Lesson 3: Ethernet

[Previous] [Next]

This lesson introduces the Ethernet network architecture. Over the years, Ethernet has become the most popular media access method to the desktop computer and is used in both small and large network environments. Ethernet is a nonproprietary industry standard that has found wide acceptance by network hardware manufacturers. Problems related to using Ethernet hardware products from different hardware manufacturers in a single network are nearly nonexistent. This lesson presents an overview of the major Ethernet components, features, and functions.

After this lesson, you will be able to:

• Identify the standard Ethernet components.
• Describe the features of each IEEE Ethernet standard topology.
• Identify the cabling for a given IEEE Ethernet standard topology.
• Determine which Ethernet topology would be appropriate for a given site.

Estimated lesson time: 50 minutes

The Origin of Ethernet

In the late 1960s, the University of Hawaii developed a WAN called ALOHA. (A WAN extends LAN technology across a larger geographical area. For more information on WANs, see Chapter 1, "Introduction to Networking.") The university occupied a wide area and sought to connect computers that were spread throughout the campus. One of the key features of the university's network was its use of CSMA/CD as the access method.

This early network was the foundation for today's Ethernet architecture. In 1972, Robert Metcalfe and David Boggs invented a cabling and signaling scheme at the Xerox Palo Alto Research Center (PARC) and in 1975 introduced the first Ethernet product. The original version of Ethernet was designed as a system of 2.94 megabits per second (Mbps) to connect over 100 computers on a 1-kilometer (.62 miles) cable.

Xerox Ethernet was so successful that Xerox, Intel Corporation, and Digital Equipment Corporation drew up a standard for a 10-Mbps Ethernet. Today, the 10-Mbps Ethernet is one of several specifications describing methods for computers and data systems to connect and share cabling.

Ethernet Specifications

Although networking standards are not discussed in detail until Chapter 5, it is important for you to be aware of them at this point. In 1978, the International Organization for Standardization (ISO) released a set of specifications for connecting dissimilar devices. This set of standards is referred to as the OSI reference model (OSI stands for Open Systems Interconnection). The Ethernet specification performs the same functions as the OSI physical and data-link layers of this model. As you will see later, these specifications affect how hardware links, or passes information to and from, ISO standards. In the 1980s the IEEE published Project 802. This project generated standards for design and compatibility for hardware components that operated within the OSI physical and data-link layers. The standard that pertains to Ethernet is the IEEE 802.3 specification.

Ethernet Features

Ethernet is currently the most popular network architecture. Figure 3.13 shows a simple Ethernet bus network. Notice that the cable is terminated at both ends. This baseband architecture uses a bus topology, usually transmits at 10 Mbps, and relies on CSMA/CD to regulate traffic on the main cable segment.

The Ethernet media is passive, which means it requires no power source of its own and thus will not fail unless the media is physically cut or improperly terminated.

Figure 3.13 Simple Ethernet bus network terminated at both ends

Ethernet Basics

Table 3.2 summarizes Ethernet features:

Table 3.2 Summary of Ethernet

Feature	Description
Traditional topology	Linear bus
Other topologies	Star bus
Type of architecture	Baseband
Access method	CSMA/CD
Specification	IEEE 802.3
Transfer speed	10 Mbps or 100 Mbps
Cable type	Thicknet, thinnet, UTP

The Ethernet Frame Format

Ethernet breaks data down into packages in a format that is different from the packets used in other networks: Ethernet breaks data down into frames. (Remember that the terms "packet" and "frame" can be used interchangeably; in the context of Ethernet, the term "frame" is used.) A frame is a package of information transmitted as a single unit. An Ethernet frame can be between 64 and 1518 bytes long, but the Ethernet frame itself uses at least 18 bytes; therefore, the data in an Ethernet frame can be between 46 and 1500 bytes long. Every frame contains control information and follows the same basic organization.

