such laws.
Although shown as a separate function, security management cuts across severalother functional areas, for example the need to limit access to users for certainconfiguration management functions.
New services such as customer network management introduced by the rapidgrowth in the internet, has now provided a focus on security. The end customerusing a PC can view performance reports relating to service level agreements oraccess network information.
Obviously the network operator would have ultimate control of the access andcontrol enabled for that particular end customer.
The appropriate steps have to be taken to ensure access is tightly controlled,there may be information on the network that may be of a sensitive nature thatunder any circumstances should not be accessible by the end customer.
Accounting Management
Accounting management aids in the preparation of bills for network users andfor tracking their payment. It also helps in the sale of network resources. It is theset of facilities which enables charges to be determined for the user of thenetwork resources and for costs to be identified and allocated to each resource.
This management function depends on statistics provided by the objects on thenetwork. Once again there is interaction between the various functions, forexample accounting management and configuration management. The lowesttransmission cost is often determined by the cheapest route or the cheapest timezone, both of which are controlled by configuration management.
Accounting management is frequently considered to include inventorymanagement. This function keeps track of the individual elements beingmanaged on the network, their characteristics, asset values and ownership andcontractual information.
Additional functionality
Customer network management(CNM)
The ubiquity of the internet together with the end customer’s need forinformation relating to their services, for example, reports relating to servicelevel agreements, has enabled network operators to offer customer networkmanagement. A PC located in the end customer’s premises for instance with theassociated security and access provides a possible window into the networkunder the full control of the network operator.
Process automation
Current work under study by the various standard bodies and forums in this areaare focusing on application plug and play with automation of processes,providing guidance and recommendations for an open framework using the besttechnology available. The goal is to reduce significantly the manual interventionrequired for performing the tasks.
Interconnection between the operators management systems is a requirementstemming from the deregulation and the subsequent competition.
109 1107 Network Management
111 110 111 110Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
Network Management Platform
A network management platform is a collection of hardware and standardisedsoftware modules, which supports the various management functions. Itcomprises the processor, storage devices, communication devices, man machineinterface (MMI), operating systems, communications software, databasemanager, maintenance software and possibly additional support software forenhanced functions. The management system may comprise a number ofplatforms using a distributed architecture.
Such a requirement exists for all management systems, however, forsynchronous systems the concept of open standards can be extended to theconcepts of an open management platform. Figure 7-3 gives a pictorialrepresentation of a modular management platform. It shows how specificsynchronous applications, down to a customer’s site can be built on to acollection of generic applications and a hardware and software platform. Detailsof the various modules are not included in this book, but the Generic ApplicationInterface is one area of particular interest. It is this package which allows thevarious management applications to operate on the platform and via this thestructure can provide an open platform for a variety of software packages.
The Telecommunication Management Network
Several groups have taken the OSI network management model and developedit further for specific applications. The best known of these is theTelecommunications Management Network (TMN). Work started on TMN in1985 and it has since been developed by ANSI, ITU-T and ESTI. Figure 7-3 Network Management Platform Structure
The basic concept behind a TMN is to provide an organised network structure,to achieve the interconnection between various types of Operations Systems(OS) and telecommunication equipment using an agreed architecture and withstandardised interfaces. Figure 7-4 shows the relationship between the TMNand a telecommunication network. The telecommunications network canconsist of both digital and analogue telecommunications equipment. As in thecase of the OSI model, the telecommunications network is considered to consistof managed objects (network elements). These may be physical elements, suchas exchanges, transmission equipment, cable, cross-connects, or it may consistof abstract elements, such as maintenance entities and support entities.
Defaultand Site
Configuration
CustomerSpecific
Applications
SDH Specific Applications
Con
figur
atio
n
Acc
ount
ing
Eve
n t
Sec
urity
Gen
eral
Adm
in
Perfo
rman
ce
Common Applications
General Applications Interface
UNIXCommsStack RDBMS HCI Etc... Etc...
Hardware Platform
111 1127 Network Management
113 112 113 112Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
Network Management Platform
A network management platform is a collection of hardware and standardisedsoftware modules, which supports the various management functions. Itcomprises the processor, storage devices, communication devices, man machineinterface (MMI), operating systems, communications software, databasemanager, maintenance software and possibly additional support software forenhanced functions. The management system may comprise a number ofplatforms using a distributed architecture.
Such a requirement exists for all management systems, however, forsynchronous systems the concept of open standards can be extended to theconcepts of an open management platform. Figure 7-3 gives a pictorialrepresentation of a modular management platform. It shows how specificsynchronous applications, down to a customer’s site can be built on to acollection of generic applications and a hardware and software platform. Detailsof the various modules are not included in this book, but the Generic ApplicationInterface is one area of particular interest. It is this package which allows thevarious management applications to operate on the platform and via this thestructure can provide an open platform for a variety of software packages.
The Telecommunication Management Network
Several groups have taken the OSI network management model and developedit further for specific applications. The best known of these is theTelecommunications Management Network (TMN). Work started on TMN in1985 and it has since been developed by ANSI, ITU-T and ESTI. Figure 7-3 Network Management Platform Structure
The basic concept behind a TMN is to provide an organised network structure,to achieve the interconnection between various types of Operations Systems(OS) and telecommunication equipment using an agreed architecture and withstandardised interfaces. Figure 7-4 shows the relationship between the TMNand a telecommunication network. The telecommunications network canconsist of both digital and analogue telecommunications equipment. As in thecase of the OSI model, the telecommunications network is considered to consistof managed objects (network elements). These may be physical elements, suchas exchanges, transmission equipment, cable, cross-connects, or it may consistof abstract elements, such as maintenance entities and support entities.
Defaultand Site
Configuration
CustomerSpecific
Applications
SDH Specific Applications
Con
figur
atio
n
Acc
ount
ing
Eve
n t
Sec
urity
Gen
eral
Adm
in
Perfo
rman
ceCommon Applications
General Applications Interface
UNIXCommsStack RDBMS HCI Etc... Etc...
Hardware Platform
111 1127 Network Management
113 112 113 112Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 7-4 Relationship between TMN and Telecommunications Network
The TMN architecture consists of three elements: functional architecture;information architecture; and physical architecture. The functional architecturemay be considered to be the building blocks that allow complex systems to bebuilt. The information architecture describes the nature of information thatneeds to be exchanged between the functional building blocks. The physicalarchitecture of TMN describes the interfaces that have to be implemented, alongwith examples of physical components that make up the TMN.
Data Communications Network
TransmissionSystem
Exchange Exchange ExchangeTransmissionSystem
OperationsSystem
OperationsSystem
OperationsSystem
TMN
113 1147 Network Management
115 114 115 114Network Management Network Management
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 7-4 Relationship between TMN and Telecommunications Network
The TMN architecture consists of three elements: functional architecture;information architecture; and physical architecture. The functional architecturemay be considered to be the building blocks that allow complex systems to bebuilt. The information architecture describes the nature of information thatneeds to be exchanged between the functional building blocks. The physicalarchitecture of TMN describes the interfaces that have to be implemented, alongwith examples of physical components that make up the TMN.
Data Communications Network
TransmissionSystem
Exchange Exchange ExchangeTransmissionSystem
OperationsSystem
OperationsSystem
OperationsSystem
TMN
113 1147 Network Management
115 114 115 114High Capacity Networks High Capacity Networks
Synchronous Transmission Systems
High Capacity Networks
The rapid increase in traffic carried on today�s networks means that manyoperators are now close to exhausting the capacity of their installed fibre. Voice,leased line and mobile traffic are increasing, but the extremely rapid increase inbandwidth requirements is due to the faster growing, newer service types.Internet, ATM, data and video traffic are not only growing at a rapid rate but havelarge bandwidth requirements.
Operators are faced with the prospect of either the costly installation of newfibre or increasing the capacity of their existing fibre. This issue is more criticalfor new, emerging operators, who have less installed infrastructure or may needto lease dark fibres. The rapid growth of their traffic is forcing them to look atnew ways of expanding capacity on their existing fibre while keeping capitalcost to a minimum.
Technology has now advanced permitting the deployment of STM-64 systems,however, Wavelength Division Multiplexing (WDM) has also emerged as atechnique to increase the capacity of fibre. This involves transmitting multipleoptical signals down a single fibre. Each optical signal operates at a differentwavelength thus preventing interaction between them. For example severalSTM-16 or STM-64 traffic streams can be transmitted down one fibre. This notonly combats fibre exhaust issues, but also reduces the requirement for multipleregenerators on long spans. A single optical amplifier can be used to amplify thecombined signal hence replacing a regenerator mountain (see Figure 8-3).
Figure 8-1 Methods of Expanding Network Capacity
Networks are now being deployed with a purely optical layer, utilising thisWDM technique. This will permit the transport and routing of traffic withoutconversion into the electrical domain. To route, protect and manage traffic in anoptical network is currently constrained by technology and lack of standards,however, offering a flexible optical overlay layer is an area of great interest anddevelopment. ATM, IP and legacy data are already being carried over DenseWDM (D-WDM), bypassing the SDH layer. Optical Add/Drop multiplexers andsimple optical protection schemes are also now being deployed.
In the short term operators are facing the challenge of how to increase thecapacity of their networks quickly, however, they must also consider the longterm evolution of their networks. Long distance and metropolitan WDM opticalnetworks are designed to meet different requirements. Long distance operatorswant to minimise the cost per managed bit per fibre. In metropolitanenvironments the WDM system not only optimises the fibre use but moreimportantly gives the operator flexibility to respond to large customer transportrequirements. For example, a large bank may initially require a 150 Mbit/s tafficcapacity but some weeks later need to expand this to STM-4; or maybe they havea large non-SDH formatted bit rate to carry as fibre channels.
8SDH Backbone
(capacity full)
Internet
Video
Businesstraffic
EXTEND CAPACITY
STM-16WDM
STM-64 STM-64WDM
115 1168 High Capacity Networks
117 116 117 116High Capacity Networks High Capacity Networks
Synchronous Transmission Systems
High Capacity Networks
The rapid increase in traffic carried on today�s networks means that manyoperators are now close to exhausting the capacity of their installed fibre. Voice,leased line and mobile traffic are increasing, but the extremely rapid increase inbandwidth requirements is due to the faster growing, newer service types.Internet, ATM, data and video traffic are not only growing at a rapid rate but havelarge bandwidth requirements.
Operators are faced with the prospect of either the costly installation of newfibre or increasing the capacity of their existing fibre. This issue is more criticalfor new, emerging operators, who have less installed infrastructure or may needto lease dark fibres. The rapid growth of their traffic is forcing them to look atnew ways of expanding capacity on their existing fibre while keeping capitalcost to a minimum.
Technology has now advanced permitting the deployment of STM-64 systems,however, Wavelength Division Multiplexing (WDM) has also emerged as atechnique to increase the capacity of fibre. This involves transmitting multipleoptical signals down a single fibre. Each optical signal operates at a differentwavelength thus preventing interaction between them. For example severalSTM-16 or STM-64 traffic streams can be transmitted down one fibre. This notonly combats fibre exhaust issues, but also reduces the requirement for multipleregenerators on long spans. A single optical amplifier can be used to amplify thecombined signal hence replacing a regenerator mountain (see Figure 8-3).
Figure 8-1 Methods of Expanding Network Capacity
Networks are now being deployed with a purely optical layer, utilising thisWDM technique. This will permit the transport and routing of traffic withoutconversion into the electrical domain. To route, protect and manage traffic in anoptical network is currently constrained by technology and lack of standards,however, offering a flexible optical overlay layer is an area of great interest anddevelopment. ATM, IP and legacy data are already being carried over DenseWDM (D-WDM), bypassing the SDH layer. Optical Add/Drop multiplexers andsimple optical protection schemes are also now being deployed.
In the short term operators are facing the challenge of how to increase thecapacity of their networks quickly, however, they must also consider the longterm evolution of their networks. Long distance and metropolitan WDM opticalnetworks are designed to meet different requirements. Long distance operatorswant to minimise the cost per managed bit per fibre. In metropolitanenvironments the WDM system not only optimises the fibre use but moreimportantly gives the operator flexibility to respond to large customer transportrequirements. For example, a large bank may initially require a 150 Mbit/s tafficcapacity but some weeks later need to expand this to STM-4; or maybe they havea large non-SDH formatted bit rate to carry as fibre channels.
8SDH Backbone
(capacity full)
Internet
Video
Businesstraffic
EXTEND CAPACITY
STM-16WDM
STM-64 STM-64WDM
115 1168 High Capacity Networks
117 116 117 116High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
To achieve the capacity which is increasingly required in backbone networks,while retaining flexibility, WDM at STM-64 is being deployed as well as atSTM-16.
This chapter will give an overview of these technologies, typical applicationsand the evolution of networks.
Wavelength Division Multiplexing
Principles
WDM involves the transmission of several high speed channels on a single fibreor fibre pair. To prevent interaction of these signals they are transmitted onseparate wavelengths.
In Time Division Multiplexing (TDM) a single optical signal is used to transmitthe signal. Capacity is increased by running at a higher bit rate, however, whatcan be done to increase the capacity once the maximum rate has been reached?WDM offers the possibility of increasing the capacity of a fibre by allowingmultiple optical signals to travel simultaneously on a single fibre.
D-WDM is a term used to describe some WDM systems. The precise differencebetween these WDM and D-WDM varies, but as a simple guide; D-WDM refersto higher capacity WDM systems, that is greater than four wavelengths on onefibre. Today systems are in development which can support up to fortywavelengths on a single fibre.
To achieve WDM the separate wavelength channels are combined ormultiplexed together in an optical coupler, these signals are then opticallyamplified and transmitted over a single fibre, then at the far end of the link, thesignal is demultiplexed or split and sent to destination nodes such as SDHnetwork elements.
Combining the light in this way does introduce its own set of optical technologychallenges such as four wave mixing, polarisation mode and chromaticdispersion. Compensation techniques for these effects is discussed later in thischapter.
WDM Equipment
The next section will outline the equipment required for a point to point WDMlink. Further optical network elements are required for a flexible optical layerand these are described later in this chapter.
Optical Sources (�A� on Figure 8-2): The input to WDM systems can be avariety of signals from PDH 565 Mbit/s to SDH STM-64 systems. The mainrequirement is that in order for them to pass transparently through the WDMequipment, each must be at a different wavelength. Tolerance of thesewavelengths must also be tightly defined to prevent interaction.
This requirement can be met directly by the optical transmission cards in theSDH multiplexer source, provided the lasers adhere to the required wavelengthtolerances. This may be referred to as an integrated WDM solution, however,this is not the case with much of the installed equipment so additional devicesare required to convert the wavelength of the signal transmitted from someinstalled source transmitter cards. These devices are called transponders,wavelength translators or wavelength converters.
Wavelength Translator (B on Figure 8-2): Transponders can convert any inputwavelength to the wavelength required for input to the WDM coupler. The trafficon each incoming signal is received and then re-transmitted by a laser whichoperates at, and stabalised to, one of the pre-selected wavelengths.
This piece of equipment is particularly important where it is required to transmitsignals from older PDH equipment. Transponders are also being developed tosupport ATM and IP signals. For example 16 optical signals, some being PDH,some SDH and some ATM could all be converted to different wavelengths,combined and carried on the same fibre.
The disadvantages of using a transponder are the cost of the additional opticalinterfaces and a reduction in the overall network availability because moreelements are placed in the signal path. These additional elements also increasemaintenance requirements and require management integration. Consequentlytransponders should only be used where an appropriate integrated WDMsolution is not available.
117 1188 High Capacity Networks
119 118 119 118High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
To achieve the capacity which is increasingly required in backbone networks,while retaining flexibility, WDM at STM-64 is being deployed as well as atSTM-16.
This chapter will give an overview of these technologies, typical applicationsand the evolution of networks.
Wavelength Division Multiplexing
Principles
WDM involves the transmission of several high speed channels on a single fibreor fibre pair. To prevent interaction of these signals they are transmitted onseparate wavelengths.
In Time Division Multiplexing (TDM) a single optical signal is used to transmitthe signal. Capacity is increased by running at a higher bit rate, however, whatcan be done to increase the capacity once the maximum rate has been reached?WDM offers the possibility of increasing the capacity of a fibre by allowingmultiple optical signals to travel simultaneously on a single fibre.
D-WDM is a term used to describe some WDM systems. The precise differencebetween these WDM and D-WDM varies, but as a simple guide; D-WDM refersto higher capacity WDM systems, that is greater than four wavelengths on onefibre. Today systems are in development which can support up to fortywavelengths on a single fibre.
To achieve WDM the separate wavelength channels are combined ormultiplexed together in an optical coupler, these signals are then opticallyamplified and transmitted over a single fibre, then at the far end of the link, thesignal is demultiplexed or split and sent to destination nodes such as SDHnetwork elements.
Combining the light in this way does introduce its own set of optical technologychallenges such as four wave mixing, polarisation mode and chromaticdispersion. Compensation techniques for these effects is discussed later in thischapter.
WDM Equipment
The next section will outline the equipment required for a point to point WDMlink. Further optical network elements are required for a flexible optical layerand these are described later in this chapter.
Optical Sources (�A� on Figure 8-2): The input to WDM systems can be avariety of signals from PDH 565 Mbit/s to SDH STM-64 systems. The mainrequirement is that in order for them to pass transparently through the WDMequipment, each must be at a different wavelength. Tolerance of thesewavelengths must also be tightly defined to prevent interaction.
This requirement can be met directly by the optical transmission cards in theSDH multiplexer source, provided the lasers adhere to the required wavelengthtolerances. This may be referred to as an integrated WDM solution, however,this is not the case with much of the installed equipment so additional devicesare required to convert the wavelength of the signal transmitted from someinstalled source transmitter cards. These devices are called transponders,wavelength translators or wavelength converters.
Wavelength Translator (B on Figure 8-2): Transponders can convert any inputwavelength to the wavelength required for input to the WDM coupler. The trafficon each incoming signal is received and then re-transmitted by a laser whichoperates at, and stabalised to, one of the pre-selected wavelengths.
This piece of equipment is particularly important where it is required to transmitsignals from older PDH equipment. Transponders are also being developed tosupport ATM and IP signals. For example 16 optical signals, some being PDH,some SDH and some ATM could all be converted to different wavelengths,combined and carried on the same fibre.
The disadvantages of using a transponder are the cost of the additional opticalinterfaces and a reduction in the overall network availability because moreelements are placed in the signal path. These additional elements also increasemaintenance requirements and require management integration. Consequentlytransponders should only be used where an appropriate integrated WDMsolution is not available.
117 1188 High Capacity Networks
119 118 119 118High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 8-2 Point to Point WDM Link
Optical Mux/Demux (C & D on Figure 8-2): An optical multiplexer combinesthe discrete wavelengths emitted by the laser sources before transmission on thelink. At the far end the signals are separated by a demultiplexer or splitter. Anoptical multiplexer/demultiplexer (also called coupler) is a small passive opticalfilter with several input/output ports coupled to a single common port. A varietyof technological approaches are used to combine/separate the wavelengthsincluding discrete wavelength filters, fibre grating technology and planar arraywaveguide assemblies.
Optical Amplifier (E on Figure 8-2): Optical amplifiers form an integral part ofWDM networks, extending the span of each link. The distance the optical signalmay travel can be limited by attenuation. This effect can be addressed byamplification or regeneration. Regeneration involves converting each channel ofthe optical signal back to an electrical form before optical re-transmission of thesignal. In contrast, when amplified the signal remains in the optical domain andall wavelengths on one fibre can be simultaneously amplified by a single erbiumdoped amplifier. In this case a single optical amplifier can be deployed ratherthan regenerator equipment for each of the channels within the WDM signal.
Figure 8-3 Optical Amplifiers Reduce Need for Regenerators
SD
HM
ux
Ch
2
WD
MM
ux/D
emux
Cou
pler
/Spl
itter
Tx
Rx
Tx
Rx
Tx
Rx
Tx
Rx
SD
HM
ux
Ch
1
SD
HM
ux
Ch
n
Tx
Rx
Tx
Rx
Tx
Rx
Tx
Rx
Tx
Rx
SD
HM
ux
Ch
2
Tx
Rx
SD
HM
ux
Ch
1
SD
HM
ux
Ch
n
Tx
Rx
Tx
Rx
Opt
ical
Am
plifi
erO
ptic
alA
mpl
ifier
Tran
spon
der
Tran
spon
der
WD
MM
ux/D
emux
Cou
pler
/Spl
itter
A
B
CD
E
Note: Some operators use dual fibre uni-directional links, however the single fibrebi-directional link as shown in the diagram has a number of advantagesassociated with the minimisation of service deployment costs.
RegeneratorMountain
OpticalAmplifier
Tx
Tx
Tx
Tx
Rx
Rx
Rx
Rx
Tx
Tx
Tx
Tx
Rx
Rx
Rx
Rx
Tx
Tx
Tx
Tx
Rx
Rx
Rx
Rx
119 1208 High Capacity Networks
121 120 121 120High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 8-2 Point to Point WDM Link
Optical Mux/Demux (C & D on Figure 8-2): An optical multiplexer combinesthe discrete wavelengths emitted by the laser sources before transmission on thelink. At the far end the signals are separated by a demultiplexer or splitter. Anoptical multiplexer/demultiplexer (also called coupler) is a small passive opticalfilter with several input/output ports coupled to a single common port. A varietyof technological approaches are used to combine/separate the wavelengthsincluding discrete wavelength filters, fibre grating technology and planar arraywaveguide assemblies.
Optical Amplifier (E on Figure 8-2): Optical amplifiers form an integral part ofWDM networks, extending the span of each link. The distance the optical signalmay travel can be limited by attenuation. This effect can be addressed byamplification or regeneration. Regeneration involves converting each channel ofthe optical signal back to an electrical form before optical re-transmission of thesignal. In contrast, when amplified the signal remains in the optical domain andall wavelengths on one fibre can be simultaneously amplified by a single erbiumdoped amplifier. In this case a single optical amplifier can be deployed ratherthan regenerator equipment for each of the channels within the WDM signal.
Figure 8-3 Optical Amplifiers Reduce Need for Regenerators
SD
HM
ux
Ch
2
WD
MM
ux/D
emux
Cou
pler
/Spl
itter
Tx
Rx
Tx
Rx
Tx
Rx
Tx
Rx
SD
HM
ux
Ch
1
SD
HM
ux
Ch
n
Tx
Rx
Tx
Rx
Tx
Rx
Tx
Rx
Tx
Rx
SD
HM
ux
Ch
2
Tx
Rx
SD
HM
ux
Ch
1
SD
HM
ux
Ch
n
Tx
Rx
Tx
Rx
Opt
ical
Am
plifi
erO
ptic
alA
mpl
ifier
Tran
spon
der
Tran
spon
der
WD
MM
ux/D
emux
Cou
pler
/Spl
itter
A
B
CD
E
Note: Some operators use dual fibre uni-directional links, however the single fibrebi-directional link as shown in the diagram has a number of advantagesassociated with the minimisation of service deployment costs.
RegeneratorMountain
OpticalAmplifier
Tx
Tx
Tx
Tx
Rx
Rx
Rx
Rx
Tx
Tx
Tx
Tx
Rx
Rx
Rx
Rx
Tx
Tx
Tx
Tx
Rx
Rx
Rx
Rx
119 1208 High Capacity Networks
121 120 121 120High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
As the signal is transmitted it will suffer degradation due to noise and non-linearinteractions within the signal. Amplifiers are analogue devices and do not �cleanup� the degradation in the signal. This �clean up� can be achieved by terminatingthe signal at a regenerator site. Transmission up to very long distances ispossible using amplifiers, without regenerating the signal. In fact up to 600 kmscan be achieved at STM-16 in conventional landline systems with installed fibre,but for longer links regenerators are required. Note that this distance will varydepending on the bit rate and installed fibre.
