Novell Del Mar College
CIS 306 - Managing NOVELL® Networks
 Instructor: Michael P. Harris
Networking Primer

8. Important WAN & High-Speed Technologies

These days, if you pick up any computer networking magazine, you'll find that among the hot topics are the technologies that make networks faster and the technologies that connect geographically distant networks—technologies such as frame relay, 100VG-AnyLAN, and ATM. We mentioned these and other high-speed and WAN technologies under the "Commonly Used Standards" heading of this primer. Brief explanations will help you more fully understand what these technologies are and why they are important. The following technologies will be treated:

100Base-TX

100Base-TX is a high-speed LAN technology. 100Base-TX is officially designated as the IEEE 802.3u standard. It functions at the data-link layer's (OSI level two) media access control sublayer and provides data transfer rates as high as 100 megabits per second (Mbit/s).

Distinguishing Characteristics

Like 10Base-T Ethernet, 100Base-TX uses carrier sense multiple access with collision detection as the media access control method. (CSMA/CD was discussed earlier under the "Bus Topology" heading of this primer.) 100Base-TX is based on the scalability of CSMA/CD. Scalability means that you can easily enlarge or downsize your network without degrading network performance, reliability, and manageability.

CSMA/CD was known to be scalable before the 100Base-TX standard was created. A scaled-down version of Ethernet (1Base-5) uses CSMA/CD, provides data transfer rates of 1 Mbit/s, and enables longer transmission distances between repeaters. If CSMA/CD could be scaled down, then it could be scaled up. Specifying changes such as decreased transmission distances between repeaters produced a reliable data transfer rate of 100 Mbit/s, 10 times faster than traditional 10Base-T Ethernet.

100Base-TX supports Category 5 (100Base-T4 can use Category 3) unshielded twisted-pair (UTP) wiring, Type-1 shielded twisted-pair (STP) wiring, and fiber-optic cable (100Base-FX). 100Base-T4 uses four wire pairs of Category 3 UTP cable—three for data and one for collision detection. However, 100Base-TX uses only two wire pairs of Category 5 (CAT5) UTP cable.

Figure 28: On 100Base-TX networks, the physical topology is a star and the logical topology is a bus. A broadcast signal travels to all parts of the cable.

Advantages

100Base-TX is widely available. Adapter cards and compatible cable are currently available from various vendors.

In addition, it's easy to upgrade from 10Base-T Ethernet to 100Base-TX Ethernet if Category 5 cable was used. Both traditional 10Base-T and 100Base-TX Ethernet use CSMA/CD, and some network cards now support both 10 Mbit/s and 100 Mbit/s Ethernet. The adapter cards automatically sense whether it is a 10 Mbit/s or 100 Mbit/s environment and adjust their speed accordingly. Because 100Base-TX and 10Base-T Ethernet can coexist, network supervisors can upgrade network stations from 10Base-T to 100Base-TX one at a time, as needed. Also, most network supervisors are already familiar with CSMA/CD, so there is no need for expensive retraining.

100Base-TX can be an inexpensive way to make your network faster. Adapter cards are not significantly more expensive than 10Base-T cards. In addition, Category 5 UTP cable is relatively inexpensive and many organizations already have either Category 5 cable installed.

Disadvantages

100Base-TX will reduce the maximum network size compared to 10Base-T because the standard specifies shorter transmission distances between repeaters.

In addition, the fact that 100Base-TX is based on CSMA/CD creates problems. 100Base-TX may scale CSMA/CD to its limit, making 100 Mbit/s the maximum data transfer rate for this standard. To increase data transfer rates, 100Base-TX specifies shorter distances between signal repeaters, and these distances may be as short as is practical. Also, because CSMA/CD is a shared media contention scheme, collisions will occur, especially under maximum loads. This results in increased overhead, which reduces actual data throughput.

100Base-T4 requires four wire pairs of Category 3 cable, and not all companies that have Category 3 cable have four wire pairs available. Thus, companies that are already using some wire pairs for a different purpose, or that installed cable with fewer than four wire pairs or cable that does not meet Category 3 standards, will have to recable to use 100Base-TX. Those companies that cabled with Category 5 (4-pair CAT5) cable are already ready to upgrade to 100Base-TX.

100VG-AnyLAN

100VG-AnyLAN, which is officially designated as the IEEE 802.12 standard, is a high-speed LAN technology that competes with 100Base-TX. Like 100Base-TX, 100VG-AnyLAN functions at the data-link layer (OSI level two) and provides data transfer rates as high as 100 Mbit/s. However, 100VG-AnyLAN differs from 100Base-TX in several important respects.