For example, the Ethernet II frame, used for Transmission Control Protocol/Internet Protocol (TCP/IP), which gets transmitted across the network, consists of the sections listed in Table 3.3 (TCP/IP has become the de facto standard for data transmission over networks, including the Internet):

Table 3.3 Components of an Ethernet II Frame

Frame field	Description
Preamble	Marks the start of the frame
Destination and source	The origin and destination addresses
Type	Used to identify the network layer protocol, usually either IP or IPX (Novell's Internetwork Packet Exchange)
Cyclical redundancy check (CRC)	Error-checking field to determine if the frame arrived without being corrupted

An illustration of an Ethernet frame is shown in Figure 3.14.

Figure 3.14 Sample Ethernet II frame

Ethernet networks include a variety of cabling and topology alternatives. The remaining sections of this lesson present these alternatives based on their IEEE specification.

The 10-Mbps IEEE Standards

This section looks at four 10 Mbps Ethernet topologies:

• 10BaseT
• 10Base2
• 10Base5
• 10BaseFL

10BaseT Standard

In 1990, the IEEE committee published the 802.3 specification for running Ethernet over twisted-pair wiring. The result, 10BaseT (10 Mbps, baseband, over twisted-pair cable), is an Ethernet network that typically uses unshielded twisted-pair (UTP) cable to connect computers. Usually, 10BaseT employs UTP, but shielded twisted-pair (STP) cabling will also work without changing any of the 10BaseT parameters.

Most networks of this type are configured in a star pattern, but internally they use a bus signaling system like other Ethernet configurations. Figure 3.15 shows a multiport hub used to extend an Ethernet LAN. Typically, the hub of a 10BaseT network serves as a multiport repeater and often is located in a wiring closet of the building. Each computer is located at the endpoint of a cable that is connected to the hub. Each computer has two pairs of wire; one pair is used to receive data, and one pair is used to transmit data.

The maximum length of a 10BaseT segment is 100 meters (328 feet). Repeaters can be used to extend this maximum cable length. The minimum cable length between computers is 2.5 meters (about 8 feet). A 10BaseT LAN will serve 1024 computers.

Figure 3.15 A multiport repeater (hub) can be used to extend an Ethernet LAN

Figure 3.16 shows how a 10BaseT solution provides the advantages of a star-wired topology. The UTP cable features data transmission at 10 Mbps. It is easy to make changes by moving a modular patch cord on the patch panel. A change at the patch panel will not affect other devices on the network; this differs from a traditional Ethernet bus network.

Figure 3.16 A patch panel makes moving computers easy

Patch panels should be tested for rates higher than 10 Mbps. The latest hubs can provide connections for both thick and thin Ethernet cable segments. In this implementation, it is also easy to convert thick Ethernet cable to 10BaseT cable by attaching a mini 10BaseT transceiver to the AUI port of any network interface card. Table 3.4 summarizes 10BaseT specifications:

Table 3.4 10BaseT Specifications Summary

Category	Notes
Cable	Category 3, 4, or 5 UTP.
Connectors	RJ-45 at cable ends.
Transceiver	Each computer needs one; some cards have built in transceivers.
Transceiver to hub distance	100 meters (328 feet) maximum.
Backbones for hubs	Coaxial or fiber-optic cable to join a larger LAN or to carry major traffic between smaller networks.
Total number of computers per LAN without connectivity components	1024 by specification.

10Base2 Standard

Another topology is 10Base2, given this name in the IEEE 802.3 specification because it transmits at 10 Mbps over a baseband wire and can carry a signal about two times 100 meters (the actual distance is 185 meters, or 607 feet).

This type of network uses thin coaxial cable, or thinnet, which has a maximum segment length of 185 meters (607 feet) and a minimum cable length of at least 0.5 meters (20 inches) between workstations. There is also a 30-computer maximum per 185-meter segment.

Thinnet cabling components include:

• BNC barrel connectors.
• BNC T connectors.
• BNC terminators.

Thinnet networks generally use a local bus topology. IEEE standards for thinnet do not allow a transceiver cable to be used from the bus T connector to a computer. Instead, a T connector fits directly on the NIC.

