LAN Technologies

It’s time to describe the basic characteristics of Ethernet, Token Ring, FDDI, and ATM—four LAN technologies that account for virtually all deployed LANs—and take a look at the Data Link and Physical layer details of each of them:

Ethernet Though one of the very first LAN technologies, the largest installed LANs employ Ethernet.

Token Ring Token Ring is an IBM creation that is widely used in a large number of corporations that migrated from mainframes to LANs.

FDDI FDDI was typically used as a backbone LAN between data closets. It is a popular campus LAN because it is very dependable and faster than Ethernet and Token Ring.

ATM ATM is taking the place of FDDI in the campus backbone arena. It’s gaining in popularity because it can run in both WAN and LAN environments at tremendous speeds.

Ethernet and IEEE 802.3

In 1980, Digital, Intel, and Xerox (DIX) created the original Ethernet I. Predictably, Ethernet II followed and was released in 1984. Ethernet II is also described as Carrier Sense, Multiple Access with Collision Detect (CSMA/CD). In response, the IEEE created the 802.3 subcommittee to come up with an Ethernet standard that happens to be almost identical to the Ethernet II version of Ethernet. The two differ only in their descriptions of the Data Link layer. Ethernet II has a Type field, whereas 802.3 has a Length field. Even so, they’re both common in their Physical layer specifications, MAC addressing, and understanding of the LLC sublayer’s responsibilities.

Ethernet II and 802.3 both define a bus-topology LAN at 10Mbps, and the cabling defined in these standards is identical:

10Base2/Thinnet Thinnet segments can extend up to 185 meters using RG58 coax at 50 ohms.

10Base5/Thicknet Thicknet segments can extend up to 500 meters using RG8 or 11 at 50 ohms.

10BaseT/UTP All hosts connect using unshielded twisted-pair (UTP) cable to a central device (a hub or switch). Category 3 UTP is specified to 10Mbps, and Category 5 UTP is specified to 100Mbps CSMA/CD.

10BaseFL 10BaseFL is 10Mbps over fiber-optic physical medium. The advantages of using 10BaseFL over a copper medium are security, a larger network diameter, and immunity to RFI and EMI.

CSMA/CD

Carrier Sense Multiple Access with Collision Detection (CSMA/CD) was created to overcome the problem of collisions that, as mentioned earlier, occur when packets are transmitted simultaneously from different nodes. Good collision management is important, because when a node transmits in a CSMA/CD network, all the other nodes on the network receive and examine that transmission. Only bridges and routers effectively prevent a transmission from propagating through the entire network.

The CSMA/CD protocol works like this: When a host wants to transmit over the network, it first checks for the presence of a digital signal on the wire. If all is clear (if no other host is transmitting), the host will then proceed with its transmission. And it doesn’t stop there. The transmitting host constantly monitors the wire to make sure no other hosts begin transmitting. If the host detects another signal on the wire, it then sends out an extended jam signal that causes all nodes on the segment to stop sending data. The nodes respond to that jam signal by waiting before attempting to transmit again. If after 15 tries collisions keep occurring, the nodes attempting to transmit will then time out.

Broadcasts

A broadcast is a frame sent to all network stations at the same time. Remember that broadcasts are built into all protocols. In the following example, the dissected frame of an Etherpeek (a protocol analyzer) trace is displayed so you can see the destination hardware address, IP address, and more:

Ethernet Header   Destination:  ff:ff:ff:ff:ff:ff Ethernet Broadcast   Source:       02:07:01:22:de:a4   Protocol Type:08-00  IP IP Header - Internet Protocol Datagram   Version:              4   Header Length:        5   Precedence:           0   Type of Service:      %000   Unused:               %00   Total Length:         93   Identifier:           62500   Fragmentation Flags:  %000   Fragment Offset:      0   Time To Live:         30   IP Type:              0x11  UDP   Header Checksum:      0x9156   Source IP Address:    10.7.1.9   Dest. IP Address:     10.7.1.255   No Internet Datagram Options

As this information shows, the source hardware and IP address are from the sending station that knows its own information. Its hardware address is 02:07:01:22:de:a4., and its source IP address is 10.7.1.9. The destination hardware address is ffffffffffff, a MAC sublayer broadcast that is monitored by all stations on the network. The destination network address is 10.7.1.255—an IP broadcast for network 10.7.1.0—meaning all devices on network 10.7.1.0.