In moderate length links a post or pre-amplifier may be located at the input to theoptical multiplexer or the output of the demultiplexer respectively. These areused to extend the reach as much as possible to reduce the need for in-lineamplifiers which require separate sites. Longer systems require line amplifiersto prevent the optical signal level being so weak that it cannot be adequatelyreceived.
Network management: The basic purpose of WDM network management isthe same as that for SDH as explained in Chapter 7. Most of the componentsrequired to further develop the optical layer of transmission networks arealready available today. The key factor that is currently limiting the rate at whichoptical layer functionality can be deployed is the complexity of the monitoringand management. The reason is that optical systems are analogue in nature. Forexample, if we look at performance management, then there is no opticalequivalent to the bit error checking that takes place in SDH equipment.
Communication between amplifiers and optical network management systemscan be achieved using an Optical Service Channel (OSC), which is a separateoptical channel outside the gain band of the amplifiers. The communicationprotocols over the optical service channels are of the same quality as theembedded communication channel found in SDH multiplexing equipment. Thismeans that the optical service channel and its terminations can support featuressuch as software download, the communication of station alarms (for exampleair conditioning failure) and data communications.
It is possible to design an optical service channel to be highly tolerant of faults.This ensures that the optical management system can be used for thecommunications required for fault isolation. This is achieved by using IS to ISrouting on the optical service channel so that communication paths can be re-established to equipment on the other side of fibre breaks. Under fault
conditions the amplifiers can also be designed to shut down in a mode thatensures that the optical service channel continues to operate.
Gain Profiles and the ITU Frequency Grid
The principal component of an optical amplifier is an Erbium Doped FibreAmplifier (EDFA) gain block, see Figure 8-4. The EDFA is composed of aspecial type of fibre optic cable manufactured with erbium. A dedicated pumplaser excites and energises the erbium in this fibre which amplifies the signal.Each amplifier has a characteristic gain profile which depends on thecharacteristics of the EDFA. The gain experienced by each band depends on itswavelength, the ideal profile is high and flat, that is all the wavelengths areapplied by a similar amount.
There is an optimum range of wavelengths for each amplifier. The ITU-T havestandardised a wavelength grid across the entire EDFA band, this grid(Recommendation G.692) is used to achieve compatibility between vendorsequipment.
Figure 8-4 Optical Amplifier
OpticalAmplifier
Doped Fibre
PumpLaser
OpticalIsolator
OpticalSignalInput
OpticalSignalOutput
121 1228 High Capacity Networks
123 122 123 122High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
As the signal is transmitted it will suffer degradation due to noise and non-linearinteractions within the signal. Amplifiers are analogue devices and do not �cleanup� the degradation in the signal. This �clean up� can be achieved by terminatingthe signal at a regenerator site. Transmission up to very long distances ispossible using amplifiers, without regenerating the signal. In fact up to 600 kmscan be achieved at STM-16 in conventional landline systems with installed fibre,but for longer links regenerators are required. Note that this distance will varydepending on the bit rate and installed fibre.
In moderate length links a post or pre-amplifier may be located at the input to theoptical multiplexer or the output of the demultiplexer respectively. These areused to extend the reach as much as possible to reduce the need for in-lineamplifiers which require separate sites. Longer systems require line amplifiersto prevent the optical signal level being so weak that it cannot be adequatelyreceived.
Network management: The basic purpose of WDM network management isthe same as that for SDH as explained in Chapter 7. Most of the componentsrequired to further develop the optical layer of transmission networks arealready available today. The key factor that is currently limiting the rate at whichoptical layer functionality can be deployed is the complexity of the monitoringand management. The reason is that optical systems are analogue in nature. Forexample, if we look at performance management, then there is no opticalequivalent to the bit error checking that takes place in SDH equipment.
Communication between amplifiers and optical network management systemscan be achieved using an Optical Service Channel (OSC), which is a separateoptical channel outside the gain band of the amplifiers. The communicationprotocols over the optical service channels are of the same quality as theembedded communication channel found in SDH multiplexing equipment. Thismeans that the optical service channel and its terminations can support featuressuch as software download, the communication of station alarms (for exampleair conditioning failure) and data communications.
It is possible to design an optical service channel to be highly tolerant of faults.This ensures that the optical management system can be used for thecommunications required for fault isolation. This is achieved by using IS to ISrouting on the optical service channel so that communication paths can be re-established to equipment on the other side of fibre breaks. Under fault
conditions the amplifiers can also be designed to shut down in a mode thatensures that the optical service channel continues to operate.
Gain Profiles and the ITU Frequency Grid
The principal component of an optical amplifier is an Erbium Doped FibreAmplifier (EDFA) gain block, see Figure 8-4. The EDFA is composed of aspecial type of fibre optic cable manufactured with erbium. A dedicated pumplaser excites and energises the erbium in this fibre which amplifies the signal.Each amplifier has a characteristic gain profile which depends on thecharacteristics of the EDFA. The gain experienced by each band depends on itswavelength, the ideal profile is high and flat, that is all the wavelengths areapplied by a similar amount.
There is an optimum range of wavelengths for each amplifier. The ITU-T havestandardised a wavelength grid across the entire EDFA band, this grid(Recommendation G.692) is used to achieve compatibility between vendorsequipment.
Figure 8-4 Optical Amplifier
OpticalAmplifier
Doped Fibre
PumpLaser
OpticalIsolator
OpticalSignalInput
OpticalSignalOutput
121 1228 High Capacity Networks
123 122 123 122High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Span Design
WDM network design is a new challenge facing network designers. There are anumber of factors which determine the spacing between amplifiers on a link, forexample the fibre type, type of signal transmitted and overall length of linkbetween regenerators. The example in Figure 8-5 illustrates how the overalllength of the link between terminal SDH network elements (or regenerators) canaffect the spacing of the intermediate optical amplifiers.
Figure 8-5 Optical Amplifier Spacing Examples
Bi-directional Transmission
On a uni-directional WDM link a single wavelength is used for each channel, buta separate fibre is required for transmission in the reverse direction (two fibresrequired, one for send and one for receive, see Figure 8-6). Bi-directionaltransmission involves using two different wavelengths for each channel, one foreach direction, but these wavelengths can all be supported on one fibre, that is
one fibre only is needed for send and receive. This means that only one fibre isneeded for transmission of the WDM signal and so particularly advantageousfor start up WDM operations.
In bi-directional transmission optical amplifiers amplify signals in both sendand receive direction, so only one amplifier is required at each site for each link.This can provide significant equipment savings over uni-directional operationwhere an amplifier is required for both send and receive fibres. Bi-directionaloperation is defined in ITU-T Recommendation G.692.
Figure 8-6 Bi-directional Optical Amplifier
Optical Signal Impairments
Optical signals can be corrupted by attenuations, back reflection and scatteringcaused by fibre breaks, micro bends and dirty connectors. These problems areinherent to the optical network and cannot be corrected by the D-WDM oroptical amplifier equipment. Other effects such as chromatic dispersion and
360 kms
120 kms 120 kms 120 kms
WDMTerminals
OpticalAmplifiers
WDMTerminals
80 kms 80 kms 80 kms 80 kms 80 kms 80 kms
560 kms
Unidirectional
Bi-directional
Tx Rx
TxRx
Tx Rx
TxRx
123 1248 High Capacity Networks
125 124 125 124High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Span Design
WDM network design is a new challenge facing network designers. There are anumber of factors which determine the spacing between amplifiers on a link, forexample the fibre type, type of signal transmitted and overall length of linkbetween regenerators. The example in Figure 8-5 illustrates how the overalllength of the link between terminal SDH network elements (or regenerators) canaffect the spacing of the intermediate optical amplifiers.
Figure 8-5 Optical Amplifier Spacing Examples
Bi-directional Transmission
On a uni-directional WDM link a single wavelength is used for each channel, buta separate fibre is required for transmission in the reverse direction (two fibresrequired, one for send and one for receive, see Figure 8-6). Bi-directionaltransmission involves using two different wavelengths for each channel, one foreach direction, but these wavelengths can all be supported on one fibre, that is
one fibre only is needed for send and receive. This means that only one fibre isneeded for transmission of the WDM signal and so particularly advantageousfor start up WDM operations.
In bi-directional transmission optical amplifiers amplify signals in both sendand receive direction, so only one amplifier is required at each site for each link.This can provide significant equipment savings over uni-directional operationwhere an amplifier is required for both send and receive fibres. Bi-directionaloperation is defined in ITU-T Recommendation G.692.
Figure 8-6 Bi-directional Optical Amplifier
Optical Signal Impairments
Optical signals can be corrupted by attenuations, back reflection and scatteringcaused by fibre breaks, micro bends and dirty connectors. These problems areinherent to the optical network and cannot be corrected by the D-WDM oroptical amplifier equipment. Other effects such as chromatic dispersion and
360 kms
120 kms 120 kms 120 kms
WDMTerminals
OpticalAmplifiers
WDMTerminals
80 kms 80 kms 80 kms 80 kms 80 kms 80 kms
560 kms
Unidirectional
Bi-directional
Tx Rx
TxRx
Tx Rx
TxRx
123 1248 High Capacity Networks
125 124 125 124High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Polarisation Mode Dispersion (PMD), however, must be considered whendesigning the optical layer.
Chromatic dispersion is the spreading of a light pulse when optical energytravels along the fibre. Several techniques exist to compensate for this type ofdispersion. For example �programmable chirp�, which involves reshaping theleading edge of the pulse in a positive or negative direction. Dispersioncompensation modules (or DCMs) can also be used to pre-distort the signal. ADCM consists of pre-coiled fibre in a box which is doped to pre-distort thesignal. New fibre types have also been introduced such as Non-zero DispersionShifted Fibre (NZDSF) which have a broad range of wavelengths that sufferminimum dispersion.
PMD is an inherent property of fibre which results from the spreading of a lightpulse. This effect depends on fibre type, the manufacturing conditions of thefibre and stresses placed on the cable. PMD must be taken into account in opticallink design.
Applications of WDM
Today WDM is commonly deployed to expand the capacity of backbone, longhaul links beyond a capacity of STM-16. WDM can present a cost effectivesolution as new fibre need not be laid. The current cost of the technology,however, can make WDM unattractive for other applications.
As the technology is developing other applications are emerging. Metropolitan,short haul links present an opportunity for WDM. Metropolitan networks differto long haul networks, in not just scale but also economics. On long haul linkstoday, fibre exhaust is a more critical problem. In metro networks it is harder tojustify the cost of an optical amplifier as the amount of traffic on a single fibreand the length of the link is likely to be lower.
Another development is the use of transponders to allow WDM systems toconnect directly to ATM switches and IP routers, as well as the transport oflower rate SDH and PDH traffic. In the years ahead IP traffic could be handledoptically, without using SDH multiplexers. Although some see the SDHoverhead as a tax on capacity, others value the overhead functionality inproviding fast fault isolation and quick problem resolution. SDH has a range ofprotection mechanisms to address network resilience.
STM-64
SDH systems are now being deployed which increase the line rate oftransmission by time division multiplexing to 10 Gbit/s or STM-64. Raising theline rate above STM-16 has previously been constrained by a number oftechnical issues, for example, chromatic dispersion and PMD. Methods toovercome and budget for these effects have been developed. These challengeshave now been overcome and there are many 10 Gbit/s systems in the field. Theneed for ever more capacity has driven the deployment of WDM on thesesystems, to achieve over 160 Gbit/s on a single fibre.
The inherent functionality of SDH network elements as described in theprevious chapters, such as cross-connect and standard protection schemes,applies equally to STM-64 systems. Lack of standards and componenttechnology constrains similar development in WDM; this is discussed further inthe following section as we look to the evolution of optical networks.
Moving Beyond 2.5 Gbit/s
So far this chapter has outlined technology to allow the capacity of a fibre to beraised above 2.5 Gbit/s, but what are the options for an operator and theadvantages and disadvantages of these? With WDM it is not enough to considerjust the link itself, but also:
� how the bandwidth is utilised,
� how the traffic is routed,
� the simplicity of the network,
� the ease with which it can be managed.
The advantages and disadvantages can be considered in terms of several criteria:
� the economics of each option,
� the existing network infrastructure,
� long term capacity needs.
125 1268 High Capacity Networks
127 126 127 126High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Polarisation Mode Dispersion (PMD), however, must be considered whendesigning the optical layer.
Chromatic dispersion is the spreading of a light pulse when optical energytravels along the fibre. Several techniques exist to compensate for this type ofdispersion. For example �programmable chirp�, which involves reshaping theleading edge of the pulse in a positive or negative direction. Dispersioncompensation modules (or DCMs) can also be used to pre-distort the signal. ADCM consists of pre-coiled fibre in a box which is doped to pre-distort thesignal. New fibre types have also been introduced such as Non-zero DispersionShifted Fibre (NZDSF) which have a broad range of wavelengths that sufferminimum dispersion.
PMD is an inherent property of fibre which results from the spreading of a lightpulse. This effect depends on fibre type, the manufacturing conditions of thefibre and stresses placed on the cable. PMD must be taken into account in opticallink design.
Applications of WDM
Today WDM is commonly deployed to expand the capacity of backbone, longhaul links beyond a capacity of STM-16. WDM can present a cost effectivesolution as new fibre need not be laid. The current cost of the technology,however, can make WDM unattractive for other applications.
As the technology is developing other applications are emerging. Metropolitan,short haul links present an opportunity for WDM. Metropolitan networks differto long haul networks, in not just scale but also economics. On long haul linkstoday, fibre exhaust is a more critical problem. In metro networks it is harder tojustify the cost of an optical amplifier as the amount of traffic on a single fibreand the length of the link is likely to be lower.
Another development is the use of transponders to allow WDM systems toconnect directly to ATM switches and IP routers, as well as the transport oflower rate SDH and PDH traffic. In the years ahead IP traffic could be handledoptically, without using SDH multiplexers. Although some see the SDHoverhead as a tax on capacity, others value the overhead functionality inproviding fast fault isolation and quick problem resolution. SDH has a range ofprotection mechanisms to address network resilience.
STM-64
SDH systems are now being deployed which increase the line rate oftransmission by time division multiplexing to 10 Gbit/s or STM-64. Raising theline rate above STM-16 has previously been constrained by a number oftechnical issues, for example, chromatic dispersion and PMD. Methods toovercome and budget for these effects have been developed. These challengeshave now been overcome and there are many 10 Gbit/s systems in the field. Theneed for ever more capacity has driven the deployment of WDM on thesesystems, to achieve over 160 Gbit/s on a single fibre.
The inherent functionality of SDH network elements as described in theprevious chapters, such as cross-connect and standard protection schemes,applies equally to STM-64 systems. Lack of standards and componenttechnology constrains similar development in WDM; this is discussed further inthe following section as we look to the evolution of optical networks.
Moving Beyond 2.5 Gbit/s
So far this chapter has outlined technology to allow the capacity of a fibre to beraised above 2.5 Gbit/s, but what are the options for an operator and theadvantages and disadvantages of these? With WDM it is not enough to considerjust the link itself, but also:
� how the bandwidth is utilised,
� how the traffic is routed,
� the simplicity of the network,
� the ease with which it can be managed.
The advantages and disadvantages can be considered in terms of several criteria:
� the economics of each option,
� the existing network infrastructure,
� long term capacity needs.
125 1268 High Capacity Networks
127 126 127 126High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Three ways to increase the capacity of a link four-fold are outlined below andevaluated in terms of these advantages and disadvantages.
Install more fibre
More links could be added to increase capacity. For example four fibre linkscould be deployed each with an STM-16 transmission system. This would beequivalent to the capacity of a 10 Gbit/s system.
� economics: expensive due the cost of installing new fibre and the costof additional repeaters
� existing network: would need to change network by installing newfibre and additional cross-connects may be required to groom thetraffic between the different fibres
� long term: no long term evolution path
Move to STM-64 systems with option to deploy WDM later
� economics: This would be a technically compact solution, using onlya single network element to increase capacity four-fold, appropriatewhen the network is growing rapidly.
� existing network: Being an SDH option, it would be a standardisedsolution offering integration with existing network management andoffering standard protection schemes. As the traffic on the networkincreases so the complexity of managing the bandwidth and routingtraffic increases. The inherent cross-connect capability of an SDHnetwork element, as described in Chapter 4, would allow the flexiblemanagement of channels across the whole of the 10 Gbit/s capacity,and removes the need to back-haul traffic to centralised cross-connects. Existing STM-16 terminals could be re-deployed or used astributary inputs for the STM-64 equipment.
� Long term: WDM deployment on STM-64 can be supported toincrease the capacity further, however, capacity is increased in stepsof 10 Gbit/s. An operator may prefer to increase the capacity insmaller steps.
Move to WDM on STM-16
� economics: A single optical amplifier can amplify several channelssimultaneously, so fewer repeaters may be required compared to thefirst option of installing more fibre. When traffic on the network isgrowing slowly this may be a more appropriate use of capitalcompared to STM-64.
� existing network: More network elements would be required than anintegrated STM-64 WDM solution. This would increasemaintenance, installation, power, spares and floor spacerequirements. WDM channels are entirely separate, so to movemanage and switch the traffic between channels it must be back-hauled to a cross-connect at the edge of the WDM link, where anoptical electrical conversion is required. This increase in distance thatthe signal must travel will increase the capacity requirementsimposed upon the WDM links. This factor and the additional cost ofthe cross-connect will raise the cost of the overall end to end link.
� evolution: Once a WDM solution and optical amplifiers are in placecapacity can be upgraded incrementally as required by the operator.
It is a common misconception that WDM will replace SDH or that to movebeyond 2.5 Gbit/s the operator must choose between WDM or STM-64. Thereality is that WDM and SDH are now being deployed together as integratedsystems. In the long term it is excepted that SDH will remain and that WDM andSDH will coexist, SDH being the input to the WDM systems. The standardprotection schemes, the bandwidth management capability and the networkmanagement systems of SDH, mean that SDH and WDM will form anintegrated solution for today�s transport networks.
127 1288 High Capacity Networks
129 128 129 128High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Three ways to increase the capacity of a link four-fold are outlined below andevaluated in terms of these advantages and disadvantages.
Install more fibre
More links could be added to increase capacity. For example four fibre linkscould be deployed each with an STM-16 transmission system. This would beequivalent to the capacity of a 10 Gbit/s system.
� economics: expensive due the cost of installing new fibre and the costof additional repeaters
� existing network: would need to change network by installing newfibre and additional cross-connects may be required to groom thetraffic between the different fibres
� long term: no long term evolution path
Move to STM-64 systems with option to deploy WDM later
� economics: This would be a technically compact solution, using onlya single network element to increase capacity four-fold, appropriatewhen the network is growing rapidly.
� existing network: Being an SDH option, it would be a standardisedsolution offering integration with existing network management andoffering standard protection schemes. As the traffic on the networkincreases so the complexity of managing the bandwidth and routingtraffic increases. The inherent cross-connect capability of an SDHnetwork element, as described in Chapter 4, would allow the flexiblemanagement of channels across the whole of the 10 Gbit/s capacity,and removes the need to back-haul traffic to centralised cross-connects. Existing STM-16 terminals could be re-deployed or used astributary inputs for the STM-64 equipment.
� Long term: WDM deployment on STM-64 can be supported toincrease the capacity further, however, capacity is increased in stepsof 10 Gbit/s. An operator may prefer to increase the capacity insmaller steps.
Move to WDM on STM-16
� economics: A single optical amplifier can amplify several channelssimultaneously, so fewer repeaters may be required compared to thefirst option of installing more fibre. When traffic on the network isgrowing slowly this may be a more appropriate use of capitalcompared to STM-64.
� existing network: More network elements would be required than anintegrated STM-64 WDM solution. This would increasemaintenance, installation, power, spares and floor spacerequirements. WDM channels are entirely separate, so to movemanage and switch the traffic between channels it must be back-hauled to a cross-connect at the edge of the WDM link, where anoptical electrical conversion is required. This increase in distance thatthe signal must travel will increase the capacity requirementsimposed upon the WDM links. This factor and the additional cost ofthe cross-connect will raise the cost of the overall end to end link.
� evolution: Once a WDM solution and optical amplifiers are in placecapacity can be upgraded incrementally as required by the operator.
It is a common misconception that WDM will replace SDH or that to movebeyond 2.5 Gbit/s the operator must choose between WDM or STM-64. Thereality is that WDM and SDH are now being deployed together as integratedsystems. In the long term it is excepted that SDH will remain and that WDM andSDH will coexist, SDH being the input to the WDM systems. The standardprotection schemes, the bandwidth management capability and the networkmanagement systems of SDH, mean that SDH and WDM will form anintegrated solution for today�s transport networks.
127 1288 High Capacity Networks
129 128 129 128High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 8-7 STM-64 Simplifies Bandwidth Management
Evolution to the Optical Network
In a typical network, it is likely that on certain links there is less fibre deployedor certain links carry a large amount of traffic. On these point to pointbottlenecks many operators are looking to WDM as a near term solution,increasing the capacity of the link without installing new fibre. What iseventually expected is that WDM will also be used to route traffic, rather thanjust increase capacity. This will create a new optical or photonic layer in thenetwork.
At its most basic, optical networking involves managing the bandwidth at theoptical layer of the network without converting the signal from optical toelectrical in order to add, drop or combine traffic. The optical traffic may also berouted around the network and protection may take place in the optical domainas well as in the underlying SDH domain. Also, these networks will be managedas an integrated part of the complete network.
STN-16 RingsVC-4 # 11
VC-4 # 4
VC-4 # 1-3
VC-4-4c
VC-4 # 4
VC-4 # 8
DXC
With large number of STM-16 channels,eventually cross-connects are needed
STM-16 D-WDM
STN-16 RingsVC-4 # 11
VC-4 # 4VC-4 # 1-3
VC-4-4c
VC-4 # 4
VC-4 # 8
STM-64 eliminates cross-connectscost through more channel capacity
STM-64 D-WDM
and built in Switcher Card
129 1308 High Capacity Networks
131 130 131 130High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 8-7 STM-64 Simplifies Bandwidth Management
Evolution to the Optical Network
In a typical network, it is likely that on certain links there is less fibre deployedor certain links carry a large amount of traffic. On these point to pointbottlenecks many operators are looking to WDM as a near term solution,increasing the capacity of the link without installing new fibre. What iseventually expected is that WDM will also be used to route traffic, rather thanjust increase capacity. This will create a new optical or photonic layer in thenetwork.
At its most basic, optical networking involves managing the bandwidth at theoptical layer of the network without converting the signal from optical toelectrical in order to add, drop or combine traffic. The optical traffic may also berouted around the network and protection may take place in the optical domainas well as in the underlying SDH domain. Also, these networks will be managedas an integrated part of the complete network.
STN-16 RingsVC-4 # 11
VC-4 # 4
VC-4 # 1-3
VC-4-4c
VC-4 # 4
VC-4 # 8
DXC
With large number of STM-16 channels,eventually cross-connects are needed
STM-16 D-WDM
STN-16 RingsVC-4 # 11
VC-4 # 4VC-4 # 1-3
VC-4-4c
VC-4 # 4
VC-4 # 8
STM-64 eliminates cross-connectscost through more channel capacity
STM-64 D-WDM
and built in Switcher Card
129 1308 High Capacity Networks
131 130 131 130High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 8-8 Network Evolution
Optical Networking Elements
With WDM it is possible to reduce the cost of the line between two end nodesbecause the cost of fibre and optical amplifiers can be shared between a largenumber of optical channels. Several channels can be transmitted on a single fibreand regenerators for several channels can be replaced by a single opticalamplifier. The real challenge of optical networks, however, will be how toreduce the cost of the node itself as well as the line and route traffic costeffectively.