A BNC barrel connector may be used to connect thinnet cable segments together, thus extending a length of cable. For example, if you need a length of cable that is nine meters (30 feet) long, but all you have is a 7.5-meter (25-foot) length and a 1.5-meter ( 5-foot) length of thinnet cable, you can join the two cable segments together using a BNC barrel connector. However, the use of barrel connectors should be kept to a minimum because each connection in the cable reduces the signal quality and adds to the risk of cable separation and disconnection.

A thinnet network is an economical way to support a small department or workgroup. The cable used for this type of network is:

• Relatively inexpensive.
• Easy to install.
• Easy to configure.

A single thinnet network can support a maximum of 30 nodes (computers and repeaters) per cable segment, as per the IEEE 802.3 specification.

The 5-4-3 Rule

A thinnet network can combine as many as five cable segments connected by four repeaters; but only three segments can have stations attached. Thus, two segments are untapped and are often referred to as "inter-repeater links." This is known as the 5-4-3 rule.

In Figure 3.17, there are five segments, four repeaters, and trunk segments 1, 2, and 5 are populated (have computers attached to them). Trunk segments 3 and 4 exist only to increase the total length of the network and to allow the computers on trunk segments 1 and 5 to be on the same network.

Figure 3.17 The thinnet 5-4-3 rule: 5 segments, 4 repeaters, and 3 populated segments

Because normal Ethernet limits are too confining for a large business, repeaters can be used to join Ethernet segments and extend the network to a total length of 925 meters (3035 feet). The following table summarizes 10Base2 specifications:

Table 3.5 10Base2 Specifications Summary

Category	Notes
Maximum segment length	185 meters (607 feet).
Connection to network interface card	BNC T connector.
Trunk segments and repeaters	Five segments can be joined using four repeaters.
Computers per segment	30 computers per segment by specification.
Segments that can have computers	Three of the five segments can be populated.
Maximum total network length	925 meters (3035 feet).

10Base5 Standard

The IEEE specification for this topology is 10 Mbps, baseband, and 500-meter (five 100-meter) segments. It is also called standard Ethernet.

This topology makes use of thick coaxial cable (see Figure 3.18), also known as thicknet. Thicknet generally uses a bus topology and can support as many as 100 nodes (stations, repeaters, and so on) per backbone segment. The backbone, or trunk segment, is the main cable from which transceiver cables are connected to stations and repeaters. The distances and tolerances for thicknet are greater than those for thinnet: a thicknet segment can be 500 meters (1640 feet) long for a total network length of 2500 meters (8200 feet).

Figure 3.18 Thicknet cable composition

The thicknet cabling components include:

• Transceivers These are devices that can both transmit and receive, provide communications between the computer and the main LAN cable, and are located in the vampire taps attached to the cable.
• Transceiver cables The transceiver cable (drop cable) connects the transceiver to the NIC.
• DIX (or AUI) connectors These are the connectors on the transceiver cable.
• N-series connectors, including N-series barrel connectors, and N-series terminators The thicknet components work in the same way as the thinnet components. Figure 3.19 shows a thicknet cable with a transceiver attached and a transceiver cable. It also shows the DIX or AUI connector on the transceiver cable.

NOTE

---

"AUI," an acronym for attachment unit interface, is a 15-pin (DB-15) connector commonly used to connect a NIC to an Ethernet cable; AUIs and DIXs are discussed in Chapter 2, "Basic Network Media."

Figure 3.19 Thicknet backbone with attached transceiver and cable

The 5-4-3 Rule in Thicknet

One thicknet Ethernet network can have a maximum of five backbone segments connected using repeaters (based on the IEEE 802.3 specification), of which up to three can accommodate computers. Figure 3.20 shows how the 5-4-3 rules are applied to thicknet. The length of the transceiver cables is not used to measure the distance supported on the thicknet cable; only the end-to-end length of the thicknet cable segment itself is used.