Distinguishing Characteristics

Instead of using CSMA/CD as the media access control method, 100VG-AnyLAN uses a method called "demand priority." Demand priority is not a contention scheme; rather, it uses a form of token passing that assigns the token (permission to broadcast data packets) based on a priority scheme that the network supervisor establishes for different types of LAN traffic and on the order in which an intelligent hub receives requests.

For example, when a workstation needs to transmit, it signals the intelligent hub that it needs access to the transmission media. If the intelligent hub receives several requests, it will give access to the workstation that has the highest priority. (100VG-AnyLAN will also function without a prioritization scheme.) If the workstations requesting access have the same priority, the intelligent hub will assign the token to the workstations in the order they request access to the transmission media.

In addition, 100VG-AnyLAN supports both Ethernet and token-ring networks. It also supports Category 3 and 5 UTP, Type-1 STP, and fiber-optic cable. 100VG-AnyLAN uses four wire pairs of Category 3 or Category 5 UTP cable.

Figure 29: On 100VG-AnyLAN networks, both the physical and logical topologies are stars. The signal from one node goes to the intelligent hub and is routed only to the correct destination node.

Advantages

Because 100VG-AnyLAN uses demand priority, it has several advantages over 100Base-TX. First, 100VG-AnyLAN provides the necessary bandwidth and timing (low latency) needed by applications such as multimedia applications. Network supervisors can, if desired, assign higher priority to the hub ports used to connect workstations or servers that frequently transmit time-sensitive data, such as audio and video. Second, the demand priority media access method avoids the collisions that can occur on 100Base-TX networks, ensuring that control overhead will not soar. Third, 100VG-AnyLAN networks are switched (they do not broadcast packets to all workstations), which makes 100VG-AnyLAN networks more secure against eavesdroppers than 100Base-TX networks.

Furthermore, unlike 100Base-TX, 100VG-AnyLAN supports token-ring networks as well as Ethernet, providing data transfer rates as high as 100 Mbit/s to the former.

100VG-AnyLAN and 100Base-TX also share many advantages. The cost of 100VG-AnyLAN is comparable to 100Base-TX: Adapter cards that support both 10 and 100 Mbit/s are not priced significantly higher than traditional 10Base-T Ethernet cards. Both standards also support the same types of transmission media. In addition, both are easy to upgrade.

Disadvantages

Unlike 100Base-TX's CSMA/CD, which is familiar to many network supervisors, demand priority is new, and network supervisors will require some training to use it effectively. Also, 100VG-AnyLAN has a smaller market share than 100Base-TX. Consequently, it is not supported by as many vendors, which means that fewer products are available for 100VG-AnyLAN.

Previously installed cable may be problematic for 100VG-AnyLAN networks, as it is for 100Base-TX. 100VG-AnyLAN uses all four wire pairs of Category 3 or 5 UTP cable. Thus, companies that are already using some wire pairs for a different purpose, or that installed cable with less than four wire pairs or cable that does not meet Category 3 standards, will have to recable to use 100VG-AnyLAN.

Fiber Distributed Data Interface (FDDI)

Fiber Distributed Data Interface (FDDI) is also a high-speed LAN technology. It is not generally used for direct connection to desktop computers, but rather as a backbone technology. A backbone connects two or more LAN segments to provide a path for transmitting packets among them. A simple backbone might connect two servers through a high-speed link consisting of network adapter cards and cable.

FDDI is officially designated as ANSI X3T9.5 and operates at the physical and data-link layers (levels one and two) of the OSI model. Like 100Base-TX and 100VG-AnyLAN, FDDI provides data transfer rates as high as 100 Mbit/s.

Figure 30: A simple server-based backbone connecting two LAN segments

Distinguishing Characteristics

FDDI networks have a dual, counter-rotating ring topology. This topology consists of two logical closed signal paths called "rings." Signals on the rings travel in opposite directions from each other. Although both rings can carry data, the primary ring usually carries data while the secondary ring serves as a backup.

On FDDI networks, every node acts as a repeater. FDDI supports four kinds of nodes: dual-attached stations (DASs), single-attached stations (SASs), single-attached concentrators (SACs), and dual-attached concentrators (DACs). DASs and DACs attach to both rings; SASs and SACs attach only to the primary ring. Several SASs often attach to the primary ring through a concentrator so that an SAS failure will not bring down the entire network. If the cable is cut or a link between nodes fails, DASs or DACs on either side of the failure route signals around the failed segment using the secondary ring to keep the network functioning.

FDDI uses token passing for the media access control method and is implemented using fiber-optic cable.