The first generation of optical networking products are being developed. Theseinclude fixed Add/Drop multiplexers, small optical cross-connects andamplifiers with advanced management.
Figure 8-9 Optical Add/Drop Multiplexer
Exchanging wavelengths is a major challenge, that is determining how toexchange or convert a signal from one wavelength to another at a node. Forexample, a signal may enter the network at one wavelength and arrive in anothercity and find this wavelength occupied. Currently there is no cost effective wayto route one wavelength into another without having to go back to the electricallayer.
ATM
Term
inal
ATM
Term
inal
Inte
rnet
Term
inal
�A
DM
�A
DM
�A
DM
�X
CO
pti
cal
Lay
er
Tran
spor
tTe
rmin
al/A
DM
Tran
spor
tTe
rmin
al/A
DM
Tran
spor
tTe
rmin
al/A
DM
Tran
spo
rtL
ayer C
onve
ntio
nal
Ser
vice
s
Inte
rnet
Term
inal
Inte
rnet
AT
M
Wav
elen
gth
serv
ices
WD
M T
raff
icA
gg
reg
atio
n a
nd
Wav
elen
gth
Man
agem
ent
TD
M t
raff
icA
gg
reg
atio
n a
nd
Tran
spo
rtB
and
wid
thM
anag
emen
t
Ser
vice
Lay
ers
Op
tica
l Net
wo
rk B
uild
ing
Blo
cks
- W
DM
am
plifi
ers
- W
avel
engt
h m
ultip
lexe
rs-
Opt
ical
AD
Ms,
opt
ical
cro
ss-c
onne
cts
- W
DM
tran
smitt
ers/
tran
spon
ders
���
12
3
���
12
3
OpticalADM
131 1328 High Capacity Networks
133 132 133 132High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 8-8 Network Evolution
Optical Networking Elements
With WDM it is possible to reduce the cost of the line between two end nodesbecause the cost of fibre and optical amplifiers can be shared between a largenumber of optical channels. Several channels can be transmitted on a single fibreand regenerators for several channels can be replaced by a single opticalamplifier. The real challenge of optical networks, however, will be how toreduce the cost of the node itself as well as the line and route traffic costeffectively.
The first generation of optical networking products are being developed. Theseinclude fixed Add/Drop multiplexers, small optical cross-connects andamplifiers with advanced management.
Figure 8-9 Optical Add/Drop Multiplexer
Exchanging wavelengths is a major challenge, that is determining how toexchange or convert a signal from one wavelength to another at a node. Forexample, a signal may enter the network at one wavelength and arrive in anothercity and find this wavelength occupied. Currently there is no cost effective wayto route one wavelength into another without having to go back to the electricallayer.
ATM
Term
inal
ATM
Term
inal
Inte
rnet
Term
inal
�A
DM
�A
DM
�A
DM
�X
CO
pti
cal
Lay
er
Tran
spor
tTe
rmin
al/A
DM
Tran
spor
tTe
rmin
al/A
DM
Tran
spor
tTe
rmin
al/A
DM
Tran
spo
rtL
ayer C
onve
ntio
nal
Ser
vice
s
Inte
rnet
Term
inal
Inte
rnet
AT
M
Wav
elen
gth
serv
ices
WD
M T
raff
icA
gg
reg
atio
n a
nd
Wav
elen
gth
Man
agem
ent
TD
M t
raff
icA
gg
reg
atio
n a
nd
Tran
spo
rtB
and
wid
thM
anag
emen
t
Ser
vice
Lay
ers
Op
tica
l Net
wo
rk B
uild
ing
Blo
cks
- W
DM
am
plifi
ers
- W
avel
engt
h m
ultip
lexe
rs-
Opt
ical
AD
Ms,
opt
ical
cro
ss-c
onne
cts
- W
DM
tran
smitt
ers/
tran
spon
ders
���
12
3
���
12
3
OpticalADM
131 1328 High Capacity Networks
133 132 133 132High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Management
Managing signals as wavelengths has its own set of challenges which constrainthe implementation of these networks. Currently carriers have a highly managedSDH infrastructure that provide performance monitoring and survivability.Optical networks need to be managed in the same way and this will delay theirdeployment. As explained previously management control information cannotbe transported over the embedded channel in the signal itself, a special �opticalservice channel� is required.
Lack of standards: For photonic networks to become part of the transportnetwork, it will be required from a network level view point to manage allnetworks of whatever technology simultaneously. As yet there are no standardsdefining how an optical network will be managed. It has not been decided whatfunctions will be required of the management, how this will be implemented andhow it will integrate with the rest of the network.
Fault isolation: Fibre faults, optical network element faults, optical non-linearinteractions as a result of element degradation and connection faults, all need tobe isolated and fixed quickly. New comprehensive fault isolation facilities arerequired to rapidly isolate and repair these faults. With no electrical regeneratorsin the span, special optical techniques are required to isolate degraded or failedunits.
Some tools are currently available such as optical analogue monitors whichcheck the peak power and signal to noise ratio per wavelength and opticalreflectometers which monitor the connector optical return loss and signalreflections.
Performance Monitoring: To monitor the signal in the electrical layers of thenetwork, established systems check the bit rate error and when it crosses a pre-set threshold, some action is taken or an alarm raised. No such measure exists forchecking the health of a wavelength of light. Technology to enable completeperformance monitoring on optical channels without having to convert to theelectrical layer is yet to be developed.
Standards
Standards for the optical network are being discussed at the moment, however,these are at the very early stages. The lack of standards for the optical layer ofthe network will slow the implementation of systems.
Protection
In the event of equipment failure, or a fibre break, the SDH layer can quickly andefficiently, re-route traffic to a back-up path. In the situation where non-SDHtraffic is being carried, or where traffic originates from a platform with very highspeed interfaces (for example an IP router), then protection can be performed atthe optical layer itself. Protection in this case is achieved by switching the lightpaths round the fault, either on all the wavelengths within the fibre at once, or tooffer greater operational flexibility, on a per wavelength basis. By providing aseparate protection scheme on each wavelength the operator can increase theflexibility of the network by offering different services on each of thesewavelengths, each backed-up with an appropriate level of protection.
Implementation of the Optical Layer
The optimum degree of functionality that it is ultimately desirable to place in theoptical layer depends on the physical size of the network, the traffic volume andthe services being carried. This means that in the future we are likely to see avariety of network structures in which the distribution of the networkfunctionality between the optical, SDH and service layers is tailored to suit thetype of traffic carried, the transmission distance and the network structure.
Today most of the hardware required to implement networks with a largeroptical layer is unavailable, the key issues are the development of systems for themanagement and control of such networks. At present it is most cost effective touse network structures and a �thin� optical layer. Optical management isdeveloping however, and this will enable the growth of optical layerfunctionality for appropriate services in future networks.
133 1348 High Capacity Networks
135 134 135 134High Capacity Networks High Capacity Networks
Synchronous Transmission Systems Synchronous Transmission Systems
Management
Managing signals as wavelengths has its own set of challenges which constrainthe implementation of these networks. Currently carriers have a highly managedSDH infrastructure that provide performance monitoring and survivability.Optical networks need to be managed in the same way and this will delay theirdeployment. As explained previously management control information cannotbe transported over the embedded channel in the signal itself, a special �opticalservice channel� is required.
Lack of standards: For photonic networks to become part of the transportnetwork, it will be required from a network level view point to manage allnetworks of whatever technology simultaneously. As yet there are no standardsdefining how an optical network will be managed. It has not been decided whatfunctions will be required of the management, how this will be implemented andhow it will integrate with the rest of the network.
Fault isolation: Fibre faults, optical network element faults, optical non-linearinteractions as a result of element degradation and connection faults, all need tobe isolated and fixed quickly. New comprehensive fault isolation facilities arerequired to rapidly isolate and repair these faults. With no electrical regeneratorsin the span, special optical techniques are required to isolate degraded or failedunits.
Some tools are currently available such as optical analogue monitors whichcheck the peak power and signal to noise ratio per wavelength and opticalreflectometers which monitor the connector optical return loss and signalreflections.
Performance Monitoring: To monitor the signal in the electrical layers of thenetwork, established systems check the bit rate error and when it crosses a pre-set threshold, some action is taken or an alarm raised. No such measure exists forchecking the health of a wavelength of light. Technology to enable completeperformance monitoring on optical channels without having to convert to theelectrical layer is yet to be developed.
Standards
Standards for the optical network are being discussed at the moment, however,these are at the very early stages. The lack of standards for the optical layer ofthe network will slow the implementation of systems.
Protection
In the event of equipment failure, or a fibre break, the SDH layer can quickly andefficiently, re-route traffic to a back-up path. In the situation where non-SDHtraffic is being carried, or where traffic originates from a platform with very highspeed interfaces (for example an IP router), then protection can be performed atthe optical layer itself. Protection in this case is achieved by switching the lightpaths round the fault, either on all the wavelengths within the fibre at once, or tooffer greater operational flexibility, on a per wavelength basis. By providing aseparate protection scheme on each wavelength the operator can increase theflexibility of the network by offering different services on each of thesewavelengths, each backed-up with an appropriate level of protection.
Implementation of the Optical Layer
The optimum degree of functionality that it is ultimately desirable to place in theoptical layer depends on the physical size of the network, the traffic volume andthe services being carried. This means that in the future we are likely to see avariety of network structures in which the distribution of the networkfunctionality between the optical, SDH and service layers is tailored to suit thetype of traffic carried, the transmission distance and the network structure.
Today most of the hardware required to implement networks with a largeroptical layer is unavailable, the key issues are the development of systems for themanagement and control of such networks. At present it is most cost effective touse network structures and a �thin� optical layer. Optical management isdeveloping however, and this will enable the growth of optical layerfunctionality for appropriate services in future networks.
133 1348 High Capacity Networks
135 134 135 134Introduction Introduction
Synchronous Transmission Systems
SDH - A Detailed Description
This chapter is intended to expand on the overview of SDH included inChapter 3, �The Need For Synchronous Transmission�. Those who have notalready read that chapter may wish to do so before proceeding.
Synchronous Operation
We have seen that plesiochronous transmission systems allow tributaries todeviate from a predefined bit rate by only a set amount. Justification methodsthen bring all tributaries up to the same bit rate before multiplexing. Thejustification method used of �stuffing� extra bits in the data stream, however,makes it impossible to identify the location of specific tributary channels withina multiplexed data stream.
In synchronous transmission systems all elements of the system aresynchronised to the same master clock so no justification is required to bringtributaries up to a common rate before multiplexing.
The basic transmission rate defined in the SDH standards is 155.520 Mbit/s(STM-1). The STM-1 frame consists of 2,430 8-bit bytes which corresponds toa frame duration of 125 � s. Three higher bit rates are also defined:622.08 Mbit/s (STM-4), 2,488.32 Mbit/s (STM-16) and 9,953.28 Mbit/s(STM-64).
The STM-1 frame structure is referred to as being a 270 column (bytes) by ninerow structure, with the first nine columns of the structure constituting the�Section Overhead� area and the remaining 261 columns being the �Payload�area. The basic structure for an STM-N frame (where N can be 1, 4, 16 or 64, orany other rates that may be standardised) is shown in Figure 9-1.
The synchronous digital hierarchy eliminates the need for a number of the lowermultiplexing levels defined in the PDH. Tributaries of 2 Mbit/s are multiplexedto the STM-1 level in a single step. In order to achieve compatibility with non-synchronous equipment, however, the SDH recommendations define methodsof subdividing the payload area of an STM-1 frame in various ways so that it cancarry different combinations of tributaries, both synchronous and asynchronous.Using this method, synchronous transmission systems can accommodatesignals generated by equipment from various levels of the plesiochronous digitalhierarchy.
Figure 9-1 STM-N Frame Structure
9
1
2
3
4
5
6
7
8
9
9 x N
AU Pointer(s)
261 x N
270 x N columns (bytes)
STM-N Payload 9 Rows
(RSOH)
Section
Overhead
(MSOH)
Section
Overhead
135 1369 SDH – A Detailed Description
137 136 137 136Introduction Introduction
Synchronous Transmission Systems
SDH - A Detailed Description
This chapter is intended to expand on the overview of SDH included inChapter 3, �The Need For Synchronous Transmission�. Those who have notalready read that chapter may wish to do so before proceeding.
Synchronous Operation
We have seen that plesiochronous transmission systems allow tributaries todeviate from a predefined bit rate by only a set amount. Justification methodsthen bring all tributaries up to the same bit rate before multiplexing. Thejustification method used of �stuffing� extra bits in the data stream, however,makes it impossible to identify the location of specific tributary channels withina multiplexed data stream.
In synchronous transmission systems all elements of the system aresynchronised to the same master clock so no justification is required to bringtributaries up to a common rate before multiplexing.
The basic transmission rate defined in the SDH standards is 155.520 Mbit/s(STM-1). The STM-1 frame consists of 2,430 8-bit bytes which corresponds toa frame duration of 125 � s. Three higher bit rates are also defined:622.08 Mbit/s (STM-4), 2,488.32 Mbit/s (STM-16) and 9,953.28 Mbit/s(STM-64).
The STM-1 frame structure is referred to as being a 270 column (bytes) by ninerow structure, with the first nine columns of the structure constituting the�Section Overhead� area and the remaining 261 columns being the �Payload�area. The basic structure for an STM-N frame (where N can be 1, 4, 16 or 64, orany other rates that may be standardised) is shown in Figure 9-1.
The synchronous digital hierarchy eliminates the need for a number of the lowermultiplexing levels defined in the PDH. Tributaries of 2 Mbit/s are multiplexedto the STM-1 level in a single step. In order to achieve compatibility with non-synchronous equipment, however, the SDH recommendations define methodsof subdividing the payload area of an STM-1 frame in various ways so that it cancarry different combinations of tributaries, both synchronous and asynchronous.Using this method, synchronous transmission systems can accommodatesignals generated by equipment from various levels of the plesiochronous digitalhierarchy.
Figure 9-1 STM-N Frame Structure
9
1
2
3
4
5
6
7
8
9
9 x N
AU Pointer(s)
261 x N
270 x N columns (bytes)
STM-N Payload 9 Rows
(RSOH)
Section
Overhead
(MSOH)
Section
Overhead
135 1369 SDH – A Detailed Description
137 136 137 136Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Differences Between SDH and SONET
The ITU-T SDH recommendations can be viewed as the worldwide standardsfor synchronous transmission. Within these standards, however, there is someroom for manoeuvre when implementing a system. As a result, the ETSIimplementation of SDH, used in Europe and much of the rest of the world,differs in some details from the North American implementation. This isbecause the North American implementation is developed from the originalANSI SONET standards. Thus the best way of viewing the differences betweenSONET and SDH as implemented elsewhere, is to consider SONET as a subsetof the worldwide SDH standard.
The first level of the SONET hierarchy is referred to as STS-1 (for an electricalsignal) or OC-1 (for an optical signal) and corresponds to a bit rate of51.84 Mbit/s. (STS = Synchronous Transport Signal, OC = Optical Carrier). Inmuch the same way as an STM-1 frame can be considered as a 270 column(byte) by nine row structure, an STS-1 frame can be viewed as 90 column (byte)by nine rows. The section overhead (SOH) constitutes the first three columns ofthe STS-1 frame.
The following gives levels from the SONET hierarchy, with the SDHequivalents:
When the basic STS-1 signal is multiplexed to STS-3 it becomes identical inframe rate and format to STM-1. There are, however, slight differences inpointer processing. Figure 9-2 shows the comparison of SDH and SONETsignals at the STM-1/STS-3 and STM-4/STS-12 levels.
Figure 9-2 Comparison of SONET and SDH signals
The STM-1 Frame
An STM-1 frame consists of 2,430 bytes which can be divided into three mainareas:
� Payload Area (2,349 bytes)
� AU Pointer Area (9 bytes)
� Section Overhead Area (72 bytes)
OC-1/STS-1 51.84 Mbit/s STM-0 or STM-1/3
OC-3/STS-3 155.52 Mbit/s STM-1
OC-12/STS-12 622.08 Mbit/s STM-4
OC-48/STS-48 2488.32 Mbit/s STM-16
OC-192/STS-192 9953.28 Mbit/s STM-64
STM-4
STS-12
STM-1 STS-3
STS-1
x 3
x 4
155.52 Mbit/s
51.84 Mbit/s
Existing North American Signals
Existing North American
3 x 3 column SOH9 column SOH
4 x 9 column (SDH SOH)
and European Signals
12 x 3 column (SONET SOH)
3 column SOH
137 1389 SDH – A Detailed Description
139 138 139 138Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Differences Between SDH and SONET
The ITU-T SDH recommendations can be viewed as the worldwide standardsfor synchronous transmission. Within these standards, however, there is someroom for manoeuvre when implementing a system. As a result, the ETSIimplementation of SDH, used in Europe and much of the rest of the world,differs in some details from the North American implementation. This isbecause the North American implementation is developed from the originalANSI SONET standards. Thus the best way of viewing the differences betweenSONET and SDH as implemented elsewhere, is to consider SONET as a subsetof the worldwide SDH standard.
The first level of the SONET hierarchy is referred to as STS-1 (for an electricalsignal) or OC-1 (for an optical signal) and corresponds to a bit rate of51.84 Mbit/s. (STS = Synchronous Transport Signal, OC = Optical Carrier). Inmuch the same way as an STM-1 frame can be considered as a 270 column(byte) by nine row structure, an STS-1 frame can be viewed as 90 column (byte)by nine rows. The section overhead (SOH) constitutes the first three columns ofthe STS-1 frame.
The following gives levels from the SONET hierarchy, with the SDHequivalents:
When the basic STS-1 signal is multiplexed to STS-3 it becomes identical inframe rate and format to STM-1. There are, however, slight differences inpointer processing. Figure 9-2 shows the comparison of SDH and SONETsignals at the STM-1/STS-3 and STM-4/STS-12 levels.
Figure 9-2 Comparison of SONET and SDH signals
The STM-1 Frame
An STM-1 frame consists of 2,430 bytes which can be divided into three mainareas:
� Payload Area (2,349 bytes)
� AU Pointer Area (9 bytes)
� Section Overhead Area (72 bytes)
OC-1/STS-1 51.84 Mbit/s STM-0 or STM-1/3
OC-3/STS-3 155.52 Mbit/s STM-1
OC-12/STS-12 622.08 Mbit/s STM-4
OC-48/STS-48 2488.32 Mbit/s STM-16
OC-192/STS-192 9953.28 Mbit/s STM-64
STM-4
STS-12
STM-1 STS-3
STS-1
x 3
x 4
155.52 Mbit/s
51.84 Mbit/s
Existing North American Signals
Existing North American
3 x 3 column SOH9 column SOH
4 x 9 column (SDH SOH)
and European Signals
12 x 3 column (SONET SOH)
3 column SOH
137 1389 SDH – A Detailed Description
139 138 139 138Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Payload
Signals from all levels of the PDH can be accommodated in the SDH bypackaging them together in the payload area of an STM-1 frame. The process ofpackaging PDH signals is a multi-step process involving a number of differentstructures.
The plesiochronous tributaries are mapped into an appropriately sized container,and a number of bytes, known as the path overhead (POH), are added to form therelevant virtual container (VC). The path overhead provides information for usein end-to-end management of a synchronous path. Further information followson the two types of path overhead, one type associated with VC-2/VC-1 and theother with VC-4/VC-3.
Figure 9-3 VC-2/VC-1 Path overhead V5 byte
VC-2/VC-1 Path Overhead
The bytes V5, J2, Z6 and Z7 are allocated to the VC-2/VC-1 POH. The V5 byteis the byte positioned at the start of the virtual container and functions of thevarious bits in this byte are described below. The V5 byte is shown in Figure 9-3:
VC-4/VC-3 Path Overhead
The VC-4 path overhead is located in the first column of the nine row by 261column VC-4 structure. For the VC-3, the path overhead is located in the firstcolumn of the nine row by 85 column structure. The nine byte path overhead forboth the VC-4 and VC-3 structure is shown in Figure 9-4. The function of eachbyte, with some explanatory text, is given in the following table.
Figure 9-4 VC-4/VC-3 Path Overhead
BIP-2: Bits 1 and 2 are used for error performance monitoring using a Bit Interleaved Parity (BIP) check on all the bytes in the previous VC-2/VC-1.
REI: Bit 3 is a path Remote Error Indication (REI) that is set to binary one and sent back towards the originating end of a VC-2/VC-1 if one or more errors are detected by the BIP-2 check.
RFI: Bit 4 is a path Remote Failure Indication (RFI) that is set to binary one and sent back by the VC-2/VC-1 assembler if a failure is declared.
BIP-2 REI RFI Signal Label RDI
1 2 3 4 5 6 7 8
Signal Label:
Indicates the type of virtual container payload. Codings can be, �path unequipped�, �asynchronous mapping�, �byte synchronous mapping�, �bit synchronous mapping�, �path equipped - non-specific�, or �path equipped - to be defined�.
RDI: Bit 8 is the path Remote Defect Indicator (RDI) that is set to binary one and sent back by the VC-2/VC-1 assembler if either a TU-2/TU-1 path AIS or a signal failure condition is being received.
J1
B3
C2
G1
F2
H4
Z3
K3
Z5
VC-4 or VC-3Payload
139 1409 SDH – A Detailed Description
141 140 141 140Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Payload
Signals from all levels of the PDH can be accommodated in the SDH bypackaging them together in the payload area of an STM-1 frame. The process ofpackaging PDH signals is a multi-step process involving a number of differentstructures.
The plesiochronous tributaries are mapped into an appropriately sized container,and a number of bytes, known as the path overhead (POH), are added to form therelevant virtual container (VC). The path overhead provides information for usein end-to-end management of a synchronous path. Further information followson the two types of path overhead, one type associated with VC-2/VC-1 and theother with VC-4/VC-3.
Figure 9-3 VC-2/VC-1 Path overhead V5 byte
VC-2/VC-1 Path Overhead
The bytes V5, J2, Z6 and Z7 are allocated to the VC-2/VC-1 POH. The V5 byteis the byte positioned at the start of the virtual container and functions of thevarious bits in this byte are described below. The V5 byte is shown in Figure 9-3:
VC-4/VC-3 Path Overhead
The VC-4 path overhead is located in the first column of the nine row by 261column VC-4 structure. For the VC-3, the path overhead is located in the firstcolumn of the nine row by 85 column structure. The nine byte path overhead forboth the VC-4 and VC-3 structure is shown in Figure 9-4. The function of eachbyte, with some explanatory text, is given in the following table.
Figure 9-4 VC-4/VC-3 Path Overhead
BIP-2: Bits 1 and 2 are used for error performance monitoring using a Bit Interleaved Parity (BIP) check on all the bytes in the previous VC-2/VC-1.
REI: Bit 3 is a path Remote Error Indication (REI) that is set to binary one and sent back towards the originating end of a VC-2/VC-1 if one or more errors are detected by the BIP-2 check.
RFI: Bit 4 is a path Remote Failure Indication (RFI) that is set to binary one and sent back by the VC-2/VC-1 assembler if a failure is declared.
BIP-2 REI RFI Signal Label RDI
1 2 3 4 5 6 7 8
Signal Label:
Indicates the type of virtual container payload. Codings can be, �path unequipped�, �asynchronous mapping�, �byte synchronous mapping�, �bit synchronous mapping�, �path equipped - non-specific�, or �path equipped - to be defined�.
