This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
The IP services market is still in its infancy but is increasingly recognised as a legitimate hope for the industry growth
• IP services are unlikely to provide a growth catalyst of the magnitude required to bring telecoms out of the current crisis
• The “converged network of the future” will be a hybrid network where IP and legacy services complement each other
• Shifts in the value chain and changing the dynamics of competition may cause vendors, service providers and systems integrators to focus on the same end-user attention
• With some exceptions (e.g. IP Storage), the technologies underpinning IP are relatively mature, and the industry is in a later stage of its life cycle:– Expect to see vendor consolidation and a shift to process innovation
– As in all other industries preceding it, the number of vendor/operators is likely to decrease over time, market leaders will consolidate their positions, and the nature of innovation will move from product to process innovation, the realm of the established vendor/operator.
• IP-based technologies have failed to live up to their promise due to the uncalculated/hidden costs associated with complexity (e.g device proliferation, unpredictable nature of traffic): – Hence, tech vendors offering products that solve complexity management issues are sure-fire winners.
– One of the few areas where we expect significant product innovation to occur and where we believe start-ups still have a reasonable chance of winning.
Overall uptake of IP services will likely be only moderate over the next 24 months
Explaining the ResultsExplaining the Results
• The IP services market is expected to grow from its current size of $14.5 billion to $17.5 billion in 2004/5, with a CAGR of 10%
• IP VPN is the foundation layer on which value-added services will be built
• VoIP will grow slowly, as it is not the optimal technology for scaling a voice business
• Video over IP and unified communications will experience high take up after 24 months
• Penetration of the IP storage market is low with huge growth potential
• The IP services market is expected to grow from its current size of $14.5 billion to $17.5 billion in 2004/5, with a CAGR of 10%
• IP VPN is the foundation layer on which value-added services will be built
• VoIP will grow slowly, as it is not the optimal technology for scaling a voice business
• Video over IP and unified communications will experience high take up after 24 months
• Penetration of the IP storage market is low with huge growth potential
Advisory
IP Services Growth/Margin MatrixIP Services Growth/Margin Matrix
$0.30bn
$4.08bn
$7.94bn
$1.36bn
$0.86bn
-10%
0%
10%
20%
30%
40%
50%
0% 10% 20% 30% 40% 50% 60%
Margin
CA
GR
(2
00
2-0
4)
Access & connectivity (incl. basic VPN)
Managed + Hosting/ Storage Services
NB: All managed and outsourced elements of IP services have been included with hosting and storage revenues to reflect similarity in delivery capabilities.
Converged Services
VPN features: Security/QoS
Players recognising IP-VPN (connectivity) as the foundation for upselling other services will likely be most successful.
Players recognising IP-VPN (connectivity) as the foundation for upselling other services will likely be most successful.
However, IP’s promise of better, cheaper and simpler services has not been fulfilled
Not BetterNot Better
• Of 50 companies interviewed, 30 cited quality and reliability as reasons for not using VoIP:a
– The benefits of VoIP are hard to justify while Quality of Service issues are not yet resolved
• 93% of IT Managers state security as a key barrier to deploying IP-VPNSb
• Of 50 companies interviewed, 30 cited quality and reliability as reasons for not using VoIP:a
– The benefits of VoIP are hard to justify while Quality of Service issues are not yet resolved
• 93% of IT Managers state security as a key barrier to deploying IP-VPNSb
Not CheaperNot Cheaper
• Equipment costs still high, payback uncertain:
– IP phones too expensive
– No cost savings until $100 each
– Traditional PBX cost is also lower when counting implementation costs for a new IP PBX
– IP PBX estimated 5-year payback may be too long
• Falling PSTN prices, have eroded VoIP’s price advantage– Sometimes referred to as Toll-bypass
• Equipment costs still high, payback uncertain:
– IP phones too expensive
– No cost savings until $100 each
– Traditional PBX cost is also lower when counting implementation costs for a new IP PBX
– IP PBX estimated 5-year payback may be too long
• Falling PSTN prices, have eroded VoIP’s price advantage– Sometimes referred to as Toll-bypass
Not SimplerNot Simpler
• Can be more complex: 60% of IT managers said ease of use/management issues were barrier to IP-VPN deploymentc
• Migration to new IP services has often resulted in requirement to operate simultaneous systems.