A frame addressed in this manner tells all the hosts on network 10.7.1.0 to receive it and process the data therein. This can be both a good thing and a bad thing. When servers or other hosts need to send data to all the other hosts on the network segment, network broadcasts are very useful indeed. But, if a lot of broadcasts are occurring on a network segment, network performance can be seriously impaired. This is one very big reason why it is so important to segment your network properly with bridges and/or routers. This process is called network segmentation.

Fiber Distributed Data Interface (FDDI)

Like Token Ring, Fiber Distributed Data Interface (FDDI), as shown in Figure 1.20, is a token-passing media access topology. American National Standards Institute (ANSI) defines the standard (ANSI X3T9.5) for a dual Token Ring LAN operating at 100Mbps over fiber-optic cable. Copper Distributed Data Interface (CDDI) can be used with UTP cable to connect servers or other stations directly into the ring, as you can see in Figure 1.20.

Figure 1.20: FDDI network topology

The advantages of FDDI include the following:

• FDDI can run very long distances and do so in electronically hostile environments where electromagnetic or radio frequency interference is present.

• It runs at a high speed compared with 10Mbps Ethernet and 4/16Mbps Token Ring LANs.

• FDDI employs a token-passing media access with dual counter-rotating rings, as shown in Figure 1.21. Typically, only one ring is active at any given time. The active ring, called the primary ring, is used for data transmission; the secondary ring usually is not active. That way, if a break or outage occurs, the FDDI ring will wrap back the other direction, keeping the ring intact.

• Some stations can be attached to both rings for redundancy reasons. These are known as dual attachment stations (DASs). These would be used mostly for high availability stations like servers. There are also single attachment stations (SASs), which are attached to the FDDI rings using a device known as a concentrator.

• Cisco routers can attach with a technique called dual homing. This provides fault tolerance by providing a primary and backup path to the FDDI ring.

• FDDI is both a logical and a physical ring—the only LAN that is an actual, physical ring. Like Token Ring, FDDI provides predictable deterministic delays and priorities.

• FDDI uses MAC addresses like other LANs do, but it uses a different numbering scheme. Instead of the eight-bit bytes that Ethernet and Token Ring uses, it applies four-bit symbols. FDDI has 12 four-bit symbols that make up its MAC addresses.

• Token Ring allows only one token on the ring at any given time, whereas FDDI permits several tokens to be present on the ring concurrently.

Some drawbacks of migrating to FDDI include the following:

• Relatively high latency occurs when Ethernet-to-FDDI and FDDI-to- Ethernet translation is performed between LANs.

• Capacity is still shared because FDDI dual ring is a shared LAN.

• There’s no full-duplex capability in shared networks.

• It’s expensive—very expensive! FDDI components, i.e., concentrators and NICs, aren’t exactly bargain equipment.

Figure 1.21 shows how an FDDI LAN would wrap the primary ring back to the standby, secondary ring if a failure occurred.

Figure 1.21: Dual-ring reliability

When a station realizes that no tokens have been received from its nearest active upstream neighbor (NAUN) for a predetermined time period, it sends out a beacon as an alert and as an attempt to locate the failure. Once it starts to receive its own beacons, the station assumes the ring is now up and running. If it doesn’t receive its beacon back for a predetermined amount of time, the primary ring will wrap to the secondary ring as shown.

Token Ring

IBM created Token Ring in the 1970s; it was popular with true-blue customers needing to migrate from a mainframe environment. It lost to Ethernet in the popularity polls because it was pricey by comparison. However, depending on what you’re looking for, Token Ring is a more resilient network, especially under heavy loads. Sometimes you actually do get what you pay for.

Like Ethernet, the IEEE came out with its own standard for Token Ring, designated 802.5. This standard was so close to the IBM standard that the IEEE is now responsible for administrating both specifications.

At the Physical layer, Token Ring runs as a star topology using shielded twisted-pair (STP) wiring. Each station connects to a central hub called a multistation access unit (MSAU). Logically, it runs in a ring where each station receives signals from its NAUN and repeats these signals to its downstream neighbors.