RDI: Bit 8 is the path Remote Defect Indicator (RDI) that is set to binary one and sent back by the VC-2/VC-1 assembler if either a TU-2/TU-1 path AIS or a signal failure condition is being received.
J1
B3
C2
G1
F2
H4
Z3
K3
Z5
VC-4 or VC-3Payload
139 1409 SDH – A Detailed Description
141 140 141 140Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
AU Pointer
After the path overhead is added the virtual container is positioned in a TributaryUnit (TU) or an Administrative Unit (AU) with a pointer indicating the start ofthe virtual container relative to the TU or AU, as appropriate. VC-1s and VC-2sare positioned in TUs; whereas VC-4s are always positioned in an AU-4, seeFigure 9-5. In Europe, VC-3s are positioned in a TU-3 but for the SONEToption are positioned in an AU-3. TUs and AUs are each bundled into theirrespective groups; Tributary Unit Groups (TUGs) for TUs, and AdministrativeUnit Groups (AUGs) for AUs. TUGs are multiplexed into higher order virtual
containers, which in turn are positioned in AUs with a pointer indicating the startof the virtual container relative to the AU. This is the AU pointer which indicatesthe position of the AU in relation to the STM-1 frame and forms part of thesection overhead area of the frame.
Figure 9-5 SDH Multiplexing Structure
The payload area of the STM-1 frame contains a VC-4 or three VC-3s with theposition of the first byte being indicated by the respective AU pointer.
The use of pointers in the STM-1 frame structure means that plesiochronoussignals can be accommodated within the synchronous network without the useof buffers. This is because the signal can be packaged into a virtual container andinserted into the frame at any position and the pointer indicates this position. Useof the pointer method was made possible by defining synchronous virtual
J1 Path Trace This byte verifies the VC-4/VC-3 path connection.
B3 Path BIP-8 This byte provides bit error monitoring over the path using an even bit parity code, BIP-8
C2 Signal Label This byte indicates the composition of the VC-4/VC-3 payload.
G1 Path Status This byte allows the status of the received signal to be returned to the transmitting end of the path from the receiving end.
F2, Z3 Path User Channels
This byte provides a user communication channel.
H4 Position Indicator
This byte provides a generalised position indicator for payloads and can be used as a multiframe position indicator for VC-2/VC-1.
K3(b1-b4)
APS These bits are allocated for Automatic Protection Switching (APS) signalling for protection at the higher order path level.
K3(b5-b8)
Spare These bits are allocated for future use.
Z5 National Operator
This byte is allocated for specific management purposes, such as tandem connection maintenance.
STM-N AUG AU-4x N x 1
x 3 TUG-3
AU-3 VC-3
x 3
TU-3 VC-3
TUG-2 TU-2 VC-2
TU-12
TU-11
VC-12
VC-11
C-3
C-2
C-12
C-11
C-4
x 1
x 7
x 7
x 1
x 3
x 4
139264
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
44736
34368
6312
2048
1544
pointer processing
multiplexing
aligning
mapping
SONET specific options
ETSI specific options
VC-4
141 1429 SDH – A Detailed Description
143 142 143 142Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
AU Pointer
After the path overhead is added the virtual container is positioned in a TributaryUnit (TU) or an Administrative Unit (AU) with a pointer indicating the start ofthe virtual container relative to the TU or AU, as appropriate. VC-1s and VC-2sare positioned in TUs; whereas VC-4s are always positioned in an AU-4, seeFigure 9-5. In Europe, VC-3s are positioned in a TU-3 but for the SONEToption are positioned in an AU-3. TUs and AUs are each bundled into theirrespective groups; Tributary Unit Groups (TUGs) for TUs, and AdministrativeUnit Groups (AUGs) for AUs. TUGs are multiplexed into higher order virtual
containers, which in turn are positioned in AUs with a pointer indicating the startof the virtual container relative to the AU. This is the AU pointer which indicatesthe position of the AU in relation to the STM-1 frame and forms part of thesection overhead area of the frame.
Figure 9-5 SDH Multiplexing Structure
The payload area of the STM-1 frame contains a VC-4 or three VC-3s with theposition of the first byte being indicated by the respective AU pointer.
The use of pointers in the STM-1 frame structure means that plesiochronoussignals can be accommodated within the synchronous network without the useof buffers. This is because the signal can be packaged into a virtual container andinserted into the frame at any position and the pointer indicates this position. Useof the pointer method was made possible by defining synchronous virtual
J1 Path Trace This byte verifies the VC-4/VC-3 path connection.
B3 Path BIP-8 This byte provides bit error monitoring over the path using an even bit parity code, BIP-8
C2 Signal Label This byte indicates the composition of the VC-4/VC-3 payload.
G1 Path Status This byte allows the status of the received signal to be returned to the transmitting end of the path from the receiving end.
F2, Z3 Path User Channels
This byte provides a user communication channel.
H4 Position Indicator
This byte provides a generalised position indicator for payloads and can be used as a multiframe position indicator for VC-2/VC-1.
K3(b1-b4)
APS These bits are allocated for Automatic Protection Switching (APS) signalling for protection at the higher order path level.
K3(b5-b8)
Spare These bits are allocated for future use.
Z5 National Operator
This byte is allocated for specific management purposes, such as tandem connection maintenance.
STM-N AUG AU-4x N x 1
x 3 TUG-3
AU-3 VC-3
x 3
TU-3 VC-3
TUG-2 TU-2 VC-2
TU-12
TU-11
VC-12
VC-11
C-3
C-2
C-12
C-11
C-4
x 1
x 7
x 7
x 1
x 3
x 4
139264
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
kbit/s
44736
34368
6312
2048
1544
pointer processing
multiplexing
aligning
mapping
SONET specific options
ETSI specific options
VC-4
141 1429 SDH – A Detailed Description
143 142 143 142Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
containers slightly larger than the payload they carry. This allows the payload toslip in time relative to the STM-1 frame in which it is contained.
Adjustment of the pointers is also possible where slight changes of frequencyand phase occur as a result of variations in propagation delay.
The result of this is that in any data stream it is possible to identify individualtributary channels, and drop or insert information, thus overcoming one of themain drawbacks of the PDH.
Figure 9-6 STM-1 Section Overhead
Section Overhead
The Section Overhead bytes are used for communication between adjacentpieces of synchronous equipment. As well as being used for framesynchronisation, they perform a variety of management and administrationfacilities. The purpose of individual bytes is detailed below. The STM-1 SectionOverhead is shown in Figure 9-6:
European Multiplexing Structure
The SDH Multiplexing Structure, as defined in ITU-T Recommendation G.707,offers a number of options in the way in which each type of container can bemultiplexed into an STM-N frame. In order to establish a common multiplexingroute for Europe, ETSI established a preferred European multiplexing routewhich is shown in Figure 9-7. (ITU-T Recommendation G.707 (03/96) is amerged and revised version of Recommendations G.707, G.708 and G.709.)
A1A1A1 A2 A2 A2 J0
B1
D1
E1
D2
F1
D3
B2 B2 B2 K1 K2
D4 D5 D6
D7 D8 D9
D10 D11 D12
S1 Z1 Z1 Z2 Z2 M1 E2
Bytes reserved for national use
RegeneratorSectionOverhead
MultiplexSectionOverhead
* *
*
(RSOH)
(MSOH)
AU Pointer(s)
9 bytes
9 ro
ws
Unscrambled bytes, therefore care should be taken
Media dependent byteswith their content
A1, A2 Framing.
J0 Regenerator section trace.
D1-D12 Bytes D1, D2 and D3 form a 192 kbit/s data communication channel for the regenerator section. Bytes D4 to D12 form a 576 kbit/s data communication channel for the multiplex section. The use of both data communication channels is for network management.
E1, E2 Orderwire channels.
F1 User channel.
B1, B2 These bytes are simple parity checks for error detection.
K1, K2 (b1-b5) Automatic Protection Switching (APS) channel.
K2 (b6-b8) Multiplex section RDI.
S1 (b5-b8) Synchronisation status.
M1 Multiplex section REI.
Z1, Z2 To be defined.
143 1449 SDH – A Detailed Description
145 144 145 144Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
containers slightly larger than the payload they carry. This allows the payload toslip in time relative to the STM-1 frame in which it is contained.
Adjustment of the pointers is also possible where slight changes of frequencyand phase occur as a result of variations in propagation delay.
The result of this is that in any data stream it is possible to identify individualtributary channels, and drop or insert information, thus overcoming one of themain drawbacks of the PDH.
Figure 9-6 STM-1 Section Overhead
Section Overhead
The Section Overhead bytes are used for communication between adjacentpieces of synchronous equipment. As well as being used for framesynchronisation, they perform a variety of management and administrationfacilities. The purpose of individual bytes is detailed below. The STM-1 SectionOverhead is shown in Figure 9-6:
European Multiplexing Structure
The SDH Multiplexing Structure, as defined in ITU-T Recommendation G.707,offers a number of options in the way in which each type of container can bemultiplexed into an STM-N frame. In order to establish a common multiplexingroute for Europe, ETSI established a preferred European multiplexing routewhich is shown in Figure 9-7. (ITU-T Recommendation G.707 (03/96) is amerged and revised version of Recommendations G.707, G.708 and G.709.)
A1A1A1 A2 A2 A2 J0
B1
D1
E1
D2
F1
D3
B2 B2 B2 K1 K2
D4 D5 D6
D7 D8 D9
D10 D11 D12
S1 Z1 Z1 Z2 Z2 M1 E2
Bytes reserved for national use
RegeneratorSectionOverhead
MultiplexSectionOverhead
* *
*
(RSOH)
(MSOH)
AU Pointer(s)
9 bytes
9 ro
ws
Unscrambled bytes, therefore care should be taken
Media dependent byteswith their content
A1, A2 Framing.
J0 Regenerator section trace.
D1-D12 Bytes D1, D2 and D3 form a 192 kbit/s data communication channel for the regenerator section. Bytes D4 to D12 form a 576 kbit/s data communication channel for the multiplex section. The use of both data communication channels is for network management.
E1, E2 Orderwire channels.
F1 User channel.
B1, B2 These bytes are simple parity checks for error detection.
K1, K2 (b1-b5) Automatic Protection Switching (APS) channel.
K2 (b6-b8) Multiplex section RDI.
S1 (b5-b8) Synchronisation status.
M1 Multiplex section REI.
Z1, Z2 To be defined.
143 1449 SDH – A Detailed Description
145 144 145 144Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 9-7 ETSI European Multiplexing Route
More detail is given in Figure 9-8, showing how PDH 2 Mbit/s signals aremultiplexed into STM-N frames.
The individual processes in Figure 9-8 are explained in more detail in thefollowing sections where PDH 2 Mbit/s signals are multiplexed into an STM-1frame.
Figure 9-8 Multiplexing method for European primary rate services
STM-N AUG AU-4x N x 1
TUG-3
x 3
TU-3 VC-3
TUG-2 TU-12 VC-12
VC-11
C-3
C-12
C-11
C-4
x 1
x 7
x 3
139264
kbit/s
kbit/s
kbit/s
kbit/s
34368
2048
1544
pointer processing
multiplexing
aligning
mapping
VC-4
C-12
TUG-3TUG-3VC-4 POH
TUG-2
AU-4 Ptr
AUG SOH
TU-12 Ptr
TU-12 Ptr
TU-12 Ptr
AU-4 Ptr
TUG-2
VC-12
TU-12
TUG-2
TUG-3
VC-4
AU-4
AUG
STM-N
Logical association
Physical association
Note: Shaded areas are phase aligned. Phase alignment between theshaded and unshaded areas is defined by the pointer and is indicatedby the arrow.
C-12
VC-12
VC-12VC-12
VC-4
VC-4
AUG
VC-12 POH
145 1469 SDH – A Detailed Description
147 146 147 146Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 9-7 ETSI European Multiplexing Route
More detail is given in Figure 9-8, showing how PDH 2 Mbit/s signals aremultiplexed into STM-N frames.
The individual processes in Figure 9-8 are explained in more detail in thefollowing sections where PDH 2 Mbit/s signals are multiplexed into an STM-1frame.
Figure 9-8 Multiplexing method for European primary rate services
STM-N AUG AU-4x N x 1
TUG-3
x 3
TU-3 VC-3
TUG-2 TU-12 VC-12
VC-11
C-3
C-12
C-11
C-4
x 1
x 7
x 3
139264
kbit/s
kbit/s
kbit/s
kbit/s
34368
2048
1544
pointer processing
multiplexing
aligning
mapping
VC-4
C-12
TUG-3TUG-3VC-4 POH
TUG-2
AU-4 Ptr
AUG SOH
TU-12 Ptr
TU-12 Ptr
TU-12 Ptr
AU-4 Ptr
TUG-2
VC-12
TU-12
TUG-2
TUG-3
VC-4
AU-4
AUG
STM-N
Logical association
Physical association
Note: Shaded areas are phase aligned. Phase alignment between theshaded and unshaded areas is defined by the pointer and is indicatedby the arrow.
C-12
VC-12
VC-12VC-12
VC-4
VC-4
AUG
VC-12 POH
145 1469 SDH – A Detailed Description
147 146 147 146Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Mapping of a 2 Mbit/s Signal into a VC-12
SDH offers two options for the mapping of a 2 Mbit/s signal into a VC-12:
� Asynchronous 2,048 kbit/s: Allows for carriage of a 2 Mbit/spayload, but without the ability to observe individual bits.
� Byte synchronous 2,048 kbit/s: Allows observation andidentification of all bits within the payload.
For asynchronous operation, the method used is based on a number ofopportunities within the signal to justify the 2,048 kbit/s data (see Figure 9-9).This justification allows for the variations between the 2048 kbit/s clock and theclock providing the timing for the synchronous network, however, it does meanthat a degree of processing is required to accommodate this justification. Theadditional fixed stuff bits and bytes shown, are included to maintain a definedsize for the VC-12 of 140 bytes (for the 500 ms multiframe). The V5 byte is theoverhead byte which carries information pertaining to the VC-12 end-to-endpath. It is this byte which provides the end-to-end path information forindividual virtual containers.
If a byte synchronous 2,048 kbit/s mapping is utilised, the 64 kbit/s channelswithin the 2 Mbit/s signal have fixed mappings within the VC-12, (seeFigure 9-10). This has great advantages if access is required to a particular64 kbit/s channel within a VC-12, for example at a point of cross-connection. Itdoes, however, introduce additional processing to provide the fixed mappinginto the VC-12 and this introduces some delay during cross-connection whilereprocessing of the pointers occurs, (125 � s for the 2 Mbit/s signal and 250 � sfor the 64 kbit/s signal).
.
Figure 9-9 Asynchronous mapping of a 2_Mbit/s signal into a VC-12
140Bytes
V5
R
32 Bytes
R
J2
C1 C2 0 0 0 0 R R
32 Bytes
R
Z6
C1 C2 0 0 0 0 R R
32 Bytes
R
K4
C1 C2 R R R R R S1
S2 I I I I I I I
31 Bytes
R
500 � s
IOCSR
- Information bit- Overhead bit- Justification control bit- Justification opportunity bit- Fixed stuff bit
147 1489 SDH – A Detailed Description
149 148 149 148Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Mapping of a 2 Mbit/s Signal into a VC-12
SDH offers two options for the mapping of a 2 Mbit/s signal into a VC-12:
� Asynchronous 2,048 kbit/s: Allows for carriage of a 2 Mbit/spayload, but without the ability to observe individual bits.
� Byte synchronous 2,048 kbit/s: Allows observation andidentification of all bits within the payload.
For asynchronous operation, the method used is based on a number ofopportunities within the signal to justify the 2,048 kbit/s data (see Figure 9-9).This justification allows for the variations between the 2048 kbit/s clock and theclock providing the timing for the synchronous network, however, it does meanthat a degree of processing is required to accommodate this justification. Theadditional fixed stuff bits and bytes shown, are included to maintain a definedsize for the VC-12 of 140 bytes (for the 500 ms multiframe). The V5 byte is theoverhead byte which carries information pertaining to the VC-12 end-to-endpath. It is this byte which provides the end-to-end path information forindividual virtual containers.
If a byte synchronous 2,048 kbit/s mapping is utilised, the 64 kbit/s channelswithin the 2 Mbit/s signal have fixed mappings within the VC-12, (seeFigure 9-10). This has great advantages if access is required to a particular64 kbit/s channel within a VC-12, for example at a point of cross-connection. Itdoes, however, introduce additional processing to provide the fixed mappinginto the VC-12 and this introduces some delay during cross-connection whilereprocessing of the pointers occurs, (125 � s for the 2 Mbit/s signal and 250 � sfor the 64 kbit/s signal).
.
Figure 9-9 Asynchronous mapping of a 2_Mbit/s signal into a VC-12
140Bytes
V5
R
32 Bytes
R
J2
C1 C2 0 0 0 0 R R
32 Bytes
R
Z6
C1 C2 0 0 0 0 R R
32 Bytes
R
K4
C1 C2 R R R R R S1
S2 I I I I I I I
31 Bytes
R
500 � s
IOCSR
- Information bit- Overhead bit- Justification control bit- Justification opportunity bit- Fixed stuff bit
147 1489 SDH – A Detailed Description
149 148 149 148Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 9-10 Byte Synchronous mapping of a 2 Mbit/s signal into a VC-12
Multiplexing of three VC-12s into a TUG-2
Figure 9-11 shows pictorially the relationship between VC-12s, TU-12s,TUG-2s and a TUG-3. The VC-12s are positioned into TU-12s with the TU-12pointer. Three TU-12s are byte interleaved into a TUG-2, and seven TUG-2s arebyte interleaved into a TUG-3.
Figure 9-11 Multiplexing of VC-12s into a TUG-2 via TU-12s
The TU-12 pointer indicates the starting point of a VC-12 within a TU-12. If thetiming of a VC-12 causes it to slip in position with regard to the TU-12, then theTU-12 pointer can be adjusted to indicate the new position. The location of thepointer is fixed within the TUG-2 regardless of the position of the virtualcontainer as shown in Figure 9-11.
500 � s
R* - May be used for timeslot 0 if requiredR = fixed stuff
140Bytes
V5RR*
Channels 1-15
Channel 16Channels 16-30
RJ2
R*R
Channels 1-15Channel 16
Channels 16-30
RZ6
R*R
Channels 1-15Channel 16
Channels 16-30R
K4
R*R
Channels 1-15
Channel 16Channels 16-30
R
NOTE: Mapping for 31 channels with common channel signalling
86 Columns
TUG-2 TUG-2
TU-12Pointers
TU-12Pointers
Fix
ed S
tuff POH
POH
TUG-3(7 x TUG-2)
VC-12
3 VC-12s in 1 x TUG-2
Fix
ed S
tuff
NOTE: A VC-12 plus TU-12 pointer forms a TU-12 structure, three of which, make up a TUG-2.
149 1509 SDH – A Detailed Description
151 150 151 150Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 9-10 Byte Synchronous mapping of a 2 Mbit/s signal into a VC-12
Multiplexing of three VC-12s into a TUG-2
Figure 9-11 shows pictorially the relationship between VC-12s, TU-12s,TUG-2s and a TUG-3. The VC-12s are positioned into TU-12s with the TU-12pointer. Three TU-12s are byte interleaved into a TUG-2, and seven TUG-2s arebyte interleaved into a TUG-3.
Figure 9-11 Multiplexing of VC-12s into a TUG-2 via TU-12s
The TU-12 pointer indicates the starting point of a VC-12 within a TU-12. If thetiming of a VC-12 causes it to slip in position with regard to the TU-12, then theTU-12 pointer can be adjusted to indicate the new position. The location of thepointer is fixed within the TUG-2 regardless of the position of the virtualcontainer as shown in Figure 9-11.
500 � s
R* - May be used for timeslot 0 if requiredR = fixed stuff
140Bytes
V5RR*
Channels 1-15
Channel 16Channels 16-30
RJ2
R*R
Channels 1-15Channel 16
Channels 16-30
RZ6
R*R
Channels 1-15Channel 16
Channels 16-30R
K4
R*R
Channels 1-15
Channel 16Channels 16-30
R
NOTE: Mapping for 31 channels with common channel signalling
86 Columns
TUG-2 TUG-2
TU-12Pointers
TU-12Pointers
Fix
ed S
tuff POH
POH
TUG-3(7 x TUG-2)
VC-12
3 VC-12s in 1 x TUG-2
Fix
ed S
tuff
NOTE: A VC-12 plus TU-12 pointer forms a TU-12 structure, three of which, make up a TUG-2.
149 1509 SDH – A Detailed Description
151 150 151 150Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Multiplexing of TUG-2s into a TUG-3
The multiplexing of seven TUG-2s into a TUG-3 is a fixed mapping as shownin Figure 9-12.
Figure 9-12 Multiplexing of seven TUG-2s into a TUG-3
Multiplexing of TUG-3s into a VC-4
Figure 9-13 shows the multiplexing of three TUG-3s into a VC-4. As in the caseof TUG-2 to TUG-3, multiplexing there is a fixed mapping with each TUG-3having a fixed position with regard to the VC-4. Column one of the VC-4contains the VC-4 path overhead and columns two and three contain fixed stuffbytes.
Figure 9-13 Multiplexing of three TUG-3s into a VC-4
Multiplexing of a VC-4 into an STM-1 Signal
The mapping of a VC-4 into the STM-1 signal follows a similar process to thatof the VC-12 into a TUG-2. The VC-4 is positioned in an AU-4 with an AU-4pointer indicating the position of the starting point of the VC-4 structure. TheAU-4 pointer provides a method of allowing flexible and dynamic alignment ofthe VC-4 within the AU-4. The AU-4 pointer bytes are shown in Figure 9-14. Itshould be noted that the AU-4 structure is identical with that of an AUG whereasa group of three AU-3s are needed to construct an AUG.
Bytes H1 and H2 indicate the location of the starting position for the VC-4payload, and where necessary byte H3 is used for justification of the VC-4.
TU-11 TU-12 TU-2
TUG-2
(1) (2) (3) (7)
67
12
34
56
7
12
34
56
7
12
12
34
56
7
12
34
56
7
12
34
56
7
12
34
56
7
12
34
56
7
12
34
12
34
12
34
12
3
12
3
12
3
12
3
1 2 3 4 5 6 7 8 9 20 30 40... 78 8682
FixedStuff
TUG-3
(A)
TUG-3
(B)
TUG-3
(C)1 1 1 868686
A BC A BC AA B CA BC
1 2 3 4 5 6 7 8 910 261
Fixed Stuff
VC-4 POH
151 1529 SDH – A Detailed Description
153 152 153 152Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Multiplexing of TUG-2s into a TUG-3
The multiplexing of seven TUG-2s into a TUG-3 is a fixed mapping as shownin Figure 9-12.
Figure 9-12 Multiplexing of seven TUG-2s into a TUG-3
Multiplexing of TUG-3s into a VC-4
Figure 9-13 shows the multiplexing of three TUG-3s into a VC-4. As in the caseof TUG-2 to TUG-3, multiplexing there is a fixed mapping with each TUG-3having a fixed position with regard to the VC-4. Column one of the VC-4contains the VC-4 path overhead and columns two and three contain fixed stuffbytes.
Figure 9-13 Multiplexing of three TUG-3s into a VC-4
Multiplexing of a VC-4 into an STM-1 Signal
The mapping of a VC-4 into the STM-1 signal follows a similar process to thatof the VC-12 into a TUG-2. The VC-4 is positioned in an AU-4 with an AU-4pointer indicating the position of the starting point of the VC-4 structure. TheAU-4 pointer provides a method of allowing flexible and dynamic alignment ofthe VC-4 within the AU-4. The AU-4 pointer bytes are shown in Figure 9-14. Itshould be noted that the AU-4 structure is identical with that of an AUG whereasa group of three AU-3s are needed to construct an AUG.
