The All-New Switch Book: The Complete Guide to LAN Switching Technology, Second Edition

Chapter 3. Bridging Between Technologies

We have made the tacit assumption in our discussion of bridges that all of the bridge ports employed the same technology. Whether interconnecting multiple Ethernets, Token Rings, or FDDI LANs, bridges operate seamlessly and transparently, shuttling frames among ports according to the rules described in Chapter 2. End stations can communicate as if they were on the same LAN, regardless of their relative location. Everything is right with the world.

If we now allow a bridge to connect LANs employing dissimilar technologies, things get ugly. The purpose of the bridge is still to allow end stations to communicate transparently. However, the technologies employed on the bridge's ports may provide different:

• Access control methods

• Frame formats

• Frame semantics (the meaning of the fields in the frame)

• Allowable data length

• Bit-ordering

How can we even consider using a device as simple as a bridge to interconnect such fundamentally incompatible technologies? The answer can be found in the way that the bridge itself operates. While the LAN interfaces on a bridge must of course support the technology of the attached LAN (that is, a bridge needs a 10 Mb/s Ethernet interface on a 10 Mb/s Ethernet port, a 16 Mb/s Token Ring interface on a 16 Mb/s Token Ring port, and so on), the bridge algorithms themselves don't care about the nature of the underlying technology. Bridges filter, forward, and learn solely on the basis of received destination and source ...