Figure 3.20 Thicknet 5-4-3 rule; 5 backbone segments, 4 repeaters, and 3 segments

Between connections, the minimum thicknet cable segment is 2.5 meters (about 8 feet). This measurement excludes transceiver cables. Thicknet was designed to support a backbone for a large department or an entire building. Table 3.6 summarizes 10Base5 specifications:

Table 3.6 10Base5 Specifications Summary

Category	Notes
Maximum segment length	500 meters (1640 feet).
Transceivers	Connected to the segment (in the tap).
Maximum computer-to-transceiver distance	50 meters (164 feet).
Minimum distance between transceivers	2.5 meters (8 feet).
Trunk segments and repeaters	Five segments can be joined using four repeaters.
Segments that can have computers	Three of the five segments can be populated.
Maximum total length of joined segments	2500 meters (8200 feet).
Maximum number of computers per segment	100 by specification.

Combining Thicknet and Thinnet Cable

It is common for larger networks to combine thick and thin Ethernet cable. Thicknet cable is good for backbones, while thinnet cable is used for branch segments. What this means is that the thicknet cable is the main cable covering the long distances. As described in Chapter 2, "Basic Network Media," thicknet cable has a larger copper core and can, therefore, carry signals for a longer distance than thinnet. The transceiver attaches to the thicknet cable, and the transceiver cable's AUI connector plugs into a repeater. The branching segments of thinnet plug into the repeater and connect the computers to the network.

10BaseFL Standard

The IEEE committee published a specification for running Ethernet over fiber-optic cable. The result, 10BaseFL (10Mbps, baseband, over fiber-optic cable) is an Ethernet network that typically uses fiber-optic cable to connect computers and repeaters.

The primary reason for using 10BaseFL is to accommodate long cable runs between repeaters, such as between buildings. The maximum distance for a 10BaseFL segment is 2000 meters (about 6500 feet).

The 100-Mbps IEEE Standards

New Ethernet standards are pushing the traditional Ethernet limits beyond the original 10 Mbps. These new capabilities are being developed to handle such highbandwidth applications as:

• Computer-aided design (CAD).
• Computer-aided manufacturing (CAM).
• Video.
• Imaging and document storage.

Two Ethernet standards that can meet the increased demands are:

• 100BaseVG-AnyLAN Ethernet.
• 100BaseX Ethernet (Fast Ethernet).

Both 100BaseVG-AnyLAN and Fast Ethernet are about 5 to 10 times faster than standard Ethernet. They are also compatible with existing 10BaseT cabling systems. This means they allow for Plug and Play upgrades from existing 10BaseT installations.

100VG-AnyLAN Standard

The 100VG (Voice Grade) AnyLAN is an emerging networking technology that combines elements of both Ethernet and Token Ring architectures. Originally developed by Hewlett-Packard, it is currently being refined and ratified by the IEEE 802.12 committee. The 802.12 specification is a standard for transmitting 802.3 Ethernet frames and 802.5 Token Ring packets.

This technology goes by any of the following names, all of which refer to the same type of network:

• 100VG-AnyLAN
• 100BaseVG
• VG
• AnyLAN

Specifications

Some of the current 100VG-AnyLAN specifications include:

• A minimum data rate of 100 Mbps.
• The ability to support a cascaded star topology over Category 3, 4, and 5 twisted-pair and fiber-optic cable.
• The demand-priority access method that allows for two priority levels (low and high).
• The ability to support an option for filtering individually addressed frames at the hub to enhance privacy.
• Support for both Ethernet frames and Token Ring packets.

Topology

A 100VG-AnyLAN network is built on a star topology in which all computers are attached to a hub. Figure 3.21 shows a parent hub with five child hubs. Adding child hubs to the central hub can expand the network. The child hubs act as computers to their parent hubs. The parent hubs control transmission of computers attached to their children.

Figure 3.21 Parent hub with five attached child hubs

Considerations

This topology requires its own hubs and cards. Also, the cable distances of 100BaseVG are limited when compared to 10BaseVG and other implementations of Ethernet. The longest cable from the 100BaseVG hub to a computer cannot exceed 250 meters (about 820 feet). Extending this limit requires special equipment used to expand the size of a LAN. These cable-length limits mean that 100BaseVG will require more wiring closets than 10BaseVG.