Bytes H1 and H2 indicate the location of the starting position for the VC-4payload, and where necessary byte H3 is used for justification of the VC-4.
TU-11 TU-12 TU-2
TUG-2
(1) (2) (3) (7)
67
12
34
56
7
12
34
56
7
12
12
34
56
7
12
34
56
7
12
34
56
7
12
34
56
7
12
34
56
7
12
34
12
34
12
34
12
3
12
3
12
3
12
3
1 2 3 4 5 6 7 8 9 20 30 40... 78 8682
FixedStuff
TUG-3
(A)
TUG-3
(B)
TUG-3
(C)1 1 1 868686
A BC A BC AA B CA BC
1 2 3 4 5 6 7 8 910 261
Fixed Stuff
VC-4 POH
151 1529 SDH – A Detailed Description
153 152 153 152Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 9-14 Multiplexing of a VC-4 into an STM-1
Concatenation
In order to support other services which do not have a container defined toaccept them, G.70X contains provision for concatenation of AU-4s and TU-2swithin an STM-N frame.
AU-4 Concatenation
AU-4s can be concatenated together to form an AU-4-Xc which can transportpayloads requiring greater capacity than one C-4. A concatenation indication inthe AU-4 pointer, is used to show that multi C-4 payloads, carried in a singleVC-4-Xc, are to be kept together. The capacity available for the mapping of themulti C-4, is X times the capacity of the C-4. For example, 599.040 Mbit/s forX=4 and 2,396.160 Mbit/s for X=16. The VC-4-Xc is illustrated in Figure 9-15.
Figure 9-15 VC-4-Xc Structure
TU-2 Concatenation
A number of TU-2s can be combined to produce a C-2 which is m times the sizeof the TU-2. An example of this is the concatenation of five TU-2s to carry a32 Mbit/s video signal. Using this method, four video signals can be carried ina VC-4, compared with only three had the video signal been mapped into astandard container. Thus a useful increase in efficiency is achieved.
J1B3
C2G1
F2H4
Z3
K3Z5
H1 YY H2 1*1* H3
H3
H3
H1 H2
H3
AU-4 PTR
SOH
SOH
VC-4POH
261
2619
3
1
5
JustificationOpportunity
1* = All 1s byteY = 1001SS11 (S bits are unspecified)
VC-4
J1B3C2G1F2H4Z3K3Z5
AU-n Pointers
RSOH
MSOH
X x 261 bytes
N x 261 bytesN x 9
3
1
5
1 X-1 X x 260
VC-4-Xc
STM-N
N x 270 bytes
C-4-XcStuffFixed
153 1549 SDH – A Detailed Description
155 154 155 154Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 9-14 Multiplexing of a VC-4 into an STM-1
Concatenation
In order to support other services which do not have a container defined toaccept them, G.70X contains provision for concatenation of AU-4s and TU-2swithin an STM-N frame.
AU-4 Concatenation
AU-4s can be concatenated together to form an AU-4-Xc which can transportpayloads requiring greater capacity than one C-4. A concatenation indication inthe AU-4 pointer, is used to show that multi C-4 payloads, carried in a singleVC-4-Xc, are to be kept together. The capacity available for the mapping of themulti C-4, is X times the capacity of the C-4. For example, 599.040 Mbit/s forX=4 and 2,396.160 Mbit/s for X=16. The VC-4-Xc is illustrated in Figure 9-15.
Figure 9-15 VC-4-Xc Structure
TU-2 Concatenation
A number of TU-2s can be combined to produce a C-2 which is m times the sizeof the TU-2. An example of this is the concatenation of five TU-2s to carry a32 Mbit/s video signal. Using this method, four video signals can be carried ina VC-4, compared with only three had the video signal been mapped into astandard container. Thus a useful increase in efficiency is achieved.
J1B3
C2G1
F2H4
Z3
K3Z5
H1 YY H2 1*1* H3
H3
H3
H1 H2
H3
AU-4 PTR
SOH
SOH
VC-4POH
261
2619
3
1
5
JustificationOpportunity
1* = All 1s byteY = 1001SS11 (S bits are unspecified)
VC-4
J1B3C2G1F2H4Z3K3Z5
AU-n Pointers
RSOH
MSOH
X x 261 bytes
N x 261 bytesN x 9
3
1
5
1 X-1 X x 260
VC-4-Xc
STM-N
N x 270 bytes
C-4-XcStuffFixed
153 1549 SDH – A Detailed Description
155 154 155 154Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
The concatenated C-2s are carried in a single VC-2-mc (concatenatedm x VC-2). Recommendation G.70X details three methods for concatenation ofTU-2s.
Concatenation of contiguous TU-2s in the higher order VC-3
TU-2s, which are contiguous in time with one another and must remain in thesame sequential order, are combined to form a TU-2-mc (m indicates thenumber of TU-2s concatenated). With this type of concatenation the VC-2-mccontains a single virtual container path overhead which appears in the first VC-2of the VC-2-mc.
This method of concatenation uses the concatenation indication �1001� withinbits 1 and 4 of the TU-2 pointer, with bits 5 and 6 unspecified and all-ones in bits7 to 16.
The capacity of the concatenated signal could in fact be larger than the m x TU-2capacity since the path overhead for the signal only occurs within the first VC-2of the VC-2-mc.
Sequential concatenation of TU-2s in the higher order VC-4
This type of concatenation allows the simultaneous transport of TU-2-mcs andTU-3s in the same VC-4. The exact method of achieving this is under study.
Virtual concatenation of TU-2s in the higher order VC-4
This method of concatenation is used for TU-2s which do not have to occupyadjacent locations (that is, are not contiguous) within the VC-4.
This method places the concatenation function solely within the operation of thepath terminating equipment and does not use the concatenation indicationwithin the TU-2 pointer bytes. The rest of the network, must still recognise thesevirtual containers carrying the concatenated load in order to retain the TU-2sequence.
Since each TU-2 contains its own virtual container path overhead the totalcapacity of the TU-2-mc is restricted to the payload capacity of the TU-2, unlikethe situation in contiguous concatenation where there is no path overheadassociated with the second and subsequent virtual containers of the TU-2-mc.
Higher Transmission Rates
Since the signals to be multiplexed at the STM level are all synchronous,multiplexing is achieved by simple byte interleaving. Some definition is neededas to the exact method used so that the signal structure at all levels remains thesame. For example, consider an STM-1 signal to be multiplexed to STM-16.Two possible routes are possible, STM-1 to STM-4 to STM-16 or direct STM-1 to STM-16. To maintain the same signal structure, ITU-T RecommendationG.70X state that to form an STM-M signal STM-N signals must be N byteinterleaved. For example, STM-1 multiplexing is achieved by interleaving onebyte from each tributary. For STM-4 multiplexing each 4 byte group is taken andinterleaved with that from the next tributary. Figure 9-16 shows how an STM-16signal would be made up from STM-1 and STM-4 tributaries.
Figure 9-16 Multiplexing of STM-1 signals
0123456789ABCDEF
STM-1 STM-4 STM-16
4
4
4
4
4
3 0 1 2 3 0
7 4 5 6 7 4 5
AB B8 89 9
CC D E FF
0 1 2 3 4 5 6 7 8 9 A B C D E F
0 1 2 3 4 5 6 7 8 9 A B C D E F16
155 1569 SDH – A Detailed Description
157 156 157 156Introduction Introduction
Synchronous Transmission Systems Synchronous Transmission Systems
The concatenated C-2s are carried in a single VC-2-mc (concatenatedm x VC-2). Recommendation G.70X details three methods for concatenation ofTU-2s.
Concatenation of contiguous TU-2s in the higher order VC-3
TU-2s, which are contiguous in time with one another and must remain in thesame sequential order, are combined to form a TU-2-mc (m indicates thenumber of TU-2s concatenated). With this type of concatenation the VC-2-mccontains a single virtual container path overhead which appears in the first VC-2of the VC-2-mc.
This method of concatenation uses the concatenation indication �1001� withinbits 1 and 4 of the TU-2 pointer, with bits 5 and 6 unspecified and all-ones in bits7 to 16.
The capacity of the concatenated signal could in fact be larger than the m x TU-2capacity since the path overhead for the signal only occurs within the first VC-2of the VC-2-mc.
Sequential concatenation of TU-2s in the higher order VC-4
This type of concatenation allows the simultaneous transport of TU-2-mcs andTU-3s in the same VC-4. The exact method of achieving this is under study.
Virtual concatenation of TU-2s in the higher order VC-4
This method of concatenation is used for TU-2s which do not have to occupyadjacent locations (that is, are not contiguous) within the VC-4.
This method places the concatenation function solely within the operation of thepath terminating equipment and does not use the concatenation indicationwithin the TU-2 pointer bytes. The rest of the network, must still recognise thesevirtual containers carrying the concatenated load in order to retain the TU-2sequence.
Since each TU-2 contains its own virtual container path overhead the totalcapacity of the TU-2-mc is restricted to the payload capacity of the TU-2, unlikethe situation in contiguous concatenation where there is no path overheadassociated with the second and subsequent virtual containers of the TU-2-mc.
Higher Transmission Rates
Since the signals to be multiplexed at the STM level are all synchronous,multiplexing is achieved by simple byte interleaving. Some definition is neededas to the exact method used so that the signal structure at all levels remains thesame. For example, consider an STM-1 signal to be multiplexed to STM-16.Two possible routes are possible, STM-1 to STM-4 to STM-16 or direct STM-1 to STM-16. To maintain the same signal structure, ITU-T RecommendationG.70X state that to form an STM-M signal STM-N signals must be N byteinterleaved. For example, STM-1 multiplexing is achieved by interleaving onebyte from each tributary. For STM-4 multiplexing each 4 byte group is taken andinterleaved with that from the next tributary. Figure 9-16 shows how an STM-16signal would be made up from STM-1 and STM-4 tributaries.
Figure 9-16 Multiplexing of STM-1 signals
0123456789ABCDEF
STM-1 STM-4 STM-16
4
4
4
4
4
3 0 1 2 3 0
7 4 5 6 7 4 5
AB B8 89 9
CC D E FF
0 1 2 3 4 5 6 7 8 9 A B C D E F
0 1 2 3 4 5 6 7 8 9 A B C D E F16
155 1569 SDH – A Detailed Description
157 156 157 156Synchronisation Synchronisation
Synchronous Transmission Systems
Synchronisation
All digital networks require timing and synchronisation. At its lowest level,synchronisation is required so that digital equipment knows when to send andreceive data; without it data becomes errored or unusable. Despite theimportance of timing, many operators give it only scant consideration, yet it candirectly impact revenue, operational costs and equally important, the operator�sreputation by delivering poor quality of service.
Timing and synchronisation are not new concepts, however, they have becomemore frequently used terms with the advent of new services and the increasedpenetration of digital technology. The changes in the telecommunicationstransmission infrastructure from a PDH network to an SDH network has alsoincreased the need for good quality synchronisation.
There are many changes taking place at the same time and because of thissynchronisation tends to be ignored or drops down the priority list, which canlead to penalties on service.
In larger networks synchronisation can be very complex, however, many of theproblems that are associated with poor synchronisation can be avoided byfollowing good design practice.
Synchronisation is as essential as power plant and is similar in that it does notdirectly generate any revenue for the operator. End users will be unaware ofgood quality synchronisation but they will notice the affects of poor
synchronisation. Poor quality synchronisation has a direct impact on the qualityof service delivered to customers with consequent loss of revenue.
Implementing a robust synchronisation network need not cost a fortune.Depending upon the types of service that are provided, synchronisation costsbetween 1 and 5% of the SDH equipment value.
The Voice Legacy
It is easy to see why some operators are sceptical about the need for specialconsideration of synchronisation.
"The network ís running just fine as it is".
Historically this was perfectly true. Voice has dominated telecommunicationsnetworks as the main traffic type, and unless there was something seriouslywrong, nobody noticed the errors caused by poor synchronisation.
Poor synchronisation causes a very short duration click during a voiceconversation. In a very badly synchronised network, the click might occur everyfew seconds. Many users have grown up with the era of analogue telephonenetworks, where clicks from mechanical Strowger telephone exchanges andcrossed lines were common-place. Compared to noisy analogue alternativeseven poorly synchronised digital networks can sound very clear.
Although synchronisation within one operator�s network may be adequate andprovide very high quality voice calls, problems can arise when a call is made intoanother operator�s network. Historically this meant international dialling whereusers tend to expect the call to be of poor quality. This is often reflected bypeople�s surprise when they say "You sound as if you are just down the road".Synchronisation is not totally responsible for poor international voice quality,but it can be an important consideration, especially on international digitalcircuits.
As deregulation sweeps the world, dialling into another network operator doesnot mean dialling overseas. Many countries now have multiple operators andindeed a single telephone call within the same country could involve three orfour different operators. The user needs to be totally oblivious to the fact thatmultiple operators are used to deliver the call. Synchronisation across multipleoperators therefore needs careful consideration.
10
157 15810 Synchronisation
159 158 159 158Synchronisation Synchronisation
Synchronous Transmission Systems
Synchronisation
All digital networks require timing and synchronisation. At its lowest level,synchronisation is required so that digital equipment knows when to send andreceive data; without it data becomes errored or unusable. Despite theimportance of timing, many operators give it only scant consideration, yet it candirectly impact revenue, operational costs and equally important, the operator�sreputation by delivering poor quality of service.
Timing and synchronisation are not new concepts, however, they have becomemore frequently used terms with the advent of new services and the increasedpenetration of digital technology. The changes in the telecommunicationstransmission infrastructure from a PDH network to an SDH network has alsoincreased the need for good quality synchronisation.
There are many changes taking place at the same time and because of thissynchronisation tends to be ignored or drops down the priority list, which canlead to penalties on service.
In larger networks synchronisation can be very complex, however, many of theproblems that are associated with poor synchronisation can be avoided byfollowing good design practice.
Synchronisation is as essential as power plant and is similar in that it does notdirectly generate any revenue for the operator. End users will be unaware ofgood quality synchronisation but they will notice the affects of poor
synchronisation. Poor quality synchronisation has a direct impact on the qualityof service delivered to customers with consequent loss of revenue.
Implementing a robust synchronisation network need not cost a fortune.Depending upon the types of service that are provided, synchronisation costsbetween 1 and 5% of the SDH equipment value.
The Voice Legacy
It is easy to see why some operators are sceptical about the need for specialconsideration of synchronisation.
"The network ís running just fine as it is".
Historically this was perfectly true. Voice has dominated telecommunicationsnetworks as the main traffic type, and unless there was something seriouslywrong, nobody noticed the errors caused by poor synchronisation.
Poor synchronisation causes a very short duration click during a voiceconversation. In a very badly synchronised network, the click might occur everyfew seconds. Many users have grown up with the era of analogue telephonenetworks, where clicks from mechanical Strowger telephone exchanges andcrossed lines were common-place. Compared to noisy analogue alternativeseven poorly synchronised digital networks can sound very clear.
Although synchronisation within one operator�s network may be adequate andprovide very high quality voice calls, problems can arise when a call is made intoanother operator�s network. Historically this meant international dialling whereusers tend to expect the call to be of poor quality. This is often reflected bypeople�s surprise when they say "You sound as if you are just down the road".Synchronisation is not totally responsible for poor international voice quality,but it can be an important consideration, especially on international digitalcircuits.
As deregulation sweeps the world, dialling into another network operator doesnot mean dialling overseas. Many countries now have multiple operators andindeed a single telephone call within the same country could involve three orfour different operators. The user needs to be totally oblivious to the fact thatmultiple operators are used to deliver the call. Synchronisation across multipleoperators therefore needs careful consideration.
10
157 15810 Synchronisation
159 158 159 158Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
The Digital Revolution
Telecommunication networks no longer carry just voice. The range of dataoriented services is growing rapidly and data adoption is growing at a staggeringpace. Facsimile machines and dial-up Internet access exploit the voiceinfrastructure but because they are data services, they need specialconsideration. Data can be transmitted in a variety forms. Operators are oftenunaware of the extent of data transmission which can even include services suchas burglar and fire alarm systems.
Many operators want to move up the value chain and providing data orientedservices is an obvious way to achieve this. As business users continue to demandhigh capacity data services such as LAN interconnects, Frame Relay, ATM orsimple leased line circuits, the need for good quality synchronisation increases.
Service Sensitivity to Synchronisation
Not all services demand good synchronisation. It has already been shown thatvoice is extremely tolerant of poor synchronisation, but the same cannot be saidof all services and in particular, data oriented services. Table 10-2 characterisesservices and how they are affected by the main symptom of poorsynchronisation.
Modem Oriented Services
Table 10-2 clearly shows that even simple services that we take for granted, suchas facsimile and dial-up Internet access, can be affected by synchronisation slips(errors caused by poor synchronisation). Slip errors usually mean that a block ofdata is lost or it is repeated.
Although a slip sounds like a short duration click to a human, to a facsimilemachine that click can result in the loss or duplication of data. Group 1 and 2type facsimile machines are particularly sensitive to errors. Most modernfacsimile machines are Group 3 or 4 which have a much higher error tolerance,however, they are not totally immune to slips.
Slips occur because the transmitting equipment is sending data at a differentspeed to that the receiver. It is practically impossible to ensure that both ends aretotally aligned to the same speed. Buffers are used to introduce some slack intothe system. Adding buffers does not eliminate the difference in speed, it meansthat the slip (an overlap in timing reference) is deferred to a later point in time.A one byte buffer results in a one byte slip. A 100 byte buffer means that slipsare reduced in frequency, but when they do occur they are 100 bytes in duration.
All digital equipment has buffers, and buffers can be found at various levels.Each level can therefore introduce a slip. The facsimile machine can introducea slip and the transmission equipment between the facsimile machines can alsoslip.
Table 10-2 Service Sensitivity to Synchronisation
Service Type Impact on Service
Voice Low impact. Occasional click
PABX Corruption of inter-exchange signalling PABX may lock-up.
Facsimile Distorted or missing facsimile lines
Modem traffic,(for example, Internet)
Loss of data or dropped calls
Cellular phone dropped calls, poor cell hand-over, poor voice quality, Cell base station may lock-up - possible visit needed.
Data(for example, leased-line)
lost, repeated or corrupted information
Digital video conferencing Distorted or frozen picture. Worse for modern systems
Digital broadcast video Picture subject to colour burst through to a distorted or frozen picture.
Encrypted data (for exam-ple, financial services)
Possible loss of data through to possible attacks on the security of the data.
159 16010 Synchronisation
161 160 161 160Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
The Digital Revolution
Telecommunication networks no longer carry just voice. The range of dataoriented services is growing rapidly and data adoption is growing at a staggeringpace. Facsimile machines and dial-up Internet access exploit the voiceinfrastructure but because they are data services, they need specialconsideration. Data can be transmitted in a variety forms. Operators are oftenunaware of the extent of data transmission which can even include services suchas burglar and fire alarm systems.
Many operators want to move up the value chain and providing data orientedservices is an obvious way to achieve this. As business users continue to demandhigh capacity data services such as LAN interconnects, Frame Relay, ATM orsimple leased line circuits, the need for good quality synchronisation increases.
Service Sensitivity to Synchronisation
Not all services demand good synchronisation. It has already been shown thatvoice is extremely tolerant of poor synchronisation, but the same cannot be saidof all services and in particular, data oriented services. Table 10-2 characterisesservices and how they are affected by the main symptom of poorsynchronisation.
Modem Oriented Services
Table 10-2 clearly shows that even simple services that we take for granted, suchas facsimile and dial-up Internet access, can be affected by synchronisation slips(errors caused by poor synchronisation). Slip errors usually mean that a block ofdata is lost or it is repeated.
Although a slip sounds like a short duration click to a human, to a facsimilemachine that click can result in the loss or duplication of data. Group 1 and 2type facsimile machines are particularly sensitive to errors. Most modernfacsimile machines are Group 3 or 4 which have a much higher error tolerance,however, they are not totally immune to slips.
Slips occur because the transmitting equipment is sending data at a differentspeed to that the receiver. It is practically impossible to ensure that both ends aretotally aligned to the same speed. Buffers are used to introduce some slack intothe system. Adding buffers does not eliminate the difference in speed, it meansthat the slip (an overlap in timing reference) is deferred to a later point in time.A one byte buffer results in a one byte slip. A 100 byte buffer means that slipsare reduced in frequency, but when they do occur they are 100 bytes in duration.
All digital equipment has buffers, and buffers can be found at various levels.Each level can therefore introduce a slip. The facsimile machine can introducea slip and the transmission equipment between the facsimile machines can alsoslip.
Table 10-2 Service Sensitivity to Synchronisation
Service Type Impact on Service
Voice Low impact. Occasional click
PABX Corruption of inter-exchange signalling PABX may lock-up.
Facsimile Distorted or missing facsimile lines
Modem traffic,(for example, Internet)
Loss of data or dropped calls
Cellular phone dropped calls, poor cell hand-over, poor voice quality, Cell base station may lock-up - possible visit needed.
Data(for example, leased-line)
lost, repeated or corrupted information
Digital video conferencing Distorted or frozen picture. Worse for modern systems
Digital broadcast video Picture subject to colour burst through to a distorted or frozen picture.
Encrypted data (for exam-ple, financial services)
Possible loss of data through to possible attacks on the security of the data.
159 16010 Synchronisation
161 160 161 160Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
If we consider our facsimile machine, it can receive or lose many bytes worth ofdata. Lost data may simply go undetected or repeated data may look �error free�since it was correctly sent before from the buffer. These conditions are calledbuffer overflow and underflow. Depending upon when the slip occurs, the resultmay look like horizontal black streaks across the page or as though the text hasbeen stretched vertically since we introduced extra lines. This may be acceptableto the user, however, if slips are occurring frequently, this may occur in severalplaces on every page.
It is extremely difficult to totally eliminate slips, but good qualitysynchronisation reduces the occurrence of a slip to once every 72 days.Facsimile users are therefore extremely unlikely to be affected.
Dial-up modem traffic is also sensitive to slips. Data received by the modem iseither lost or duplicated. In some cases the modem may be able to recover if thedata is detected as errored or missing. Modems are, however, a rapidly changingtechnology, and modem speeds are rapidly increasing. The latest generation ofmodems now in common use are capable of 56 kbit/s from the exchange to theuser and 33.2 kbit/s from the user to the exchange. As modem speeds increase,the modems become more sensitive to errors and may be unable to recover froma slip, by re-negotiating the link. The direct digital modulation technique used todeliver 56 kbit/s is sensitive to sips, and relies upon the receiving modem beingwell designed to recover from slips. Modems generally drop the connectionwhen they are unable to recover the link. If slips are occurring frequently, themodems may negotiate a lower link speed such as 19.2 kbit/s, which mayprompt the user to ring the operator asking why their modem never connects atthe fastest speed.
The normal reaction of a user when the connection is dropped is to re-dial.Indeed the computer may do this automatically. Dropped connections can beextremely frustrating for the user, particularly when the call drops at the end ofa long file transfer. If calls are dropped frequently, users tend to complain andwill burden technical support personnel. Additionally fast re-dialling can burdenthe call processing engine of the telephone exchange. Poor synchronisation canas a direct consequence place an additional load on the exchange.
It should be noted however, that synchronisation is not the sole cause of droppedcalls and slow link speed negotiation. Some of the popular modemmanufacturers have poorly designed modems which exhibit these symptoms
when connecting to modems from other vendors! For example, a perfectlysynchronised SDH network could be carrying traffic between facsimilemachines which are badly synchronised. The errors would not be due to theSDH synchronisation.