• Issues of inter-operability, billing and customer service have also emerged for service providers
• Can be more complex: 60% of IT managers said ease of use/management issues were barrier to IP-VPN deploymentc
• Migration to new IP services has often resulted in requirement to operate simultaneous systems.
• Issues of inter-operability, billing and customer service have also emerged for service providers
a. See IDC, “Attitudes towards IP Telephony in European Corporations”, October 2001. b. Internet Week Research, VPN usage survey based in US users (multiple responses accepted). c. Forrester, “Surviving A Metro Bandwidth Crunch”, December 2001.
Recent Academic Research is coming to the same conclusions
• From a paper presented at Hotnes, Princeton, October 2002: “Is IP going to take over the world (of communications)?”– Pablo MolineroFernandez, Nick McKeown, Stanford University; Hui Zhang, Turin Networks and Carnegie
Mellon University• "It remains ill suited as a means to provide many other types of service; and is too crude to form the transport
infrastructure in its own right."
• "The growth and success of IP has given rise to some widely held assumptions amongst researchers, the networking industry and the public at large. One common assumption is that it is only a matter of time before IP becomes the sole global communication infrastructure, dwarfing and eventu ally displacing existing communication infrastructures such as telephone, cable and TV networks."
• "But for all its strengths, we (the authors) do not believe that IP will displace existing networks; in fact, we believe that many of the assumptions discussed above are not supported by reality, and do not stand up to close scrutiny."
• "It is the goal of this paper to question the assumption that IP will be the network of the future. We will conclude that if we started over - with a clean slate - it is not clear that we would argue for a universal, packet-switched IP network."
• "We take the position that while IP will be the network layer of choice for best-effort, non-mission critical and non-real time data communications (such as information exchange and retrieval), it will live alongside other networks, such as circuit-switched networks, that are optimized for high revenue time-sensitive applications that demand timely delivery of data and guaranteed availability of service."
• "At the core of the network, we expect the circuit switched transport network to remain as a means to interconnect the packet switched routers, and as a means to provide high reliability, and performance guarantees. Over time, more and more optical technology will be introduced into the transport network, leading to capacities that electronic routers cannot achieve."
. . . But IP will dominate the market for smaller enterprise networks
• Small IP networks (such as Corporate LANs and WANs) are cheaper and simpler to operate than small circuit-switched networks:– Enterprise packet-switching is considerably simpler and cheaper to manage.
– The number of components in an all IP corporate LAN can be significantly smaller than a circuit- switched infrastructure with N^2 characteristics, especially in highly interactive corporates like banks, the pharmaceutical sectors, large engineering firms, and the military.
– The power/density/operational complexity characteristics of IP infrastructures in a core network. [NOT CLEAR]
• QoS solutions for corporate LANs and WANs are becoming more manageable with the advent of policy-based management technologies:– Both from a hardware/software and operational perspective.
• Service providers will also need to maintain packet-switched architectures to enable Internet access and other services:– E.g. managed services, such as storage and hosting, where most corporate interfaces will be IP.
– E.g. multi-party services, such as audio, video, and web-enabled conferencing, due to the N-squared complexities associated with trying to provide these services efficiently off a circuit-switched infrastructure.
From our conversations with buyers, systems integrators, and service providers, we are convinced the future of enterprise LANs and WANs will be all IP.
From our conversations with buyers, systems integrators, and service providers, we are convinced the future of enterprise LANs and WANs will be all IP.