100BaseX Ethernet Standard

This standard, sometimes called Fast Ethernet, is an extension of the existing Ethernet standard. It runs on UTP Category 5 data-grade cable and uses CSMA/CD in a star-wired bus topology, similar to 10BaseT where all cables are attached to a hub.

Media Specifications

100BaseX incorporates three media specifications:

• 100BaseT4 (4-pair Category 3, 4, or 5 UTP)
• 100BaseTX (2-pair Category 5 UTP or STP)
• 100BaseFX (2-strand fiber-optic cable)

These media are described further in Table 3.7:

Table 3.7 100BaseX Media Specifications

Value	Represents	Actual meaning
100	Transmission speed	100 Mbps
Base	Signal type	Baseband
T4	Cable type	Indicates twisted-pair cable using four telephone-grade pairs
TX	Cable type	Indicates twisted-pair cable using two data-grade pairs
FX	Cable type	Indicates fiber-optic link using two strands of fiber-optic cable

Performance Considerations

Ethernet architecture can use multiple communication protocols and can connect mixed computing environments such as Netware, UNIX, Windows, and Macintosh.

Segmentation

Ethernet performance can be improved by dividing a crowded segment into two less-populated segments and joining them with either a bridge or a router. Bridges and routers are discussed later in more detail in Chapter 7, "Elements of Network Connectivity." Figure 3.22 shows how a bridge is used to extend a network. This reduces traffic on each segment. Because fewer computers are attempting to transmit onto the segment, access time improves.

Figure 3.22 Using a bridge to segment a network and reduce network traffic

Consider dividing segments if large numbers of new users are joining the network or if new, high-bandwidth applications, such as database or video programs, are being added to the network.

Network Operating Systems on Ethernet

Ethernet will work with most popular network operating systems including:

• Microsoft Windows 95, Windows 98, and Windows 2000.
• Microsoft Windows NT Workstation and Windows NT Server.
• Microsoft Windows 2000 Professional and Windows 2000 Server.
• Microsoft LAN Manager.
• Microsoft Windows for Workgroups.
• Novell NetWare.
• IBM LAN Server.
• AppleShare.
• UNIX.

Lesson Summary

The following points summarize the main elements of this lesson:

• Ethernet is one of the most popular network architectures.
• Ethernet is governed by the specifications found in the OSI reference model physical layer and data-link layer, as well as IEEE 802.3.

Table 3.8 summarizes the specifications for Ethernet architecture discussed in this lesson. It outlines the minimum set of standards required to conform to IEEE specifications. A particular implementation of the network architecture may differ from the information in the table.

Table 3.8 Ethernet Specifications (IEEE 802.3)

	10Base2	10Base5	10BaseT
Topology	Bus	Bus	Star bus
Cable type	RG-58 (thinnet coaxial cable)	Thicknet; one-centimeter (3/8-inch) shielded transceiver cable	Category 3, 4, or 5 unshielded twisted-pair cable
Connection to NIC	BNC T connector	DIX or AUI connector	RJ-45
Terminator resistance, Ω (ohms)	50	50	Not applicable
Impedance, Ω	50 ± 2	50 ± 2	85-115 unshielded twisted-pair; 135-165 shielded twisted-pair
Distance	0.5 meters between computers (23 inches)	2.5 meters (8 feet) between taps and maximum of 50 meters (164 feet) between the tap and the computer	100 meters (328 feet) between the transceiver (the computer) and the hub
Maximum cable segment length	185 meters (607 feet)	500 meters (1640 feet)	100 meters (328 feet)
Maximum connected segments	5 (using 4 repeaters); Only 3 segments can have computers connected.	5 (using 4 repeaters). Only 3 segments can have computers connected.	Not applicable
Maximum total network length	925 meters (3035 feet)	2460 meters (8000 feet)	Not applicable
Maximum computers per segment	30 (There can be a maximum of 1024 computers per network.)	100	1 (Each station has its own cable to the hub. There can be a maximum of 12 computers per hub and a maximum of 1024 transceivers per LAN without some type of connectivity.)