Token Ring uses MAC addresses like Ethernet does, but that’s where the similarities end. Token Ring media access is described point by point below:

• Stations can’t transmit whenever they want to, like Ethernet stations can. Instead, they have to wait to be given a special frame called a token. When a station receives a token, it does one of two things:

  • It appends the data it wants to send onto the end of the frame and then changes the T bit in the frame. Doing that alerts the receiving station that data is attached.

  • If the station that gets a token doesn’t need to send any data, it simply passes on the token to the next station in the ring.

• The information frame circles the ring until it gets to the destination station. The destination station copies the frame and then tags the frame as being copied. The frame continues around until it reaches the originating station, which then removes the tag.

• Typically, only one frame can be on a ring at any given time. However, by using early token release, a station can transmit a new token onto the ring after transmitting its first frame.

• Collisions don’t happen because stations can’t transmit unless they have a token.

The frame in a Token Ring network is different from the frames in Ethernet. As shown in Figure 1.22, the token frame uses a priority system that permits certain user-designated, high-priority stations to use the network more frequently. The media access control field of the frame is shown in this figure.

Figure 1.22: Token Ring media access control field

The two fields that control priority are predictably the priority field and the reservation field. If a priority token is transmitted, only stations with a priority equal to or higher than the priority of that token can claim it. Priority levels are configured by the network administrator. After the token is claimed and changed to an information frame, only stations with a priority rating higher than the transmitting station can reserve the token for the next pass around the network. When the next token is generated, it includes the highest priority for the reserving station. Stations that raise a token’s priority level must reinstate the previous lower priority level after their transmission is complete.

The frame status field is shown in Figure 1.23. The address (A) bit and the copied (C) bit are used to indicate the status of an outstanding frame.

Figure 1.23: Token Ring frame status field

Both the bits are turned off when the sending station transmits the frame. When the sending station receives the frame back again, the station reads this information to ensure that the data was either received correctly by the destination computer or that it needs to be retransmitted.

Active Monitor

One station on a Token Ring network is always an active monitor. The active monitor makes sure that there is no more than one token on the ring at any given time. It is the only device that can generate a new token. It also provides timing on the ring. Finally, if a transmitting station fails, it isn’t able to remove the token as it makes its way back through the ring. Should this occur, the active monitor would step in, remove the token, and then generate a new one. Also, many stations on the ring will be designated as standby monitors (to act as backups) in case the active monitor goes offline.

ATM

The ATM protocol dictates how two end devices communicate with each other across an ATM network through switches. The ATM protocol model contains three functional layers:

ATM Physical layer The ATM Physical layer performs Bit timing on the physical medium.

ATM layer The ATM layer performs Generic flow control, call header generation, multiplexing, and demultiplexing.

ATM Adaptation layer The ATM Adaptation layer (AAL) provides support for higher layer services such as signaling, circuit emulation, voice, and video.

The ATM Physical Layer

The ATM Physical layer is in charge of sending and receiving bits on the physical level. This layer also manages ATM cell boundaries and controls the cell packaging in the correct frame type for the ATM media you use. The ATM Physical layer consists of two sublayers:

• Physical medium dependent (PMD) sublayer

• Transmission convergence (TC) sublayer

The PMD sublayer sends and receives a constant flow of bits that contain associated timing information to synchronize transmission and reception. The PMD sublayer relies on the media used for transport, and thus, ATM only works on ATM-specific media. Standards include DS-3/E3, FDDI, 155Mbps local fiber, and SONET (Synchronous Optical Network)/SDH. The ATM Forum is considering proposals for twisted-pair wire.

The TC sublayer maintains several functions. It mainly extracts and inserts ATM cells within either a Plesiochronous Digital Hierarchy (PDH) or Synchronous Digital Hierarchy (SDH) time-division multiplexed (TDM) frame and passes this to and from the ATM layer. The other functions it provides are as follows:

• Cell delineation for maintaining ATM cell boundaries.

• Header error control sequence generation and verification. The TC layer creates and checks header error control to ensure valid data.