Cellular
Mobile phone networks make extensive use of signalling so that the callcontinues, as the user moves around. Each cell has a base station that connectsa mobile phone to the network. Whenever a user moves from one cell to another,the base stations negotiate amongst themselves to agree which cell the user hasmoved into. This process is called �handover� .
If the base stations are not synchronised, the chances of the call being droppedare increased. Even if the call is not dropped, the user will notice distortion in thevoice quality. Digital phone systems such as GSM or PCN are especiallysusceptible to this problem.
Because digital phone systems, such as GSM use highly compressed voice data,data corruption can cause serious degradation in voice quality. To the user thiscan sound like unnatural periods of silence or unnatural voice quality. Unlikeuncompressed voice, slips do not sound like a simple click.
Poor synchronisation therefore, is directly noticeable by the end user. He willperceive the system at best as poor voice quality and at worst as an almostunusable service where calls have to be constantly re-dialled.
Revenue can also be impacted. There is the obvious scenario where the userswitches to another network to get a usable service, but in the now highlycompetitive mobile phone market, operators are turning to tariffing to attractusers. Some operators now offer rebates to customers if the call is dropped. If theuser re-dials a number within a few seconds of the connection being closed, theuser is automatically credited. A poorly synchronised network can thereforeresult in a drop in revenue.
There may also be an operational costs associated with poor synchronisation.Most base station equipments treat slips as a symptom of network unavailability.This can cause the base station to go �off-air�. Some types of equipment requirea site visit to reset the base-station.
161 16210 Synchronisation
163 162 163 162Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
If we consider our facsimile machine, it can receive or lose many bytes worth ofdata. Lost data may simply go undetected or repeated data may look �error free�since it was correctly sent before from the buffer. These conditions are calledbuffer overflow and underflow. Depending upon when the slip occurs, the resultmay look like horizontal black streaks across the page or as though the text hasbeen stretched vertically since we introduced extra lines. This may be acceptableto the user, however, if slips are occurring frequently, this may occur in severalplaces on every page.
It is extremely difficult to totally eliminate slips, but good qualitysynchronisation reduces the occurrence of a slip to once every 72 days.Facsimile users are therefore extremely unlikely to be affected.
Dial-up modem traffic is also sensitive to slips. Data received by the modem iseither lost or duplicated. In some cases the modem may be able to recover if thedata is detected as errored or missing. Modems are, however, a rapidly changingtechnology, and modem speeds are rapidly increasing. The latest generation ofmodems now in common use are capable of 56 kbit/s from the exchange to theuser and 33.2 kbit/s from the user to the exchange. As modem speeds increase,the modems become more sensitive to errors and may be unable to recover froma slip, by re-negotiating the link. The direct digital modulation technique used todeliver 56 kbit/s is sensitive to sips, and relies upon the receiving modem beingwell designed to recover from slips. Modems generally drop the connectionwhen they are unable to recover the link. If slips are occurring frequently, themodems may negotiate a lower link speed such as 19.2 kbit/s, which mayprompt the user to ring the operator asking why their modem never connects atthe fastest speed.
The normal reaction of a user when the connection is dropped is to re-dial.Indeed the computer may do this automatically. Dropped connections can beextremely frustrating for the user, particularly when the call drops at the end ofa long file transfer. If calls are dropped frequently, users tend to complain andwill burden technical support personnel. Additionally fast re-dialling can burdenthe call processing engine of the telephone exchange. Poor synchronisation canas a direct consequence place an additional load on the exchange.
It should be noted however, that synchronisation is not the sole cause of droppedcalls and slow link speed negotiation. Some of the popular modemmanufacturers have poorly designed modems which exhibit these symptoms
when connecting to modems from other vendors! For example, a perfectlysynchronised SDH network could be carrying traffic between facsimilemachines which are badly synchronised. The errors would not be due to theSDH synchronisation.
Cellular
Mobile phone networks make extensive use of signalling so that the callcontinues, as the user moves around. Each cell has a base station that connectsa mobile phone to the network. Whenever a user moves from one cell to another,the base stations negotiate amongst themselves to agree which cell the user hasmoved into. This process is called �handover� .
If the base stations are not synchronised, the chances of the call being droppedare increased. Even if the call is not dropped, the user will notice distortion in thevoice quality. Digital phone systems such as GSM or PCN are especiallysusceptible to this problem.
Because digital phone systems, such as GSM use highly compressed voice data,data corruption can cause serious degradation in voice quality. To the user thiscan sound like unnatural periods of silence or unnatural voice quality. Unlikeuncompressed voice, slips do not sound like a simple click.
Poor synchronisation therefore, is directly noticeable by the end user. He willperceive the system at best as poor voice quality and at worst as an almostunusable service where calls have to be constantly re-dialled.
Revenue can also be impacted. There is the obvious scenario where the userswitches to another network to get a usable service, but in the now highlycompetitive mobile phone market, operators are turning to tariffing to attractusers. Some operators now offer rebates to customers if the call is dropped. If theuser re-dials a number within a few seconds of the connection being closed, theuser is automatically credited. A poorly synchronised network can thereforeresult in a drop in revenue.
There may also be an operational costs associated with poor synchronisation.Most base station equipments treat slips as a symptom of network unavailability.This can cause the base station to go �off-air�. Some types of equipment requirea site visit to reset the base-station.
161 16210 Synchronisation
163 162 163 162Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
If slips are occurring frequently, users will notice that the network is phasing inand out of service.
Within Europe, asynchronous hand-over is used which is more tolerant to timingproblems such as jitter, than synchronous hand-over. This means that calls areless likely to be dropped moving between networks, but the use of this techniquedoes not provide a substitute for good synchronisation design.
Video Services
Video is playing an increasingly important role in our lives. Video conferencingusage is growing rapidly, especially amongst multi-national companies, anddigital television is delivering potentially hundreds of channels.
Video signals contain vast amounts of data compared to voice, and users aremore tolerant of poor sound quality than poor video. A user might happilytolerate an audio click every ten seconds yet would find a picture freezing everyten seconds extremely annoying.
Synchronisation slips affect digital video signals in very noticeable ways. Someof the symptoms include freezing of the picture, picture distortion and colourbursts. The latest generation of digital video systems are very bandwidthefficient, but rely upon virtually error free transmission. The more compressedthe video information is, the more susceptible the signal is to synchronisationproblems.
Networks, which carry video traffic, therefore, need to be designed to avoidslips. This precaution may be enough for video conferencing where picturequality is not a premium concern, but for broadcast quality video, simplyavoiding slips is still not good enough.
Another symptom of poor synchronisation is jitter. Jitter can be thought of asnoise on the clock signal. Most services such as voice are not usually affected byjitter, but the same cannot be said for broadcast video, and to a lesser extentvideo conferencing. Jitter affects picture quality, which users would notice ascolour burst patterns on the screen. These effects can be minimised by liberal useof specialised external timing equipment or special pointer leaking mechanismsas found on Nortel�s TN-16X product.
Another consideration is the service provided to advertisers who indirectly payfor a lot of television programming. Television companies consequently take theviews of advertisers very seriously and will endeavour to provide the bestpossible quality video.
With broadcast TV, it is possible to introduce very large buffers to accommodateany feasible slip or jitter. This would be required on networks where the data rateis not fixed like SDH. Large buffers would be found in ATM or IP networks.Buffers do however, introduce delays. With broadcast TV, large delays ofpotentially seconds, may not be a problem, but with video conferencing, delaysas large as this, would make discussions difficult. Although large buffers maysolve a problem at one level, they may introduce a problem at another level,requiring the need for time of day synchronisation, where time stamps areperiodically required to define when data is delivered.
Synchronisation Across Multiple Operators
Establishing a synchronisation policy within a single operator�s network is arealistic task, yet modern-day communication often requires data to crossmultiple networks. It is unrealistic for an operator to force their synchronisationpolicy onto another network.
Few operators are willing to rely upon another operator for the timing for theirwhole network. Loss of that timing could dramatically degrade the performanceof their network and hence start impacting service quality and revenues.
Competitive operators are faced with the dilemma of balancing independenceagainst service quality. The new operator�s traffic tends to be routed via theestablished operator, hence there is a clear need to minimise slips between thenetworks. The obvious solution to this is to synchronise the network to theestablished operator, however, this is a high risk strategy since the quality ofservice delivered by the network is highly reliant upon the established operator.The quality of service delivered to all the new operator�s customers willultimately be dependent upon the established operator.
In addition to this problem, the established operator probably still has anextensive PDH legacy while the new operator has built its green-field networkusing SDH. The synchronisation needs of SDH are much higher than PDH,
163 16410 Synchronisation
165 164 165 164Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
If slips are occurring frequently, users will notice that the network is phasing inand out of service.
Within Europe, asynchronous hand-over is used which is more tolerant to timingproblems such as jitter, than synchronous hand-over. This means that calls areless likely to be dropped moving between networks, but the use of this techniquedoes not provide a substitute for good synchronisation design.
Video Services
Video is playing an increasingly important role in our lives. Video conferencingusage is growing rapidly, especially amongst multi-national companies, anddigital television is delivering potentially hundreds of channels.
Video signals contain vast amounts of data compared to voice, and users aremore tolerant of poor sound quality than poor video. A user might happilytolerate an audio click every ten seconds yet would find a picture freezing everyten seconds extremely annoying.
Synchronisation slips affect digital video signals in very noticeable ways. Someof the symptoms include freezing of the picture, picture distortion and colourbursts. The latest generation of digital video systems are very bandwidthefficient, but rely upon virtually error free transmission. The more compressedthe video information is, the more susceptible the signal is to synchronisationproblems.
Networks, which carry video traffic, therefore, need to be designed to avoidslips. This precaution may be enough for video conferencing where picturequality is not a premium concern, but for broadcast quality video, simplyavoiding slips is still not good enough.
Another symptom of poor synchronisation is jitter. Jitter can be thought of asnoise on the clock signal. Most services such as voice are not usually affected byjitter, but the same cannot be said for broadcast video, and to a lesser extentvideo conferencing. Jitter affects picture quality, which users would notice ascolour burst patterns on the screen. These effects can be minimised by liberal useof specialised external timing equipment or special pointer leaking mechanismsas found on Nortel�s TN-16X product.
Another consideration is the service provided to advertisers who indirectly payfor a lot of television programming. Television companies consequently take theviews of advertisers very seriously and will endeavour to provide the bestpossible quality video.
With broadcast TV, it is possible to introduce very large buffers to accommodateany feasible slip or jitter. This would be required on networks where the data rateis not fixed like SDH. Large buffers would be found in ATM or IP networks.Buffers do however, introduce delays. With broadcast TV, large delays ofpotentially seconds, may not be a problem, but with video conferencing, delaysas large as this, would make discussions difficult. Although large buffers maysolve a problem at one level, they may introduce a problem at another level,requiring the need for time of day synchronisation, where time stamps areperiodically required to define when data is delivered.
Synchronisation Across Multiple Operators
Establishing a synchronisation policy within a single operator�s network is arealistic task, yet modern-day communication often requires data to crossmultiple networks. It is unrealistic for an operator to force their synchronisationpolicy onto another network.
Few operators are willing to rely upon another operator for the timing for theirwhole network. Loss of that timing could dramatically degrade the performanceof their network and hence start impacting service quality and revenues.
Competitive operators are faced with the dilemma of balancing independenceagainst service quality. The new operator�s traffic tends to be routed via theestablished operator, hence there is a clear need to minimise slips between thenetworks. The obvious solution to this is to synchronise the network to theestablished operator, however, this is a high risk strategy since the quality ofservice delivered by the network is highly reliant upon the established operator.The quality of service delivered to all the new operator�s customers willultimately be dependent upon the established operator.
In addition to this problem, the established operator probably still has anextensive PDH legacy while the new operator has built its green-field networkusing SDH. The synchronisation needs of SDH are much higher than PDH,
163 16410 Synchronisation
165 164 165 164Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
hence the timing information may be inadequate for an SDH network. The newoperator, therefore, needs a �firewall� approach to gain independence from theestablished operator.
The use of external timing equipment at the boundary between the network canaddress this problem. The incoming timing can be vetted and used provided itmeets the appropriate quality level. If the quality drops below a certain level, theequipment can swap over to its own high quality reference. This equipment canalso compile extensive statistics on the quality of the incoming timing signal.Since operators tend to charge significant sums for a timing feed, the newoperator will have evidence of the established operator�s performance and hencecan negotiate rebates.
Global Timing
As communication becomes increasingly global, and the world becomes anetworked planet, the need for reliable error free communication across multipleoperators will continue to grow. This raises the question, should the Worldsynchronise to the same timing source?
If all network operators synchronised their networks to the same timing source,the potential for slips, and hence synchronisation related errors, would bevirtually eliminated.
The Global Positioning System (GPS), lets network operators do exactly that.Most people are more familiar with GPS as a method for navigation, but GPS isalso an extremely accurate source of timing.
Use of GPS as a network timing source is growing, and as the number increases,many of today�s timing problems associated with connecting networks together,will be reduced.
GPS can be viewed as the timing source for the planet, and each network in theworld will be synchronised to this single source. All networks will besynchronised to the same source, hence clock slips will be almost eliminated.
GPS timing availability, however, cannot be guaranteed, hence each networkwill still needs its own back-up clock to provide a redundant timing sourceduring periods of GPS unavailability.
Operators that design their network today with GPS technology will be ready forthe future.
Synchronisation Basics
The terms timing and synchronisation are often confused. A timing source is themaster frequency reference whereas synchronisation is the alignment of localequipment frequency to the master and the distribution of timing around anetwork. The term clock is often used to refer to a timing source. The termsynchronisation is also confusingly used to describe both timing sources as wellas distributing timing information.
If the timing source is poor quality, it does not matter how good the networksynchronisation is, it will still be distributing poor quality timing. Similarly ahigh quality timing source, used with a poorly designed synchronisation systemwill distribute poor quality timing information, affecting the quality of servicedelivered to customers.
All digital equipment has an internal timing source, the accuracy of which isdependent upon the type of equipment. This internal timing source has atolerance which varies with temperature and voltage, and hence must besynchronised to another clock to minimise the small timing differences that willexist between internal equipment clocks. Most equipment usually has a facilityto synchronise the internal clock to a range of external timing sources. SDHequipment has this facility built-in, and provides timing distribution features.
If equipment timing differences are not minimised, slips (bursts of errors) willoccur periodically. In the simple example where two equipments are connectedtogether, one clock is free-running while the other equipment is synchronised tothe other equipment�s free-running clock. Timing is essentially �analogue� andhence even when clocks are synchronised, or locked together, small errors arepresent. These errors are due to wander and jitter. When we start building morecomplex networks where another piece of equipment is synchronised toequipment that is synchronised to another free-running clock, these errors startto accumulate as illustrated in Figure 10-1. The underlying synchronisationsignal frequency, however, remains accurate. Fairly long or complexsynchronisation networks can be quickly constructed.
165 16610 Synchronisation
167 166 167 166Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
hence the timing information may be inadequate for an SDH network. The newoperator, therefore, needs a �firewall� approach to gain independence from theestablished operator.
The use of external timing equipment at the boundary between the network canaddress this problem. The incoming timing can be vetted and used provided itmeets the appropriate quality level. If the quality drops below a certain level, theequipment can swap over to its own high quality reference. This equipment canalso compile extensive statistics on the quality of the incoming timing signal.Since operators tend to charge significant sums for a timing feed, the newoperator will have evidence of the established operator�s performance and hencecan negotiate rebates.
Global Timing
As communication becomes increasingly global, and the world becomes anetworked planet, the need for reliable error free communication across multipleoperators will continue to grow. This raises the question, should the Worldsynchronise to the same timing source?
If all network operators synchronised their networks to the same timing source,the potential for slips, and hence synchronisation related errors, would bevirtually eliminated.
The Global Positioning System (GPS), lets network operators do exactly that.Most people are more familiar with GPS as a method for navigation, but GPS isalso an extremely accurate source of timing.
Use of GPS as a network timing source is growing, and as the number increases,many of today�s timing problems associated with connecting networks together,will be reduced.
GPS can be viewed as the timing source for the planet, and each network in theworld will be synchronised to this single source. All networks will besynchronised to the same source, hence clock slips will be almost eliminated.
GPS timing availability, however, cannot be guaranteed, hence each networkwill still needs its own back-up clock to provide a redundant timing sourceduring periods of GPS unavailability.
Operators that design their network today with GPS technology will be ready forthe future.
Synchronisation Basics
The terms timing and synchronisation are often confused. A timing source is themaster frequency reference whereas synchronisation is the alignment of localequipment frequency to the master and the distribution of timing around anetwork. The term clock is often used to refer to a timing source. The termsynchronisation is also confusingly used to describe both timing sources as wellas distributing timing information.
If the timing source is poor quality, it does not matter how good the networksynchronisation is, it will still be distributing poor quality timing. Similarly ahigh quality timing source, used with a poorly designed synchronisation systemwill distribute poor quality timing information, affecting the quality of servicedelivered to customers.
All digital equipment has an internal timing source, the accuracy of which isdependent upon the type of equipment. This internal timing source has atolerance which varies with temperature and voltage, and hence must besynchronised to another clock to minimise the small timing differences that willexist between internal equipment clocks. Most equipment usually has a facilityto synchronise the internal clock to a range of external timing sources. SDHequipment has this facility built-in, and provides timing distribution features.
If equipment timing differences are not minimised, slips (bursts of errors) willoccur periodically. In the simple example where two equipments are connectedtogether, one clock is free-running while the other equipment is synchronised tothe other equipment�s free-running clock. Timing is essentially �analogue� andhence even when clocks are synchronised, or locked together, small errors arepresent. These errors are due to wander and jitter. When we start building morecomplex networks where another piece of equipment is synchronised toequipment that is synchronised to another free-running clock, these errors startto accumulate as illustrated in Figure 10-1. The underlying synchronisationsignal frequency, however, remains accurate. Fairly long or complexsynchronisation networks can be quickly constructed.
165 16610 Synchronisation
167 166 167 166Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 10-1 Synch Chains
External Timing Equipment
The accuracy of the internal equipment clocks is deliberately limited becauseaccuracy is expensive. When long synchronisation chains are constructed,however, the accuracy of the free-running clock needs to be increased. Thenormal way to achieve this is to use equipment dedicated for this purpose.Specialised equipment, termed a reference clock or timing generator, provides ahighly accurate timing source, usually derived from an atomic clock. The headof the chain then needs to synchronise its clock to this external reference clock.
Every piece of equipment that passes timing information along the chain slightlydegrades the accuracy of the clock signal. After passing through about 20 piecesof equipment, the clock signal degrades to a point where slips can occur andhence traffic becomes errored. A special purpose external clock filter, termed aSynchronisation Supply Unit (SSU) is then required to process and recover theclock. Another name for this type of unit is a Transit Node Clock (TNC).
Reference Clock Sources
Reference clocks come in a variety of forms, each of which has strengths andweaknesses. The ultimate timing accuracy is derived from Cesium, however,cheaper alternatives such as Rubidium and Quartz are available. GPS is rapidlygrowing in popularity as a very high accuracy timing source, usually used inconjunction with Cesium or Rubidium.
GPS uses a ring of satellites around the earth distributing timing information,primarily intended for military use. GPS provides Cesium level accuracy for afraction of the cost. GPS timing information, however, is not entirely reliableand is subject to outages, sometimes deliberately introduced by the US military.When used with Cesium or Rubidium as a back-up, the combination can providethe ultimate global timing reference.
Synchronising SDH Networks
High accuracy clocks are expensive so the fibre optic transmission network isused to distribute synchronisation. This avoids the need to put a very highaccuracy timing source at every location in the network.
The principles used to synchronise PDH networks are very different to thoseused by SDH networks. A key difference is that PDH embeds timinginformation in the multiplexing structure and hence a multiplexed signal can beused to carry timing, whereas SDH transfers timing using the optical line signal.
It should be noted that SDH network elements that are operating asynchronouslywill carry traffic error free, however, they are designed to be synchronous.Pointer movements will result in asynchronous operation, causing jitter andwander to the tributary output signals.
Coupled with the fact that SDH enables complex mesh/ring networks to berealised, something which was not possible with PDH, means that thesynchronisation design for SDH networks is more complex than PDH networks.
Timing loops occur when each piece of equipment in a ring is synchronised tothe adjacent equipment. If the previous chain example is considered, a loopoccurs when the headend becomes synchronised to the other end of the chain. Atiming loop rapidly degrades the accuracy of every equipment clock in the loop,and traffic becomes seriously errored.
SDH networks make extensive use of rings, hence timing loops are a real threat.SDH networks, therefore, need to be designed to minimise the opportunity fortiming loops to occur, especially under fault scenarios.
Although synchronisation of SDH networks is more complex than that of PDHnetworks, SDH provides functionality such as Synchronisation Status
Equipment EquipmentEquipment
ClockSource
Quality= 100%
Quality= 97%
Quality= 94%
SynchDistribution
SynchDistribution
167 16810 Synchronisation
169 168 169 168Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
Figure 10-1 Synch Chains
External Timing Equipment
The accuracy of the internal equipment clocks is deliberately limited becauseaccuracy is expensive. When long synchronisation chains are constructed,however, the accuracy of the free-running clock needs to be increased. Thenormal way to achieve this is to use equipment dedicated for this purpose.Specialised equipment, termed a reference clock or timing generator, provides ahighly accurate timing source, usually derived from an atomic clock. The headof the chain then needs to synchronise its clock to this external reference clock.
Every piece of equipment that passes timing information along the chain slightlydegrades the accuracy of the clock signal. After passing through about 20 piecesof equipment, the clock signal degrades to a point where slips can occur andhence traffic becomes errored. A special purpose external clock filter, termed aSynchronisation Supply Unit (SSU) is then required to process and recover theclock. Another name for this type of unit is a Transit Node Clock (TNC).
Reference Clock Sources
Reference clocks come in a variety of forms, each of which has strengths andweaknesses. The ultimate timing accuracy is derived from Cesium, however,cheaper alternatives such as Rubidium and Quartz are available. GPS is rapidlygrowing in popularity as a very high accuracy timing source, usually used inconjunction with Cesium or Rubidium.
GPS uses a ring of satellites around the earth distributing timing information,primarily intended for military use. GPS provides Cesium level accuracy for afraction of the cost. GPS timing information, however, is not entirely reliableand is subject to outages, sometimes deliberately introduced by the US military.When used with Cesium or Rubidium as a back-up, the combination can providethe ultimate global timing reference.
Synchronising SDH Networks
High accuracy clocks are expensive so the fibre optic transmission network isused to distribute synchronisation. This avoids the need to put a very highaccuracy timing source at every location in the network.
The principles used to synchronise PDH networks are very different to thoseused by SDH networks. A key difference is that PDH embeds timinginformation in the multiplexing structure and hence a multiplexed signal can beused to carry timing, whereas SDH transfers timing using the optical line signal.
It should be noted that SDH network elements that are operating asynchronouslywill carry traffic error free, however, they are designed to be synchronous.Pointer movements will result in asynchronous operation, causing jitter andwander to the tributary output signals.
Coupled with the fact that SDH enables complex mesh/ring networks to berealised, something which was not possible with PDH, means that thesynchronisation design for SDH networks is more complex than PDH networks.
Timing loops occur when each piece of equipment in a ring is synchronised tothe adjacent equipment. If the previous chain example is considered, a loopoccurs when the headend becomes synchronised to the other end of the chain. Atiming loop rapidly degrades the accuracy of every equipment clock in the loop,and traffic becomes seriously errored.
