Chapter 6

1

Chapter 6

Wide Area Networking

2

Wide area networks (WANs)

Provide data communications that traverse a broad geographic area.

WANs typically utilize the transmission infrastructure provided by a third party such as a telephone company.

WAN technologies function at the lowest three layers of the OSI reference model.

WANs consist of physical media and connectors at the physical layer, physical addressing and media access at the data-link layer, and routing at the network layer.

3

Point-to-point WANs

A point-to-point connection in a WAN is not a typical direct connection. Rather, it’s a pre-established path from one site to another that passes through a carrier network.

A point-to-point connection, including wiring and hardware, is usually rented from a carrier and thus called a leased line.

4

Switching Methods

Circuit switching Packet switching Cell switching

5

Circuit Switching

Circuit-switched networks use temporary paths created through the network along which to transmit data.

There are two types of virtual circuits: permanent virtual circuits (PVC) switched virtual circuits (SVC)

6

Circuit Switching

Permanent virtual circuits (PVC) dedicated path between two points used when data

communications must take place 24 hours a day, 7 days a week. A PVC is similar to a telephone call that never ends. (more expensive option)

Switched virtual circuits (SVC) SVCs are established on demand and ended when

the communication is complete. SVCs are often called a dialup service because the SVC is established just like a telephone call and terminated in the same way. SVCs are used when the need to transfer data is sporadic. (circuit must be established & terminated each time data is sent, requiring some overhead)

7

Packet Switching

The carrier creates virtual circuits, a system that specifies a path between specific sites, within its network.

The path is not dedicated, however, and data can flow across different routers but still arrive at the same destination. FLASH VIDEO

8

Packet Switching

Each packet is transmitted in a store-and-forward process. When a router receives a packet, it stores the

packet temporarily. After reviewing the packet’s address information, the router forwards the packet.

The time that it takes to store the packet causes a slight latency in the transmission.

9

Packet Switching

Because packet-switched protocols provide for error checking and flow control, packet switching is very efficient.

The virtual circuit minimizes the connection time between any two systems, reducing the load on the network.

Because virtual circuits are closed after each packet is forwarded, a router is available to accept information from a router for a different virtual circuit. Voice over IP (VoIP) phone calls use this

10

Cell Switching

Cell switching is a form of packet switching. The main difference between a packet-

switched network and a cell-switched network is the size of the cell.

Cells are extremely small and do not vary in size. Their size makes them fast and provides for a network with a low latency.

11

Cell Switching An example of a cell-switched network is

Asynchronous Transfer Mode (ATM). The cell in an ATM network is 53 bytes in length,

including the data portion. Because a cell does not vary in size, each router in the

cell-switched network knows how much data to expect with each cell and is built to take advantage of it.

The tiny cell is small enough to be stored in random access memory, whereas a packet-switching router must store a packet to disk.

Because the router need only switch the cell in and out of its fastest memory, there is little latency in a cell-switched network.

12

ISDN

developed as the digital offering by the telephone companies to run over the existing telephone copper wiring.

ISDN allows subscribers to transmit data, voice, and multimedia.

13

ISDN

Basic Rate Interface (BRI): Used for home and small office connectivity.

BRI services include two B channels and a single D channel.

A B channel offers 64 Kbps and carries user data. A BRI D channel operates at 16 Kbps and carries control and signaling information.

Through these two channels, a home connection can reach 128 Kbps of data throughput.

14

ISDN

Primary Rate Interface (PRI): Used for WANs and runs across leased lines.

The PRI service is composed of 23 B channels at 64 Kbps each for user data, along with a single D channel, also operating at 64 Kbps to handle control information.

Overall, the PRI service provides a throughput rate of 1.544 Mbps.

15

ISDN

A computer with an ISDN line is able to connect to any other computer that also uses ISDN simply by dialing its ISDN number.

Terminal adapter (TA): Also called an ISDN modem, this is either an internal or external adapter to connect equipment to an ISDN line. Question: Why is this device not really a

MODEM?

16

Using Fiber Optics with FDDI

Light signals are capable of traveling long distances, and fiber optics aren’t subject to either electromagnetic interference or radio frequency interference.