• Cell rate decoupling. The TC layer inserts or suppresses unassigned ATM cells to adapt the rate of valid ATM cells to the payload capacity of the transmission system.

• Transmission frame adaptation. The TC layer packages ATM cells in appropriate frames for physical layer implementation.

• Transmission frame generation and recovery. The TC layer generates and maintains the given Physical layer frame structure.

The ATM Layer

The ATM layer maintains the virtual connections and carries ATM cells through the network. It accomplishes this by using information contained within the header of each ATM cell. The ATM layer is responsible for the following functions:

• Multiplexing and demultiplexing ATM cells from different virtual connections. You can identify these different connections by their VCI and VPI values.

  Note

  A VCI (Virtual Circuit Identifier) can also be called a virtual channel. This is simply the identifier for the logical connection between the two ends of a connection. A VPI (Virtual Path Identifier) is the identifier for a group of VCIs that allows an ATM switch to perform operations on a group of VCs.

• Translation of VCI and VPI values at the ATM switch or cross-connect.

• Extraction and insertion of the header before or after the cell is delivered from or to the ATM Adaptation layer.

• Governing the implementation of a flow-control mechanism at the user-network interface (UNI). UNI is basically two ports connected by a pair of wires, typically fiber.

• Passes cells and accepts cells from the AAL.

ATM Adaptation Layer (AAL)

The AAL provides the translation between the larger service data units of the upper layers of the OSI reference model and the ATM cells. This function works by receiving packets from the upper-level protocols and breaking them into 48-byte segments to be dumped into the payload of an ATM cell. The AAL has two different sublayers: segmentation and reassembly (SAR) and convergence sublayer (CS). The CS contains sublayers within itself: the common part (CP) and the service specific (SS). Like protocols specified in the OSI reference model, protocol data units (PDUs) are used to pass information between these layers.

Specifications exist for a few different ATM adaptation layers:

AAL1 (Class A) AAL1 is used for transporting telephone traffic and uncompressed video traffic. AAL1 uses constant bit rate (CBR) service and end-to-end timing and is connection-oriented. Examples of AAL1 are DS1, E!, and nx64 kbps emulation.

AAL2 (Class B) AAL2 does not use the CS and SAR sublayers. It multiplexes short packets from multiple sources into a single cell. AAL2 uses a variable bit rate (VBR) and end-to-end timing and is connection-oriented. Examples of AAL2 are packet, video, and audio.

AAL3/4 (Class C) AAL3/4 is designed for network service providers, uses VBR with no timing required, but is still connection-oriented. Examples of AAL3/4 are Frame Relay and X.25.

AAL5 (Class D) AAL5 is used to transfer most non-SMDS data and LAN emulation. AAL5 also uses VBR with no timing required.

LANE (LAN Emulation) Components

ATM networks can provide the transport for several different independent emulated LANs. This process is called LAN emulation (LANE). As an attached device to these emulated LANs, the physical location no longer matters to the administrator or implementation. This process allows you to connect several LANs in different locations with switches to create one large emulated LAN. This can make a big difference, since attached devices can now be moved easily between emulated LANs. Thus, an engineering group can belong to one LANE and a design group can belong to another LANE, without ever residing in the same location.

LANE also provides translation between multiple media environments, allowing data sharing. Token Ring or FDDI networks can share data with Ethernet networks as if they were part of the same network.

LANE consists of several components that interact and relate in different ways to provide network connectivity based on the client/server model. The interaction of these components allows broadcast searching, address registration, and address caching. The LANE model is made up of the following components:

LEC (LAN Emulation Client) A LANE client emulates a LAN interface to higher layer protocols and applications. It proxies for users attached into ATM via a non-ATM path.

LES (LAN Emulation Server) A LANE server provides address resolution and registration services to the LANE clients in that emulated LAN. The LES keeps a database of all LANE servers. It also manages the stations that make up the ELAN.

LECS (LAN Emulation Configuration Server) Via a database, an LECS keeps track of which emulated LAN a device belongs to (each configuration server can have a differently named database).

BUS (Broadcast-and-Unknown Server) A BUS is used for broadcasting, sequencing, and distributing multicast and broadcast packets. The BUS also handles unicast flooding.

Warning

Notice that LEC and LECS are completely different terms and components!