However, there are significant advantages for service providers from hybrid trunk side architectures
Advantages for Service Providers
Internet offloadInternet offload Intercept dial-up calls handled by the IP infrastructure avoid clogging up Class-5 and Class-4 switchesIntercept dial-up calls handled by the IP infrastructure avoid clogging up Class-5 and Class-4 switches
Improve mesh performance
Improve mesh performance
Operational savingsOperational savings
Deliver new hybrid applications
Deliver new hybrid applications
Adding switching capacity has decreasing advantages as more ports get allocated to interconnect with other elements of the mesh rather than to handle new calls
Adding switching capacity has decreasing advantages as more ports get allocated to interconnect with other elements of the mesh rather than to handle new calls
By using, for example, an ATM cloud for the interconnect service providers can greatly reduce the number of trunks they need to manageBy using, for example, an ATM cloud for the interconnect service providers can greatly reduce the number of trunks they need to manage
Many of the applications that mix voice and data, which generally involve SIP manipulations, require IP anyway to interact with intelligent clientsMany of the applications that mix voice and data, which generally involve SIP manipulations, require IP anyway to interact with intelligent clients
• Obligation to deliver local number portability, emergency and directory services
• Web-based invitation, notification, scheduling and device control
• Unified network to allow unified and simplified management
New ArchitectureNew Architecture
• Interoperable technology is finally becoming available to enable VoIP over LANs and increasingly WANs– PBX interoperability is improving in brownfields
• In next-generation carrier networks:– Voice traffic between traditional voice networks and new packet-
based networks will be directed by media gateways and media gateway controllers, which will be handled by soft-switches
• There are two key concepts behind these new networks:– Media gateways, signalling gateways, media gateway controllers
and application servers will be divided into separate logical network components
– These components will communicate with one another through the use of intra-switch protocols such as Media Gateway Control (MEGACO), Media Gateway Control Protocol (MGCP), SCTP/M3UA and Session Initiation Protocol (SIP)
• Interoperable technology is finally becoming available to enable VoIP over LANs and increasingly WANs– PBX interoperability is improving in brownfields
• In next-generation carrier networks:– Voice traffic between traditional voice networks and new packet-
based networks will be directed by media gateways and media gateway controllers, which will be handled by soft-switches
• There are two key concepts behind these new networks:– Media gateways, signalling gateways, media gateway controllers
and application servers will be divided into separate logical network components
– These components will communicate with one another through the use of intra-switch protocols such as Media Gateway Control (MEGACO), Media Gateway Control Protocol (MGCP), SCTP/M3UA and Session Initiation Protocol (SIP)
VoIP today is more a question of delivering a suite of hybrid voice/data services than replacing TDM networks.
VoIP today is more a question of delivering a suite of hybrid voice/data services than replacing TDM networks.
• Centrex (central office exchange service) is a service from local telephone companies in which up-to-date telephone facilities at the service provider’s central office are offered to business users who do not wish to purchase their own facilities.
• The Centrex service consists of allocating centralized capabilities among different business customers. The customer is spared the expense of having to keep up with fast-moving technology changes (for example, having to continually update their private branchexchange infrastructure) and the phone company has a new set of services to bill for.
• According to statistics compiled by RHK, the Centrex service is most popular with small to mid-sized companies with up to 400 lines. However, customers with over 1,000 lines represent over 20% of the existing Centrex base.
• These larger customers have been increasingly investing in their own infrastructure, as suggested by the following data showing erosion of 8% per year from 1996 to 2001. The rate of decline is expected to increase to 12% per year from 2001 to 2005.
• Meanwhile, smaller customers are expected to show only modest growth in the adoption of Centrex. Another source of risk with Centrex customers is that as they switch from outsourcing to owning their own facilities, service providers risk losing other precious sources of revenue such as Web hosting, VPN offerings and Unified Communications. According to Lucent, such losses could add up to $750,000 per customer per week, or approximately $39 m per year.