Fiber Distributed Data Interface (FDDI) is the American National Standards Institute (ANSI) specification for a 100-Mbps token-passing dual ring network over fiber optic cable and is often used as a backbone for a campus network because it can attain distances up to two kilometers with multimode fiber or farther with single-mode fiber.

17

Using Fiber Optics with FDDI

Both multimode and single mode fiber optics transmit light signals. Multimode fiber uses a light emitting diode (LED)

to transmit the signals. Single mode fiber depends on lasers.

18

Using Fiber Optics with FDDI

FDDI’s dual ring structure enables traffic to flow on each ring, but in opposite directions one runs clockwise and the other counterclockwise.

One of the rings is considered primary, the other secondary.

The primary ring is used for data transmission. Because fiber optic cables are brittle, the secondary

ring acts as a backup in case of a break in the primary ring.

19

Using Fiber Optics with FDDI

20

Using Fiber Optics with FDDI

FDDI is a physical and data-link layer protocol.

Upper-layer protocols such as TCP/IP and IPX/SPX can run across a FDDI ring.

The FDDI frame is similar to a token ring frame format. There are two frames: the token frame and the data frame.

The FDDI data frame can become as large as 4500 bytes in length.

21

Gaining WAN Speed with ATM

WANs are typically slow. They are transmitted across long distances over networks that traditionally were unreliable.

As a result, the oldest WAN technologies were loaded with failsafe measures, such as error checking, and were clocked to match the slowest expected link.

Demand for greater data throughput was one of the driving forces behind Asynchronous Transfer Mode (ATM).

22

Gaining WAN Speed with ATM

ATM is a network made up of a series of ATM switches (also called ATM routers) and ATM nodes.

The ATM nodes can be routers that connect the ATM network to another type of network, such as an Ethernet LAN.

Any ATM node can communicate with any other ATM node by transmitting data across the ATM network.

23

Gaining WAN Speed with ATM

24

Gaining WAN Speed with ATM

ATM is a connection-oriented protocol in which the two ATM end points must establish a connection before transferring data.

Each ATM cell consists of 53 total bytes, including the data. The header is 5 bytes in length, with 48 bytes of data.

25

The ATM Model

Three layers in the ATM networking model map roughly to the physical and data-link layers of the OSI reference model: ATM physical layer ATM layer ATM adaptation layer

26

ATM physical layer

The ATM physical layer is concerned with transmission of bit-stream data on the physical media.

This layer is divided into logical sublayers. The physical medium sublayer commands the sending and

receiving of the bit stream and uses timing information to synchronize the transmitted data.

The physical medium sublayer is dependent on the type of cabling used. ATM can transmit across SONET and T3/E3, fiber optics, shielded twisted pair (STP), and unshielded twisted pair (UTP).

27

The ATM layer

Creates the virtual connection between the sending and receiving node and then switches the ATM cells through the virtual path.

When two virtual connections temporarily share the same path, the ATM layer will multiplex the cells to be able to transmit them as a single bit stream, and then demultiplex the cells when the virtual connections split off into two different directions.

28

The ATM layer

29

The ATM adaptation layer

Receives the packets from upper-layer processes and translates them into ATM cells.

These upper-layer packets can come from AppleTalk, TCP/IP, or IPX/SPX.

30

Frame Relay

Widely applied WAN protocol that uses packet switching.

Originally designed to run across ISDN interfaces.

The frame relay protocol provides an efficient data transmission, even though the packets vary in length.

The variable length of the packets makes data transmission very flexible.

31

Frame Relay

Even though frame relay was designed to transmit across ISDN, it has since been updated to transmit data across a variety of different protocols.

Most of the data transmission takes place within the carrier’s network, also called a cloud, which provides for congestion signaling, physical media, and switching.

32

Frame Relay

Frame relay supplies data-link and physical layer services.

Both permanent virtual circuits (PVC) and switched virtual circuits (SVC) can be used. It is most common to find frame relay with PVCs.

33

Frame Relay

34

Frame Relay

Data Link Connection Identifiers (DLCI) are key to frame relay connections.

The DLCI identifies the point-to-point link that begins the virtual circuit to the cloud.

It is a logical connection, because a single interface can transmit to different DLCIs.