SDH networks make extensive use of rings, hence timing loops are a real threat.SDH networks, therefore, need to be designed to minimise the opportunity fortiming loops to occur, especially under fault scenarios.
Although synchronisation of SDH networks is more complex than that of PDHnetworks, SDH provides functionality such as Synchronisation Status
Equipment EquipmentEquipment
ClockSource
Quality= 100%
Quality= 97%
Quality= 94%
SynchDistribution
SynchDistribution
167 16810 Synchronisation
169 168 169 168Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
Messaging (SSMB) and remote management of synchronisation functions, tohelp operational staff to manage the synchronisation network.
Figure 10-2 Timing Loops
Synchronisation Status Messaging
SSMB is an enhancement to SDH to provide timing configuration managementin SDH rings. SSMB can minimise the risks of timing loops and free-runningclocks in SDH rings. Although SSMB can improve the resilience of asynchronous network, it is not a substitute for design. SSMB does not design thenetwork for you, it is basically a protection mechanism. Although it is feasibleto use SSMB in mesh topologies, care must be taken to ensure that the protocoldoes not cause timing loops.
SSMB is a messaging protocol where SDH elements can distribute informationabout the quality of clocks that are available. In the event of a fibre break or
similar network fault, the multiplexers exchange information about clockquality, so that each element can decide on which clocks to use.
Used carefully, SSMB can greatly improve the resilience of the network, byeliminating timing loops and free-running clocks. SSMB does, however, havesome weaknesses which need to be taken into consideration when thesynchronisation network is being designed. SSMB does not provide anymechanism to �degrade� the clock as it is propagated around a network. If anetwork element is locked to a high quality clock it will pass on a good SSMBvalue. The next network element locks to this good clock and passes on the samevalue. Each network element does the same while the clock quality valueremains the same. It must be remembered that each network element slightlydegrades the quality of the clock and hence after the clock has passed through anumber of network elements, its real quality will not be as good as the originalSSMB value.
Another factor, which needs to be given consideration in synchronisationnetwork design, is that distribution of the SSMB information around thenetwork takes time. This means that in meshes, and to some extent rings, thetime taken for a new SSMB value to be distributed by each network elementtakes time. If a network element locks to a new clock and hence a new SSMBvalue too quickly, the network many not have had enough time to stabilise. Outof date values may still be in the network and the network may end up chasinga non-existent clock, rather like a dog chases its tail. This, at best, results in rapidswitching between clock sources at each network element or at worst, a massivetiming loop. The problem cannot be totally eliminated but the chances of itoccurring can be minimised by careful network design and by configuration ofthe network elements to slow their response time to switch to a new clock.
Although networks using SSMB must be carefully designed, they permit rapidautomatic reconfiguration of a synchronous network when there is a failure of asynchronisation source and greatly reduce the duration and quantity of pointermovements.
Well synchronised ring Timing loop
169 17010 Synchronisation
170Synchronisation Synchronisation
Synchronous Transmission Systems Synchronous Transmission Systems
Messaging (SSMB) and remote management of synchronisation functions, tohelp operational staff to manage the synchronisation network.
Figure 10-2 Timing Loops
Synchronisation Status Messaging
SSMB is an enhancement to SDH to provide timing configuration managementin SDH rings. SSMB can minimise the risks of timing loops and free-runningclocks in SDH rings. Although SSMB can improve the resilience of asynchronous network, it is not a substitute for design. SSMB does not design thenetwork for you, it is basically a protection mechanism. Although it is feasibleto use SSMB in mesh topologies, care must be taken to ensure that the protocoldoes not cause timing loops.
SSMB is a messaging protocol where SDH elements can distribute informationabout the quality of clocks that are available. In the event of a fibre break or
similar network fault, the multiplexers exchange information about clockquality, so that each element can decide on which clocks to use.
Used carefully, SSMB can greatly improve the resilience of the network, byeliminating timing loops and free-running clocks. SSMB does, however, havesome weaknesses which need to be taken into consideration when thesynchronisation network is being designed. SSMB does not provide anymechanism to �degrade� the clock as it is propagated around a network. If anetwork element is locked to a high quality clock it will pass on a good SSMBvalue. The next network element locks to this good clock and passes on the samevalue. Each network element does the same while the clock quality valueremains the same. It must be remembered that each network element slightlydegrades the quality of the clock and hence after the clock has passed through anumber of network elements, its real quality will not be as good as the originalSSMB value.
Another factor, which needs to be given consideration in synchronisationnetwork design, is that distribution of the SSMB information around thenetwork takes time. This means that in meshes, and to some extent rings, thetime taken for a new SSMB value to be distributed by each network elementtakes time. If a network element locks to a new clock and hence a new SSMBvalue too quickly, the network many not have had enough time to stabilise. Outof date values may still be in the network and the network may end up chasinga non-existent clock, rather like a dog chases its tail. This, at best, results in rapidswitching between clock sources at each network element or at worst, a massivetiming loop. The problem cannot be totally eliminated but the chances of itoccurring can be minimised by careful network design and by configuration ofthe network elements to slow their response time to switch to a new clock.
Although networks using SSMB must be carefully designed, they permit rapidautomatic reconfiguration of a synchronous network when there is a failure of asynchronisation source and greatly reduce the duration and quantity of pointermovements.
Well synchronised ring Timing loop
169 17010 Synchronisation
171
Synchronous Transmission Systems
171 Evolving SDH to support Data-centric and Wavelength Services
Evolving SDH to support Data-centric and Wavelength Services
There is an increasing need to transport new, non-SDH framed data ser-
vices in the metro area and to extend the reach of these services from the
metro area across national or international networks. Typical examples
are Gigabit Ethernet services for corporate LAN extension, and storage
services to provide Business Continuity and Disaster Recovery capabili-
ties.
This chapter examines two key technologies in this field – Ethernet over
SDH (EoSDH) using Generic Framing Procedure (GFP) and Resilient
Packet Ring (RPR), both which allow such data-centric services to be
carried across SDH core networks.
GFP will typically be used more frequently in point-to-point and point-
to-multipoint applications similar to wavelength services, whereas RPR
is targeted towards LAN Services across ring-based architectures. Which
of these technologies is better suited will depend on the specific network
and service mix being addressed and can be decided on a case by case
basis.
Deploying metro-to-long haul wavelength services us-ing Generic Framing Procedure (GFP)
Introduction
Modern SDH solutions meet the key enterprise need to transport a wide
range of non-SDH protocols including ESCON, FICON (IBM main-
frame networking), Fibre Channel (Storage Area Networks), Ethernet
and Gigabit Ethernet (GbE) over a Metro Area Network (MAN).
11
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 172
Previously, enterprises deployed translation devices routers to convert
these protocols to standard TDM interfaces such as E1 and E3, or data
interfaces such as ATM, carried over SDH channels, usually leased from
network operators. This approach often required the client signal to be
transmitted at less than its native rate, and to undergo complex protocol
translations at the network edge. For the fixed cost of SDH/DWDM
equipment and dark fibre, an enterprise could significantly reduce its
monthly leased line costs while transporting these protocols at their na-
tive rates.
As the enterprise demand for these protocols has increased, so has the
requirement to extend their reach – for example, in storage networking
where Business Continuity sites must be suitably remote from the main
data centre. Modern solutions can be extended to create regional optical
networks, but in most countries there is an existing huge investment in
regional and long haul networking – the SDH network. The challenge is
to enable non-SDH services to be carried effectively over this existing
infrastructure.
For this purpose Generic Framing Procedure (GFP) has been introduced
- a standard developed in North America (T1X1) and internationally
(ITU G.7041) as a generic encapsulation method for data transport. GFP
thus provides a bridge linking data and SDH networks. By mapping
broadband data traffic using GFP, these enterprise-focused services can
now be transported and managed over the vast number of SDH networks
available today.
Network Architecture
Ethernet over SDH is a convenient access technology for the enterprise,
since it allows client signals to be in their native formats at the point
where they are handed over to the network. Most regional and long haul
core networks are primarily based on SDH today, and SDH provides the
necessary tools to deliver SLA measurements and end-to-end service
assurance.
The best approach is to combine EoSDH access with the current SDH
metro/long haul core infrastructure. In this approach broadband services
are wrapped with a SDH wrapper at the ingress of network access sys-
tems. This will enable these services to be transported over the existing
SDH network using existing SDH networking tools and methods.
11 Evolving SDH to support Data-centric and Wavelength Services
172
Synchronous Transmission Systems
171 Evolving SDH to support Data-centric and Wavelength Services
Evolving SDH to support Data-centric and Wavelength Services
There is an increasing need to transport new, non-SDH framed data ser-
vices in the metro area and to extend the reach of these services from the
metro area across national or international networks. Typical examples
are Gigabit Ethernet services for corporate LAN extension, and storage
services to provide Business Continuity and Disaster Recovery capabili-
ties.
This chapter examines two key technologies in this field – Ethernet over
SDH (EoSDH) using Generic Framing Procedure (GFP) and Resilient
Packet Ring (RPR), both which allow such data-centric services to be
carried across SDH core networks.
GFP will typically be used more frequently in point-to-point and point-
to-multipoint applications similar to wavelength services, whereas RPR
is targeted towards LAN Services across ring-based architectures. Which
of these technologies is better suited will depend on the specific network
and service mix being addressed and can be decided on a case by case
basis.
Deploying metro-to-long haul wavelength services us-ing Generic Framing Procedure (GFP)
Introduction
Modern SDH solutions meet the key enterprise need to transport a wide
range of non-SDH protocols including ESCON, FICON (IBM main-
frame networking), Fibre Channel (Storage Area Networks), Ethernet
and Gigabit Ethernet (GbE) over a Metro Area Network (MAN).
11
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 172
Previously, enterprises deployed translation devices routers to convert
these protocols to standard TDM interfaces such as E1 and E3, or data
interfaces such as ATM, carried over SDH channels, usually leased from
network operators. This approach often required the client signal to be
transmitted at less than its native rate, and to undergo complex protocol
translations at the network edge. For the fixed cost of SDH/DWDM
equipment and dark fibre, an enterprise could significantly reduce its
monthly leased line costs while transporting these protocols at their na-
tive rates.
As the enterprise demand for these protocols has increased, so has the
requirement to extend their reach – for example, in storage networking
where Business Continuity sites must be suitably remote from the main
data centre. Modern solutions can be extended to create regional optical
networks, but in most countries there is an existing huge investment in
regional and long haul networking – the SDH network. The challenge is
to enable non-SDH services to be carried effectively over this existing
infrastructure.
For this purpose Generic Framing Procedure (GFP) has been introduced
- a standard developed in North America (T1X1) and internationally
(ITU G.7041) as a generic encapsulation method for data transport. GFP
thus provides a bridge linking data and SDH networks. By mapping
broadband data traffic using GFP, these enterprise-focused services can
now be transported and managed over the vast number of SDH networks
available today.
Network Architecture
Ethernet over SDH is a convenient access technology for the enterprise,
since it allows client signals to be in their native formats at the point
where they are handed over to the network. Most regional and long haul
core networks are primarily based on SDH today, and SDH provides the
necessary tools to deliver SLA measurements and end-to-end service
assurance.
The best approach is to combine EoSDH access with the current SDH
metro/long haul core infrastructure. In this approach broadband services
are wrapped with a SDH wrapper at the ingress of network access sys-
tems. This will enable these services to be transported over the existing
SDH network using existing SDH networking tools and methods.
11 Evolving SDH to support Data-centric and Wavelength Services
173
Synchronous Transmission Systems
173 Evolving SDH to support Data-centric and Wavelength Services
Figure 11-1
Broadband Service Adaptation
The goal is to provide ‘managed transparency’ for broadband services.
In other words, to the network user the network looks like a transparent
pipe, but the carrier has comprehensive visibility of the performance of
the service end-to-end across the network. In this way a detailed and
flexible set of SLAs can be defined and monitored. This section will
address the functions that must occur at the edge of the network for ser-
vice transparency.
The service mapping function shown in Figure 11-1 includes the physi-
cal service interface, bi-directional service performance measurement
(input to the network, and egress from the network), and service encap-
sulation.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 174
Figure 11-2
Network adaptation creates common networking containers compatible
with the network technology. In the case of optical networking, SDH or
SONET compatible containers are optimum. To provide great flexibility
in service offerings, GFP is an optimal solution for broadband traffic.
GFP containers enable multiplexing and switching, and make service
monitoring information readily available for both network management
and control systems. The networking attributes are key functions that
allow the carrier to manage connectivity and support efficient network
and service operations. Additionally, this adaptation is critical for man-
agement of traffic that is handed off between carriers. Unambiguous
service performance data must be readily available, without de-mapping,
at key parts of the network.
Generic Framing Procedure - An overview
Generic Framing Procedure is an encapsulation method whose definition
has been completed and documented as a global standard by ITU-T
G.7041. Two modes of GFP are defined: transparently-coded and frame-
mapped.
11 Evolving SDH to support Data-centric and Wavelength Services
174
Synchronous Transmission Systems
173 Evolving SDH to support Data-centric and Wavelength Services
Figure 11-1
Broadband Service Adaptation
The goal is to provide ‘managed transparency’ for broadband services.
In other words, to the network user the network looks like a transparent
pipe, but the carrier has comprehensive visibility of the performance of
the service end-to-end across the network. In this way a detailed and
flexible set of SLAs can be defined and monitored. This section will
address the functions that must occur at the edge of the network for ser-
vice transparency.
The service mapping function shown in Figure 11-1 includes the physi-
cal service interface, bi-directional service performance measurement
(input to the network, and egress from the network), and service encap-
sulation.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 174
Figure 11-2
Network adaptation creates common networking containers compatible
with the network technology. In the case of optical networking, SDH or
SONET compatible containers are optimum. To provide great flexibility
in service offerings, GFP is an optimal solution for broadband traffic.
GFP containers enable multiplexing and switching, and make service
monitoring information readily available for both network management
and control systems. The networking attributes are key functions that
allow the carrier to manage connectivity and support efficient network
and service operations. Additionally, this adaptation is critical for man-
agement of traffic that is handed off between carriers. Unambiguous
service performance data must be readily available, without de-mapping,
at key parts of the network.
Generic Framing Procedure - An overview
Generic Framing Procedure is an encapsulation method whose definition
has been completed and documented as a global standard by ITU-T
G.7041. Two modes of GFP are defined: transparently-coded and frame-
mapped.
11 Evolving SDH to support Data-centric and Wavelength Services
175
Synchronous Transmission Systems
175 Evolving SDH to support Data-centric and Wavelength Services
Transparently coded mode
( GFP-T )
Frame-mapped mode ( GFP-F )
allows 8B10B-coded SAN and
LAN signals (e.g. GbE, FC, FI-
CON, ESCON) to be transported
and switched across an optical
network while preserving the rele-
vant 8B10B code information.
Permits multiplexing, at a client
frame granularity, of multiple cli-
ent ports/types
“streaming” mode—even if full
interface capacity (e.g. GbE) is not
being used.
“PDU-aware” mode—gaps be-
tween packets are indicated using
null frames—enabling the option
of statistical multiplexing of dif-
ferent GFP channels.
GFP is a powerful tool for relatively simple multi-service adaptation
(independence from Layer 2 or Layer 3 protocols) with broadband and
ultra-broadband capacity.
It is an interoperable global ITU standard with mapping to SDH,
SONET, and OTN networks. Especially when combined with virtual
concatenation, it provides efficient network resource utilization with low
overhead and the ability to utilize paths through the network that are
diversely routed and lower in capacity than the actual payload. Minimal
buffering is required at adaptation points for low latency and jitter per-
formance.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 176
Figure 11-3
Case studies in GFP for services
Figure 11-3 shows an example of three separate customers who want to
deploy GigE and Fibre Channel services across a metro core. This ap-
proach is typical of current deployments of Metro DWDM services.
Three pairs of fibre are used to support the three customer’s end-to-end
across the network. The cost of this approach becomes more significant
as we move to the core of the network as these fibre pairs could poten-
tially be shared by many customers. The end-to-end distance of the links
is limited to 150 km due to current DWDM limitations. These systems
are also difficult to engineer as the end-to-end fibre performance must be
characterized for each individual customer.
Finally, current DWDM technology does not provide comprehensive
service management and SLA enforcement so each system must be
monitored separately. Failures within the network become difficult to
track. All of these challenges have restricted this technology to only the
largest customers, so far.
11 Evolving SDH to support Data-centric and Wavelength Services
176
Synchronous Transmission Systems
175 Evolving SDH to support Data-centric and Wavelength Services
Transparently coded mode
( GFP-T )
Frame-mapped mode ( GFP-F )
allows 8B10B-coded SAN and
LAN signals (e.g. GbE, FC, FI-
CON, ESCON) to be transported
and switched across an optical
network while preserving the rele-
vant 8B10B code information.
Permits multiplexing, at a client
frame granularity, of multiple cli-
ent ports/types
“streaming” mode—even if full
interface capacity (e.g. GbE) is not
being used.
“PDU-aware” mode—gaps be-
tween packets are indicated using
null frames—enabling the option
of statistical multiplexing of dif-
ferent GFP channels.
GFP is a powerful tool for relatively simple multi-service adaptation
(independence from Layer 2 or Layer 3 protocols) with broadband and
ultra-broadband capacity.
It is an interoperable global ITU standard with mapping to SDH,
SONET, and OTN networks. Especially when combined with virtual
concatenation, it provides efficient network resource utilization with low
overhead and the ability to utilize paths through the network that are
diversely routed and lower in capacity than the actual payload. Minimal
buffering is required at adaptation points for low latency and jitter per-
formance.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 176
Figure 11-3
Case studies in GFP for services
Figure 11-3 shows an example of three separate customers who want to
deploy GigE and Fibre Channel services across a metro core. This ap-
proach is typical of current deployments of Metro DWDM services.
Three pairs of fibre are used to support the three customer’s end-to-end
across the network. The cost of this approach becomes more significant
as we move to the core of the network as these fibre pairs could poten-
tially be shared by many customers. The end-to-end distance of the links
is limited to 150 km due to current DWDM limitations. These systems
are also difficult to engineer as the end-to-end fibre performance must be
characterized for each individual customer.
Finally, current DWDM technology does not provide comprehensive
service management and SLA enforcement so each system must be
monitored separately. Failures within the network become difficult to
track. All of these challenges have restricted this technology to only the
largest customers, so far.
11 Evolving SDH to support Data-centric and Wavelength Services
177
Synchronous Transmission Systems
177 Evolving SDH to support Data-centric and Wavelength Services
Figure 11-4
In the scenario shown in Fig 11-4, both access and core networks are
deployed with shared DWDM rings. The last mile CPE-to-metro edge
links are carried on private DWDM systems, but the multiple customer
�are transported over shared DWDM infrastructure in the metro-edge to
core networks. This architecture is an improvement because:
Traffic from multiple users is carried over the same fibre pair,
making efficient use of access and core fibre. The architecture
will facilitate networking to different locations as any POP
traversed by these shared access and core DWDM rings can
now support DWDM services.
DWDM distances will also improve, as now the spans of the
DWDM shared rings are no longer limited to a 150 km end-to-
end distance limit. Engineering is simplified because the shared
rings only have to be engineered once. The only custom engi-
neering required is from the CPE to the metro edge POP and in-
terconnection of the customer �between shared rings.
Issues that must still be addressed include the continued need for end-to
end service management and support for increased distances. The other
key issue is the need to optically engineer each end-to-end service even
though it travels over a shared infrastructure. Every time a new service is
added, an engineer would need to calculate the fibre losses for each indi-
vidual lightpath being provisioned on an end-to-end basis. Finally, a
technician would be required to physically go to each and every POP
along the service path to patch the optical cables.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 178
Figure 11-5
This example in Figure 11-5 introduces Ethernet over SDH. Multiple
transparent services are GFP mapped into SDH VC4-4c payloads and are
carried as a VC4-4c lambda to the metro edge office. At this point, the
signal can be carried as a virtually concatenated SDH signal to a metro
core office and onward through the core long haul SDH network. The
network operator can take advantage of SDH overhead for line and path
troubleshooting and can transported the signal over extended distances
beyond the technical limitations of DWDM. At the final metro edge POP
to the CPE, the reverse process will take place taking the GFP mapped
SDH signal back to the corresponding transparent service.
Figure 11-6
This architecture addresses all of the issues previously identified:
λ
λ
11 Evolving SDH to support Data-centric and Wavelength Services
178
Synchronous Transmission Systems
177 Evolving SDH to support Data-centric and Wavelength Services
Figure 11-4
In the scenario shown in Fig 11-4, both access and core networks are
deployed with shared DWDM rings. The last mile CPE-to-metro edge
links are carried on private DWDM systems, but the multiple customer
�are transported over shared DWDM infrastructure in the metro-edge to
core networks. This architecture is an improvement because:
Traffic from multiple users is carried over the same fibre pair,
making efficient use of access and core fibre. The architecture
will facilitate networking to different locations as any POP
traversed by these shared access and core DWDM rings can
now support DWDM services.
DWDM distances will also improve, as now the spans of the
DWDM shared rings are no longer limited to a 150 km end-to-
end distance limit. Engineering is simplified because the shared
rings only have to be engineered once. The only custom engi-
neering required is from the CPE to the metro edge POP and in-
terconnection of the customer �between shared rings.
Issues that must still be addressed include the continued need for end-to
end service management and support for increased distances. The other
key issue is the need to optically engineer each end-to-end service even
though it travels over a shared infrastructure. Every time a new service is
added, an engineer would need to calculate the fibre losses for each indi-
vidual lightpath being provisioned on an end-to-end basis. Finally, a
technician would be required to physically go to each and every POP
along the service path to patch the optical cables.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 178
Figure 11-5
This example in Figure 11-5 introduces Ethernet over SDH. Multiple
transparent services are GFP mapped into SDH VC4-4c payloads and are
carried as a VC4-4c lambda to the metro edge office. At this point, the
signal can be carried as a virtually concatenated SDH signal to a metro
core office and onward through the core long haul SDH network. The
network operator can take advantage of SDH overhead for line and path
troubleshooting and can transported the signal over extended distances
beyond the technical limitations of DWDM. At the final metro edge POP
to the CPE, the reverse process will take place taking the GFP mapped
SDH signal back to the corresponding transparent service.
Figure 11-6
This architecture addresses all of the issues previously identified:
λ
λ
11 Evolving SDH to support Data-centric and Wavelength Services
179
Synchronous Transmission Systems
179 Evolving SDH to support Data-centric and Wavelength Services
It optimizes the use of metro core and access fibre and existing
SDH facilities
It solves any distance limitations of DWDM and can now trav-
erse SDH facilities over vast metro core or long haul networks
It provides full service assurance and SLA enforcement as pro-
vided by SDH today
No special engineering is needed as the core of the network fol-
lows the same engineering as SDH today. Services can also be
turned up easily since end-to-end provisioning is becoming
available for SDH-based core bandwidth
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 180
Resilient Packet Ring ( RPR ) technology
The metro bottleneck
Today, interconnecting enterprise local area networks (LANs) across
metro and wide area networks (MANs/WANs) calls for a mix of skills
and technologies that challenge even the most adept network manager.
Complex schemes are required, involving multiple data conversion steps,
that complicate and slow down transmission. The LAN’s native Ethernet
frames are translated to a variety of protocols and mapped onto point-to-
point circuits using time division multiplexing (TDM), frame relay,
ATM, and other techniques.
These circuits are the key bottlenecks in today’s network, often slowing
data transmission to one tenth the speed of the LANs that they intercon-
nect. The conversion and mapping process inevitably leads to higher
expense and lower reliability. The service provider faces the overwhelm-
ing task of provisioning the network’s multiple systems in a timely fash-
ion. Service velocity suffers as weeks or even months go by between the
order and the delivery of a new circuit.