Figure 31: If a cable section on an FDDI network goes down, DASs on either side of the failed section automatically reconnect the primary and secondary rings. Also note that the server has a redundant connection to improve reliability.

Advantages

FDDI is a fast, reliable standard. The dual, counter-rotating ring topology increases the network's reliability by keeping the network functioning even if a cable is damaged. FDDI also offers network management support, which was designed directly into the standard. Also, the standard includes the Copper Distributed Data Interface (CDDI) specification for building a network using UTP cable (which is less expensive than fiber-optic cable).

Disadvantages

FDDI's main disadvantage is price. FDDI adapter cards and fiber-optic cable are both relatively expensive compared to other technologies offering the same speed. Fiber-optic cable installation also requires more expert technicians. Even CDDI adapters (for copper wire), which are less expensive than FDDI adapters, are more expensive than either 100Base-TX or 100VG-AnyLAN adapters.

X.25

X.25 is a commonly used WAN standard at the network layer (level three) of the OSI model. It is a CCITT (now known as the International Telecommunications Union (ITU)) standard and includes data-link and physical layer protocols (LAP-B and X.21), as shown in Figure 17. X.25 provides data transfer rates of 9.6 kilobits per second (kbit/s) to 256 kbit/s, depending on the connection method.

Distinguishing Characteristics

X.25 specifies the interface for connecting computers on different networks by means of an intermediate connection through a packet-switched network (for example, CompuServe, Tymnet, or Telnet). X.25 was defined when the quality of transmission media was relatively poor. As a result, the standard specifies that each node in the packet-switched network must fully receive each packet and check it for errors before forwarding it.

Figure 32: X.25 networks are often provided by telecommunication carriers. CompuServe uses X.25 on its network.

Advantages

X.25 is well understood and reliable. Connections to X.25 networks can be made through the existing telephone system, ISDN, and leased lines. Because access is so simple, it is comparatively inexpensive. X.25 is also available worldwide. In countries with little digital telecommunications infrastructure, X.25 may be the best WAN technology available.

Disadvantages

Although it is widely available, X.25 is slow compared to newer technologies. The process of checking each packet for errors at each node limits data transfer rates. X.25 also uses variable-size packets, which can cause transmission delays at intermediate nodes. In addition, many people connect to X.25 networks through modems, which limit data transfer rates from 9.6 kbit/s to 56 kbit/s. Although X.25 is likely to remain in common use for some time, newer, faster standards are already replacing it.

Frame Relay

Frame relay, like X.25, is a WAN technology. Approved by ANSI and the ITU, frame relay works at the data-link layer (level two) of the OSI model, providing data transfer rates from 56 kbit/s to 1.544 Mbit/s.

Distinguishing Characteristics

Frame relay is an interface specification for connecting LANs over public packet-switched networks. This standard can be thought of as a simplified version of X.25 designed to take advantage of digital transmission media.

Frame relay services are typically provided by telecommunications carriers. Customers install a router and lease a line (often a T1 or fractional T1 line) to provide a permanent connection from the customer's site to the telecommunications carrier's network. This connection enables frame relay to use permanent virtual circuits (PVCs), which are predefined network paths between two locations.

With frame relay, the router encapsulates (or frames) network layer packets, such as IP and IPX packets, directly into a data-link level protocol and sends them on to the packet-switched network. Like X.25, frame relay uses variable-size frames, but it eliminates the error checking required on X.25 networks. A frame relay switch simply reads the header and forwards the packet, perhaps without even fully receiving a frame before forwarding it. Intelligent end stations must identify missing or corrupted frames and request retransmission.

Figure 33: Frame relay is a WAN technology that enables companies to connect LANs through a telecommunicationscarrier's network. AT&T WorldNet Intranet Connect Service currently uses this technology.

Advantages

Frame relay offers several advantages over X.25. Most importantly, frame relay is faster than X.25. Frame relay uses PVCs over leased lines rather than a modem connection. Unlike modem connections, PVCs transmit and receive data immediately, eliminating the call setup and handshaking that modems must perform. In addition, as mentioned above, frame relay does not require error checking and flow control at the switches, reducing overhead and leaving more bandwidth for data transmission. Also, although not as prevalent as X.25, frame relay is a common standard in many countries. Finally, frame relay is less expensive than other WAN technologies because it provides bandwidth on demand, rather than dedicating bandwidth whether data is being transmitted or not.

Although frame relay is fairly complex to implement, value-added resellers and some telephone companies will assist customers in determining their needs and will help install the technology.