It’s easiest to think of a DLCI as a phone number. When transmitting data across frame relay, the interface into the cloud will specify the destination by its DLCI in the same way that a telephone call identifies the recipient by that recipient’s telephone number.

35

Frame Relay

Controlling Congestion Congestion is one problem that can occur in a

cloud network shared by multiple customers. Because the switches can participate in any number of virtual circuits, they can become overloaded by traffic.

36

Frame Relay The frame relay protocol includes 3 bits in the address of each

frame relay frame header to manage congestion. The first bit is called the Forward Explicit Congestion Notification

(FECN). When this bit is set to a one (1) value, it means that the frame encountered congestion. A router that receives a frame with the FECN bit set to 1 will send a reply frame with a different bit named the Backward Explicit Congestion Notification (BECN). Both the FECN and BECN bits are used to notify upper-layer protocols of congestion so that they can initiate flow-control mechanisms to reduce data transmissions.

The third congestion control bit, the Discard Eligibility (DE) bit, is assigned to unimportant frames. Frames with the DE bit set to 1 are discarded when there is congestion on the network. This is an additional method to help manage high-traffic situations within the frame relay cloud.

37

Frame Relay Given that congestion does exist in the frame relay

network, the carrier provides a guaranteed minimum data transmission speed: the committed information rate (CIR).

A benefit of frame relay is that bursts of speed up to double the CIR level are generally allowed by the carrier network, but only for short periods of time.

Bursting is available only during non-congested time periods when extra bandwidth is available.

38

Frame Relay Not only is frame relay flexible because of its ability to

burst extra traffic, but its data transmission rates are also dependent solely on configuration.

A frame relay link can be installed and configured at a lower speed (and at a lower cost) using the same equipment that is used for a much faster link.

If a link needs to be upgraded because of an increased need for bandwidth, or downgraded because bandwidth is no longer required, the carrier needs only to make a configuration change.

39

SONET/SDH The Synchronous Optical Network (SONET), or Synchronous

Digital Hierarchy (SDH) as it’s known in Europe, offers the ability to construct large-scale, high-speed IP networks over fiber optics.

The SONET topology can be either a dual ring architecture or a star. The dual ring is preferable because it can be reconfigured in case of a break in the fiber optic cable to ensure the network’s survivability.

SONET is often used for Internet and large internetwork backbone services.

Using time division multiplexing (TDM), SONET is capable of providing high-bandwidth capacity for data transmission as well as voice traffic and even cable television.

40

SONET/SDH

SONET is a global standard focusing on synchronous communications that are multiplexed.

In synchronous networking, all the clocks are synchronized to the same time.

The time division multiplexing enables signals from slower networks to be intermingled directly with SONET signals as they are moved onto the SONET network.

This is also partially because of the advanced network management and maintenance features inherent in SONET.

41

SONET/SDH

SONET’s basic transmission rate, Synchronous Transport Signal level 1 (STS-1), also considered Optical Carrier 1 (OC1), is 51.84 Mbps. The SONET multiplexing scheme can transmit at rates that are multiples of 51.84 Mbps.

42

SONET/SDH

43

T-Carrier System

• The T-carrier system is a series of data transmission formats developed by Bell Telephone for use in the telephone network system in North America and Japan.

• The base unit of a T-carrier is DS0, which is 64 Kbps. • The T-carrier system uses in-band signaling, a method

that actually robs bits from being used for data and uses them instead for overhead.

• This reduces the transmission rates used for T-carrier signals. The E-carrier system used in Europe doesn’t perform bit-robbing and as a result has a higher throughput rate.

44

T-Carrier System T1/E1

T1 and E1 lines are each multiples of DS0 signals. The T1 line provides 1.544 Mbps, while the E1 line provides 2.048 Mbps. The difference in data rates results from the T-carrier system’s method of bit-robbing.

Customers can purchase fractional-T1 lines, which are actually multiples of the DS0 signal. With Frac-T1, the customer rents a number of the 24 channels within a T1 line. The remaining channels go unused. For example, a Frac-T1 line can be 128 Kbps, 256 Kbps, 512 Kbps, and so on.

T3/E3 T3 lines are digital carriers, equivalent to 28 T1 lines, that

can transmit at the rate of 44.736 Mbps. E3 lines provide 16 E1 lines, with a transmission rate of 34.368 Mbps.