Router-based networks make the problem worse since they require carri-
ers to deploy additional, complex routing devices to scale Ethernet over
the fibre infrastructure. Highly qualified IP engineers are required to roll
out and manage these services on a wide scale. Initial customers are eas-
ily provisioned, but adding new customers involves significant planning
and effort because the routed network must be engineered to scale and
maintain optimal utilization levels. Most importantly, these networks
must be engineered to convert complex addressing schemes and to avoid
the problem of address overlap. Security and customer separation need
to be carefully addressed.
Service providers need a solution that leverages their investment in their
infrastructure and operations staff. Today, that proven infrastructure is
the synchronous transmission network (SDH / SONET). Its advantages
in terms of scalability, manageability, bandwidth capacity, reliability,
and reach are well known. However, since the SDH / SONET network is
circuit-based, traditional approaches to transporting data over SDH /
SONET have been cumbersome to provision. The ultimate solution is a
highly efficient metro network that can leverage the existing SDH /
11 Evolving SDH to support Data-centric and Wavelength Services
180
Synchronous Transmission Systems
179 Evolving SDH to support Data-centric and Wavelength Services
It optimizes the use of metro core and access fibre and existing
SDH facilities
It solves any distance limitations of DWDM and can now trav-
erse SDH facilities over vast metro core or long haul networks
It provides full service assurance and SLA enforcement as pro-
vided by SDH today
No special engineering is needed as the core of the network fol-
lows the same engineering as SDH today. Services can also be
turned up easily since end-to-end provisioning is becoming
available for SDH-based core bandwidth
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 180
Resilient Packet Ring ( RPR ) technology
The metro bottleneck
Today, interconnecting enterprise local area networks (LANs) across
metro and wide area networks (MANs/WANs) calls for a mix of skills
and technologies that challenge even the most adept network manager.
Complex schemes are required, involving multiple data conversion steps,
that complicate and slow down transmission. The LAN’s native Ethernet
frames are translated to a variety of protocols and mapped onto point-to-
point circuits using time division multiplexing (TDM), frame relay,
ATM, and other techniques.
These circuits are the key bottlenecks in today’s network, often slowing
data transmission to one tenth the speed of the LANs that they intercon-
nect. The conversion and mapping process inevitably leads to higher
expense and lower reliability. The service provider faces the overwhelm-
ing task of provisioning the network’s multiple systems in a timely fash-
ion. Service velocity suffers as weeks or even months go by between the
order and the delivery of a new circuit.
Router-based networks make the problem worse since they require carri-
ers to deploy additional, complex routing devices to scale Ethernet over
the fibre infrastructure. Highly qualified IP engineers are required to roll
out and manage these services on a wide scale. Initial customers are eas-
ily provisioned, but adding new customers involves significant planning
and effort because the routed network must be engineered to scale and
maintain optimal utilization levels. Most importantly, these networks
must be engineered to convert complex addressing schemes and to avoid
the problem of address overlap. Security and customer separation need
to be carefully addressed.
Service providers need a solution that leverages their investment in their
infrastructure and operations staff. Today, that proven infrastructure is
the synchronous transmission network (SDH / SONET). Its advantages
in terms of scalability, manageability, bandwidth capacity, reliability,
and reach are well known. However, since the SDH / SONET network is
circuit-based, traditional approaches to transporting data over SDH /
SONET have been cumbersome to provision. The ultimate solution is a
highly efficient metro network that can leverage the existing SDH /
11 Evolving SDH to support Data-centric and Wavelength Services
181
Synchronous Transmission Systems
181 Evolving SDH to support Data-centric and Wavelength Services
SONET-based optical infrastructure, while taking advantage of LAN
data networks.
Resilient Packet Ring (RPR) technology is a solution that addresses the
needs of data-centric carriers and enterprises. This high-performance
solution breaks through the bottleneck in the metro while delivering the
key features of simplicity, speed, and reliability.
Introducing Resilient Packet Ring (RPR)
Resilient Packet Ring (RPR) technology addresses the growing need to
explode the bandwidth bottleneck found in today’s metro access net-
works. Efficient use of bandwidth and reliability of transmission are tar-
geted at ring topologies deployed in metro areas worldwide. A resilient
packet ring combines the low cost and simplicity of packet-based, con-
nectionless networking with the reliability, bandwidth, and scalability of
optical networks. The result is the best of both worlds - a resilient,
packet-oriented, ring-based solution that provides virtual mesh network
connectivity.
Figure 11-7
The Nortel Networks OPTera Packet Edge System makes use of the con-
cept of a LAN switch extending across the MAN and WAN. The nodes on the RPR ring act as a distributed Ethernet learning bridge, effectively
connecting buildings in a MAN as if they were simply segments on a
LAN.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 182
RPR shares bandwidth among the sites on the optical transport ring to
provide logical connectivity and optimal bandwidth sharing. Built on the
proven SDH / SONET infrastructure already available in virtually all
metro fibre rings, introducing RPR allows some or all of the ring’s
bandwidth to be used as shared bandwidth among many sources of
packet data. The result is a dramatic improvement in bandwidth effi-
ciency and simplified provisioning of service.
A Layer 2 encapsulation header is used to segregate customer traffic and
to switch it to the multiple egress points across the service provider net-
work. There is no need to manage the IP addressing or the Q-tagging
VLAN schemes used by the enterprise. The Nortel Networks encapsula-
tion technique scales to support millions of unique customer VLANs.
Nortel Networks OPTera Packet Edge System effectively transports
these packets within a distributed Ethernet learning bridge across the
metro.
Scalability
The Nortel Networks RPR implementation leverages the existing SDH /
SONET transport architecture - transparently. Traffic from multiple cus-
tomers is packed into standard SONET STS-Nc envelopes, in Resilient
Packet Rings, around the physical ring. These STS-Nc payload enve-
lopes can be sent through any intervening SDH / SONET
Add/Drop/Multiplex (MUX) or DWDM transport system for transmis-
sion over any distance (across town, the continent, or the world). At the
far end, a terminating OPTera Packet Edge System module delivers the
packets to their destination ports. As a result, the SDH / SONET trans-
port system becomes a distributed backplane where packets can find
their way to the appropriate destinations by means of standard Layer 2
techniques and without the complexities of overlay networks. Since mul-
tiple OPTera Packet Edge System modules can be put on a virtual ring,
they provide an efficient multicast technology that “drops” copies of
packets at multiple destinations.
Efficiency
Bandwidth efficiency is achieved through the following mechanisms that
are available with RPR technology:
11 Evolving SDH to support Data-centric and Wavelength Services
182
Synchronous Transmission Systems
181 Evolving SDH to support Data-centric and Wavelength Services
SONET-based optical infrastructure, while taking advantage of LAN
data networks.
Resilient Packet Ring (RPR) technology is a solution that addresses the
needs of data-centric carriers and enterprises. This high-performance
solution breaks through the bottleneck in the metro while delivering the
key features of simplicity, speed, and reliability.
Introducing Resilient Packet Ring (RPR)
Resilient Packet Ring (RPR) technology addresses the growing need to
explode the bandwidth bottleneck found in today’s metro access net-
works. Efficient use of bandwidth and reliability of transmission are tar-
geted at ring topologies deployed in metro areas worldwide. A resilient
packet ring combines the low cost and simplicity of packet-based, con-
nectionless networking with the reliability, bandwidth, and scalability of
optical networks. The result is the best of both worlds - a resilient,
packet-oriented, ring-based solution that provides virtual mesh network
connectivity.
Figure 11-7
The Nortel Networks OPTera Packet Edge System makes use of the con-
cept of a LAN switch extending across the MAN and WAN. The nodes on the RPR ring act as a distributed Ethernet learning bridge, effectively
connecting buildings in a MAN as if they were simply segments on a
LAN.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 182
RPR shares bandwidth among the sites on the optical transport ring to
provide logical connectivity and optimal bandwidth sharing. Built on the
proven SDH / SONET infrastructure already available in virtually all
metro fibre rings, introducing RPR allows some or all of the ring’s
bandwidth to be used as shared bandwidth among many sources of
packet data. The result is a dramatic improvement in bandwidth effi-
ciency and simplified provisioning of service.
A Layer 2 encapsulation header is used to segregate customer traffic and
to switch it to the multiple egress points across the service provider net-
work. There is no need to manage the IP addressing or the Q-tagging
VLAN schemes used by the enterprise. The Nortel Networks encapsula-
tion technique scales to support millions of unique customer VLANs.
Nortel Networks OPTera Packet Edge System effectively transports
these packets within a distributed Ethernet learning bridge across the
metro.
Scalability
The Nortel Networks RPR implementation leverages the existing SDH /
SONET transport architecture - transparently. Traffic from multiple cus-
tomers is packed into standard SONET STS-Nc envelopes, in Resilient
Packet Rings, around the physical ring. These STS-Nc payload enve-
lopes can be sent through any intervening SDH / SONET
Add/Drop/Multiplex (MUX) or DWDM transport system for transmis-
sion over any distance (across town, the continent, or the world). At the
far end, a terminating OPTera Packet Edge System module delivers the
packets to their destination ports. As a result, the SDH / SONET trans-
port system becomes a distributed backplane where packets can find
their way to the appropriate destinations by means of standard Layer 2
techniques and without the complexities of overlay networks. Since mul-
tiple OPTera Packet Edge System modules can be put on a virtual ring,
they provide an efficient multicast technology that “drops” copies of
packets at multiple destinations.
Efficiency
Bandwidth efficiency is achieved through the following mechanisms that
are available with RPR technology:
11 Evolving SDH to support Data-centric and Wavelength Services
183
Synchronous Transmission Systems
183 Evolving SDH to support Data-centric and Wavelength Services
Statistical Multiplexing: Since packet traffic tends to be
“bursty” in nature, the shared bandwidth of RPR reduces the
wasted bandwidth of circuit-based connections whose fixed
bandwidth is “consumed” even in periods of low-packet traffic.
Spatial Reuse: RPR frames traverse the shortest path between
communicating nodes on the fibre ring. Bandwidth is consumed
only on those link segments that interconnect those nodes, not
around the entire ring. Spatial reuse is another form of statisti-
cal gain that increases with the number of nodes in a ring.
Ring Protection: Since protection is provided using Layer 2
mechanisms, the reservation of 50 percent of the ring’s band-
width for protection is no longer required. Depending on the
types of traffic offered, this reserved protection bandwidth can
be reduced for other customer data traffic.
RPR provides bandwidth efficiency through statistical multiplexing of
bursty packet traffic, through spatial reuse around the ring, and by using
Layer 2 ring protection. The result is more capacity available for revenue
generating packet services.
OPTera Packet Edge Solution
The Nortel Networks OPTera Packet Edge System brings data network-
ing to a new level by taking the simplicity of LAN Ethernet natively into
the metro. Plug-in modules transform the Nortel Networks Metro 3300 /
3400 / 3500 and OME6500 Multiservice Platforms into hybrid systems
that are capable of concurrently supporting high speed data services,
voice traffic, and other traditional TDM (circuit-based) service offerings.
The RPR technology integrated into the OPTera Packet Edge System
uses an advanced ring architecture that provides native Ethernet inter-
faces, distributed Layer 2 packet switching, and data transport over
shared bandwidth. The solutions enabled by this architecture include
Internet access, Virtual Private Ethernet, data centre connectivity, and
outsourced IP applications. To the service provider, these solutions
translate into new opportunities for revenue-generating, value-added
services. To the enterprise, economical and scalable bandwidth with low
end-to-end delay enables a host of applications never before possible.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 184
With the advantage of the OPTera Packet Edge System, carriers and ser-
vice providers can deploy and upgrade a single network infrastructure
that delivers a wide range of revenue generating, resource-efficient data
and voice services.
Standards Evolution
The IEEE 802.17 Working Group brings together the leading technical
experts in the field of networking to generate a standard specification for
the RPR MAC and PHY(s). Nortel Networks has been a leader in this
standards effort since its original “call-for-interest” session, held in early
2000. John Hawkins, Marketing Manager of Optical Ethernet for Nortel
Networks, is the current chairman of the Resilient Packet Ring Alliance,
an industry advocacy group promoting RPR technology and its stan-
dardization by IEEE.
11 Evolving SDH to support Data-centric and Wavelength Services
184
Synchronous Transmission Systems
183 Evolving SDH to support Data-centric and Wavelength Services
Statistical Multiplexing: Since packet traffic tends to be
“bursty” in nature, the shared bandwidth of RPR reduces the
wasted bandwidth of circuit-based connections whose fixed
bandwidth is “consumed” even in periods of low-packet traffic.
Spatial Reuse: RPR frames traverse the shortest path between
communicating nodes on the fibre ring. Bandwidth is consumed
only on those link segments that interconnect those nodes, not
around the entire ring. Spatial reuse is another form of statisti-
cal gain that increases with the number of nodes in a ring.
Ring Protection: Since protection is provided using Layer 2
mechanisms, the reservation of 50 percent of the ring’s band-
width for protection is no longer required. Depending on the
types of traffic offered, this reserved protection bandwidth can
be reduced for other customer data traffic.
RPR provides bandwidth efficiency through statistical multiplexing of
bursty packet traffic, through spatial reuse around the ring, and by using
Layer 2 ring protection. The result is more capacity available for revenue
generating packet services.
OPTera Packet Edge Solution
The Nortel Networks OPTera Packet Edge System brings data network-
ing to a new level by taking the simplicity of LAN Ethernet natively into
the metro. Plug-in modules transform the Nortel Networks Metro 3300 /
3400 / 3500 and OME6500 Multiservice Platforms into hybrid systems
that are capable of concurrently supporting high speed data services,
voice traffic, and other traditional TDM (circuit-based) service offerings.
The RPR technology integrated into the OPTera Packet Edge System
uses an advanced ring architecture that provides native Ethernet inter-
faces, distributed Layer 2 packet switching, and data transport over
shared bandwidth. The solutions enabled by this architecture include
Internet access, Virtual Private Ethernet, data centre connectivity, and
outsourced IP applications. To the service provider, these solutions
translate into new opportunities for revenue-generating, value-added
services. To the enterprise, economical and scalable bandwidth with low
end-to-end delay enables a host of applications never before possible.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 184
With the advantage of the OPTera Packet Edge System, carriers and ser-
vice providers can deploy and upgrade a single network infrastructure
that delivers a wide range of revenue generating, resource-efficient data
and voice services.
Standards Evolution
The IEEE 802.17 Working Group brings together the leading technical
experts in the field of networking to generate a standard specification for
the RPR MAC and PHY(s). Nortel Networks has been a leader in this
standards effort since its original “call-for-interest” session, held in early
2000. John Hawkins, Marketing Manager of Optical Ethernet for Nortel
Networks, is the current chairman of the Resilient Packet Ring Alliance,
an industry advocacy group promoting RPR technology and its stan-
dardization by IEEE.
11 Evolving SDH to support Data-centric and Wavelength Services
185
Synchronous Transmission Systems
183 Evolving SDH to support Data-centric and Wavelength Services
Statistical Multiplexing: Since packet traffic tends to be
“bursty” in nature, the shared bandwidth of RPR reduces the
wasted bandwidth of circuit-based connections whose fixed
bandwidth is “consumed” even in periods of low-packet traffic.
Spatial Reuse: RPR frames traverse the shortest path between
communicating nodes on the fibre ring. Bandwidth is consumed
only on those link segments that interconnect those nodes, not
around the entire ring. Spatial reuse is another form of statisti-
cal gain that increases with the number of nodes in a ring.
Ring Protection: Since protection is provided using Layer 2
mechanisms, the reservation of 50 percent of the ring’s band-
width for protection is no longer required. Depending on the
types of traffic offered, this reserved protection bandwidth can
be reduced for other customer data traffic.
RPR provides bandwidth efficiency through statistical multiplexing of
bursty packet traffic, through spatial reuse around the ring, and by using
Layer 2 ring protection. The result is more capacity available for revenue
generating packet services.
OPTera Packet Edge Solution
The Nortel Networks OPTera Packet Edge System brings data network-
ing to a new level by taking the simplicity of LAN Ethernet natively into
the metro. Plug-in modules transform the Nortel Networks Metro 3300 /
3400 / 3500 and OME6500 Multiservice Platforms into hybrid systems
that are capable of concurrently supporting high speed data services,
voice traffic, and other traditional TDM (circuit-based) service offerings.
The RPR technology integrated into the OPTera Packet Edge System
uses an advanced ring architecture that provides native Ethernet inter-
faces, distributed Layer 2 packet switching, and data transport over
shared bandwidth. The solutions enabled by this architecture include
Internet access, Virtual Private Ethernet, data centre connectivity, and
outsourced IP applications. To the service provider, these solutions
translate into new opportunities for revenue-generating, value-added
services. To the enterprise, economical and scalable bandwidth with low
end-to-end delay enables a host of applications never before possible.
Synchronous Transmission Systems
Evolving SDH to support Data-centric and Wavelength Services 184
With the advantage of the OPTera Packet Edge System, carriers and ser-
vice providers can deploy and upgrade a single network infrastructure
that delivers a wide range of revenue generating, resource-efficient data
and voice services.
Standards Evolution
The IEEE 802.17 Working Group brings together the leading technical
experts in the field of networking to generate a standard specification for
the RPR MAC and PHY(s). Nortel Networks has been a leader in this
standards effort since its original “call-for-interest” session, held in early
2000. John Hawkins, Marketing Manager of Optical Ethernet for Nortel
Networks, is the current chairman of the Resilient Packet Ring Alliance,
an industry advocacy group promoting RPR technology and its stan-
dardization by IEEE.
List of Abbreviations List of Abbreviations
Synchronous Transmission Systems
List of Abbreviations
ADM Add Drop Multiplexer.
ANSI American National Standards Institute
APS Automatic Protection Switching
ATM Asynchronous Transfer Mode
AU Administrative Unit
AUG Adminstrative Unit Group
BML Business Management Level
CCITT Consultative Committee on International Telegraphy andTelephony (now the ITU-T)
Ch Channel
DCN Data Communications Network
DPRing Dedicated Protection Ring
DSC Digital Switching Centre
D-WDM Dense Wavelength Division Multiplexer
DXC Digital Cross-connect
EM Element Manager
EML Element Manager Layer
ETSI European Telecommunications Standards Institute
FDM Frequency Division Multiplexer
GPS Global Positioning System
GSC Group Switching Centre
HCI Human Computer Interface
HDSL High-speed Digital Subscriber Line
HO High Order
IP Internet Protocol
ISDN Integrated Services Digital Network
ISO International Organisation for Standardisation
ITU-T International Telecommunications Union TelecommunicationsStandardisation Sector
LAN Local Area Network
LE Local Exchange
LO Low Order
MSC Main Switching Centre
MSOH Multiplex Section Overhead
MSP Multiplex Section Protection
MS-SPRing Multiplex Section Shared Protection Ring
MUX Multiplexer
12
185 18612 List of Abbreviations
187 186 187 186List of Abbreviations List of Abbreviations
Synchronous Transmission Systems
List of Abbreviations
ADM Add Drop Multiplexer.
ANSI American National Standards Institute
APS Automatic Protection Switching
ATM Asynchronous Transfer Mode
AU Administrative Unit
AUG Adminstrative Unit Group
BML Business Management Level
CCITT Consultative Committee on International Telegraphy andTelephony (now the ITU-T)
Ch Channel
DCN Data Communications Network
DPRing Dedicated Protection Ring
DSC Digital Switching Centre
D-WDM Dense Wavelength Division Multiplexer
DXC Digital Cross-connect
EM Element Manager
EML Element Manager Layer
ETSI European Telecommunications Standards Institute
FDM Frequency Division Multiplexer
GPS Global Positioning System
GSC Group Switching Centre
HCI Human Computer Interface
HDSL High-speed Digital Subscriber Line
HO High Order
IP Internet Protocol
ISDN Integrated Services Digital Network
ISO International Organisation for Standardisation
ITU-T International Telecommunications Union TelecommunicationsStandardisation Sector
LAN Local Area Network
LE Local Exchange
LO Low Order
MSC Main Switching Centre
MSOH Multiplex Section Overhead
MSP Multiplex Section Protection
MS-SPRing Multiplex Section Shared Protection Ring
MUX Multiplexer
12
185 18612 List of Abbreviations
187 186 187 186List of Abbreviations List of Abbreviations
Synchronous Transmission Systems Synchronous Transmission Systems
NE Network Element
NEL Network Element Layer
NM Network Manager
NML Network Manager Layer
OA&M Operation, Administration and Maintenance
OAM&P Operation, Administration, Maintenance and Provisioning
OC Optical Carrier
OSC Optical Supervisory Channel
OSI Open Systems Interconnect
OSS Operation Support System
PCM Pulse Code Modulation
PDH Plesiochronous Digital Hierarchy
POH Path Overhead
POTS Plain Old Telephone Service
PSTN Public Switched Telephone Network
PTT Post, Telephones and Telegraph - former name for government-controlled public telecommunications operator
RDI Remote Defect Indication
REI Remote Error Indication
RSOH Repeater Section Overhead
Rx Receive
SDH Synchronous Digital Hierarchy
SLA Service Level Agreement
SML Service Management Layer
SNCP Subnetwork Connection Protection
SOH Section Overhead
SONET Synchronous Optical Network
SPRing Shared Protection Ring
SSMB Synchronisation Status Message Byte
SSU Synchronisation Supply Unit
STM Synchronous Transport Module
STS Synchronous Transport Signal
TDM Time Division Multiplex
TMN Telecommunications Managed Netork
TNC Transit Node Clock
TU Tributary Unit
TUG Tributary Unit Group
Tx Transmit
VC Virtual Container
WDM Wavelength Division Multiplexer
187 18812 List of Abbreviations
188List of Abbreviations List of Abbreviations
Synchronous Transmission Systems Synchronous Transmission Systems
NE Network Element
NEL Network Element Layer
NM Network Manager
NML Network Manager Layer
OA&M Operation, Administration and Maintenance
OAM&P Operation, Administration, Maintenance and Provisioning
OC Optical Carrier
OSC Optical Supervisory Channel
OSI Open Systems Interconnect
OSS Operation Support System
PCM Pulse Code Modulation
PDH Plesiochronous Digital Hierarchy
POH Path Overhead
POTS Plain Old Telephone Service
PSTN Public Switched Telephone Network
PTT Post, Telephones and Telegraph - former name for government-controlled public telecommunications operator
RDI Remote Defect Indication
REI Remote Error Indication
RSOH Repeater Section Overhead
Rx Receive
SDH Synchronous Digital Hierarchy
SLA Service Level Agreement
SML Service Management Layer
SNCP Subnetwork Connection Protection
SOH Section Overhead
SONET Synchronous Optical Network
SPRing Shared Protection Ring
SSMB Synchronisation Status Message Byte
SSU Synchronisation Supply Unit
STM Synchronous Transport Module
STS Synchronous Transport Signal
TDM Time Division Multiplex
TMN Telecommunications Managed Netork
TNC Transit Node Clock
TU Tributary Unit
TUG Tributary Unit Group
Tx Transmit
VC Virtual Container
WDM Wavelength Division Multiplexer
187 18812 List of Abbreviations
MF 190
1KH
A-0
0891
-SEN
50
0 - 1
1.07
- Pr
inte
d in
Sw
itzer
land
© A
BB
Sw
itzer
land
Ltd
, Nov
embe
r 200
7. T
he ri
ght t
o m
odifi
catio
ns o
r dev
iatio
ns d
ue to
tech
nica
l pro
gres
s is r
eser
ved.
top related