Disadvantages

Although frame relay is faster than X.25, its speed is limited because it uses variable-size frames, which can cause delays at switches along the frame's path. As a result, frame relay cannot support applications that require low latency, such as real-time video.

In addition, frame relay is more complex to implement than X.25. Customers must negotiate a service agreement with the phone company, lease a line, and have it installed. They must also purchase and install a frame relay-compatible router.

Asynchronous Transfer Mode (ATM)

Asynchronous Transfer Mode (ATM) is both a LAN and a WAN technology. It is generally implemented as a backbone technology. The exact relationship of the ATM layers to the OSI model is currently undefined, although ATM LAN Emulation works at the data-link layer (level two).

ATM is extremely scalable; data transfer rates range from 25 Mbit/s to 2.4 gigabits per second (Gbit/s). This wide range of data transfer rates reflects the various ways in which ATM can be used. The 25 Mbit/s rate is a new offering meant for desktop environments. In LAN backbones, ATM provides data transfer rates of 100 Mbit/s and 155 Mbit/s. At the high end, WAN implementations using ATM and SONET together have achieved data transfer rates of 2.4 Gbit/s. (For more information about SONET, see the "Synchronous Optical Network" heading later in this primer.)

Distinguishing Characteristics

ATM is a cell relay technology, meaning that it uses standard-sized packets called cells. The size of an ATM cell is 53 bytes.

In a LAN implementation, ATM functions at the data-link layer's media access control sublayer. It further divides the MAC sublayer into three layers: LAN Emulation, ATM Adaptation Layer (AAL), and ATM. LAN Emulation enables you to integrate ATM with Ethernet and token-ring networks without modifying existing Ethernet or token-ring protocols.

On a mixed network, LAN Emulation hardware sits between the Ethernet or token-ring segment and the ATM part of the network. It uses the three layers mentioned above to convert packets moving toward the ATM segment into cells and to assemble cells moving toward the Ethernet or token-ring segment into packets. AAL and ATM put data into standard-sized cells. In most network computing situations, ATM Adaptation Layer 5 breaks packets into 48-byte blocks that are then passed to the ATM layer, where the five-byte header is attached to form a complete 53-byte cell.

Advantages

Many people believe that ATM will become the industry-standard transmission technology for LANs and WANs. The scalability, discussed above, seems to be limitless. Data transfer rates have climbed into the gigabit range and are still growing.

One reason that ATM is so fast is its use of cells. Because cells are a standard size, ATM networks handle data in a predictable, efficient manner at the switches. Standard-sized cells and high-bandwidth media like fiber-optic cable also enable ATM to support real-time voice, video, and data traffic.

ATM also offers flexibility in its transmission media. As many as 22 ATM specifications exist for media like unshielded twisted-pair, shielded twisted-pair, and fiber-optic cable. (ATM is generally implemented with fiber-optic cable.)

Although it is seen as a technology of the future, ATM can currently be integrated with Ethernet and token-ring networks, through use of LAN Emulation.

Disadvantages

ATM standards are still developing. Without industry standards, interoperability between equipment from different vendors is not guaranteed. Furthermore, ATM is more expensive than the other high-speed LAN technologies. The expense is preventing many companies from taking ATM to the desktop.

Integrated Services Digital Network (ISDN)

Integrated Services Digital Network (ISDN) is a set of protocols defined by CCITT to integrate data, voice, and video signals into digital telephone lines. It functions at the physical, data-link, network, and transport layers (levels one through four) of the OSI model. ISDN offers data transfer rates between 128 kbit/s and either 1.544 Mbit/s or 2.048 Mbit/s, depending on the country where it is implemented.

Distinguishing Characteristics

ISDN makes end-to-end digital connections over telephone lines. Although many telephone networks are almost completely digital, the local loop that connects a home or office to the telephone company's network is not: Most local loops send analog rather than digital signals. ISDN replaces local analog signaling with digital signaling, enabling end-to-end digital communications.

ISDN offers Basic Rate Interface (BRI) for individuals or small branch offices and Primary Rate Interface (PRI) for larger companies.

BRI uses two bearer, or B, channels (providing 64 kbit/s each) to transmit and receive data and one delta, or D, channel for call setup and management.

PRI is the same thing as a T1 line. A T1 line in the United States consists of 23 B channels and one D channel, providing a total data transfer rate of 1.544 Mbit/s. A T1 line in Europe consists of 30 B channels and one D channel, providing a total data transfer rate of 2.048 Mbit/s. A fractional T1 uses only some of the B channels in a T1 line (and thus offers some fraction of the total T1 data transfer rate).