No appetite for forklift upgradesNo appetite for forklift upgrades Gradual migration in softwsitch architecture in multiprotocol world
Gradual migration in softwsitch architecture in multiprotocol world
Veraz networksVeraz networks
Integrating voice with other real-time apps
Integrating voice with other real-time apps
Multi-channel interface for application delivery
Multi-channel interface for application delivery
Iperia, Voyant, AspectIperia, Voyant, Aspect
Source: CGE&Y and DrKW Analysis.
Product InnovationProduct Innovation Scale packet based voice servicesScale packet based voice services Separate voice processing from application logic
Separate voice processing from application logic
IP Unity (media server) Sylantro, Pactolus (apps server)
IP Unity (media server) Sylantro, Pactolus (apps server)
Process InnovationProcess InnovationEnd-to-end solutions absentEnd-to-end solutions absent Control the value chain from the
user interface upControl the value chain from the user interface up
AvayaAvaya
We have identified the potential winners in the VoIP space...
100% VoIP for core voice transport is many years into the future
• IP is not the optimal technology for scaling a voice business:– QoS slippage is not an option
– Legacy infrastructure is cheap and very efficient
• The transition to VoIP will be slow as each new service introduced must be value accretive and protect the value of legacy networks
• Work on VoIP has led to developments in technology that enable delivery of real time services e.g. call centre automation and conferencing
• The market take-up of these services depends on the design, pricing and bundling of attractive end-user applications:– It is up to the carriers to design and deliver IP-enabled differentiated services that leverage existing TDM networks …
without falling back into a minutes / bits price war
Source: CGEY and DrKW Analysis
Conclusions on VoIP
The near-term significance of VoIP is in the real-time IP Services that it enables.The near-term significance of VoIP is in the real-time IP Services that it enables.
The future of real-time IP Services depends on successfully developing and deploying emerging QoS tools
Over-provisioningOver-provisioning
• Common over-provisioning is for typical bandwidth utilisation
• QoS over-provisioning is for peak utilisation
• May raise questions of affordability
• Common over-provisioning is for typical bandwidth utilisation
• QoS over-provisioning is for peak utilisation
• May raise questions of affordability
Guaranteed bandwidth allocations or resource
reservation
Guaranteed bandwidth allocations or resource
reservation
Traffic managementTraffic management
Application-driven QoSApplication-driven QoS
QoS Solution/Method Tools
• Make sure preferred “bursty” bandwidth is available to QoS-sensitive applications
• Determine preferred paths in network
• Make sure preferred “bursty” bandwidth is available to QoS-sensitive applications
• Determine preferred paths in network
• Admission control
• Data management: packet prioritisation, policing and shaping
• Compression
• Admission control
• Data management: packet prioritisation, policing and shaping
• Compression
• Prioritise by application, not packet type or port
• Measure and police network traffic in real time
• Content management & caching
• Prioritise by application, not packet type or port
• Measure and police network traffic in real time
• Content management & caching
• Increased edge/access infrastructure (servers, switches, routers and gateways)
• Increased edge/access infrastructure (servers, switches, routers and gateways)
• Implement circuit-switching with ATM
• Implement connection-oriented IP with MPLS
• Signalled resource reservation with RSVP
• Implement circuit-switching with ATM
• Implement connection-oriented IP with MPLS
• Signalled resource reservation with RSVP
• Provision-class packet marking with DiffServ
• Standardised shaping/queue reordering/etc.
• IPv6
• Provision-class packet marking with DiffServ
• Standardised shaping/queue reordering/etc.
• IPv6
• Implement centrally controlled traffic-management tools and drive policies down to network elements in the LAN
• Multicast & content-aware networking
• Implement centrally controlled traffic-management tools and drive policies down to network elements in the LAN
• Multicast & content-aware networking
Description
Delivering quality is an inter-domain traffic engineering problem encountered at both the interface of the WAN and LAN and between service provider networks.
Delivering quality is an inter-domain traffic engineering problem encountered at both the interface of the WAN and LAN and between service provider networks.