ISDN requires special equipment at the customer's site, including a digital phone line and a network termination unit (NT-1). An NT-1 converts the bandwidth coming over the line into the B and D channels and helps the phone company with diagnostic testing. The NT-1 also provides a connection for terminal equipment, such as ISDN telephones and computers that have an ISDN interface. In addition, the NT-1 provides terminal adapter (TA) equipment to connect equipment that is not compatible with ISDN. TA equipment provides an intermediary connection point: Such equipment has an ISDN interface, for connection to the NT-1, and a non-ISDN interface, for connection to non-ISDN equipment.

Advantages

ISDN increases speed and broadens data transmission capabilities, especially for those currently using analog modems to remotely connect to an office or to access the Internet. It offers faster call setup and faster data transfer rates. The transfer rates are acceptable for transmitting voice, data, limited video, fax, and images. ISDN can also be used for limited LAN-to-LAN communications.

With ISDN, you can transmit voice and data traffic simultaneously: An ISDN user can simultaneously talk on the phone and download a data file to his or her computer, over the same telephone line. For example, one BRI ISDN configuration enables users to use the two B channels (128 kbit/s) for data and part of the D channel for a phone conversation.

Disadvantages

Although widely available in Australia, Japan, and Western Europe, ISDN is available in only 50 percent of the United States. Presently, telephone companies are working to make it available throughout the United States.

Acceptance of ISDN in the United States has been slow for several reasons. First, to understand ISDN well enough to even order services requires considerable effort. Furthermore, configuration can be difficult. In addition, ISDN lacks the standards that ensure interoperability. As a result, customers must be careful to purchase equipment that is compatible with the local phone company's equipment. Another problem is that not all phone companies offer the same services, so customers must ensure that the services they need are available in their area. Finally, to take full advantage of ISDN, customers must communicate with others who also have ISDN.

Synchronous Optical Network (SONET)

Synchronous Optical Network, also known in some countries as Synchronous Digital Hierarchy, is a WAN technology that functions at the physical layer (level one) of the OSI model. Telecommunications companies are implementing SONET on some of their networks: A typical business would not implement this standard on its network. SONET has been accepted by ANSI and recommended by the ITU. It specifies a number of data transfer rates from 51.8 Mbit/s to 2.48 Gbit/s.

Distinguishing Characteristics

SONET defines a fiber-optic standard for high-speed digital traffic. This standard provides the flexibility to transport many digital signals with different capacities.

Data communications sometimes prove difficult because digital signaling rates can vary. For example, in the United States, a T1 line provides 1.544 Mbit/s; in Europe, a T1 line (sometimes called an E1 line) provides 2.048 Mbit/s. SONET resolves such problems by defining how switches and multiplexers coordinate communications over lines with different speeds, including defining data transfer rates and frame format.

SONET defines a number of Optical Carrier (OC) levels. Each level defines an optical signal and a corresponding electrical signal called Synchronous Transport Signal (STS). The base level is OC-1/STS-1 or 51.84 Mbit/s. Each level's rate is a multiple of 51.84 Mbit/s. The table below shows the OC levels and the corresponding data transfer rates that SONET defines.

OC Level Data Rate
OC-1 51.8 Mbit/s
OC-3 155.5 Mbit/s
OC-9 466.5 Mbit/s
OC-12 622.0 Mbit/s
OC-18 933.1 Mbit/s
OC-24 1.24 Gbit/s
OC-36 1.86 Gbit/s
OC-48 2.48 Gbit/s

SONET also provides easy access for low-speed signals, such as DS-0 (64 kbit/s) and DS-1 (1.544 Mbit/s) by assigning them to sub-STS-1 signals called Virtual Tributaries.

Advantages

The SONET standard defines data transfer rates and a frame format that all vendors and telephone companies throughout the world can use, creating the potential for global networking. SONET also includes management capabilities for telephone company equipment. Cell relay technologies such as Switched Multimegabit Data Services and ATM operate above SONET, making SONET the expected foundation for future broadband service.

Disadvantages

Some telephone companies are currently using SONET in their networks, but they are not yet offering it to the public on a tariffed basis. Unless your company is a large corporation in a metropolitan area, you probably cannot get dedicated SONET service. Also, some countries do not yet have a digital, fiber-optic telecommunications infrastructure, which means they cannot take advantage of SONET.

[Prior] Prior         [Index] Index [Index]         Next [Next]
[TOP of Page]     This page is maintained by:   Michael P. Harris
[Viking Home Page] viking.delmar.edu
Last Updated: Aug 29		 [E-mail] mpharris@surf.delmar.edu
Copyright © 1998