What does it do?
The function of a router, as the name implies, is to direct or route traffic between networks. A router is concerned with the network addresses of the messages it handles. A message may pass through several devices, each with a MAC (hardware) address on the way to its ultimate destination. This destination, as well as the message source, have a network address that is programmed by the network administrator. A network address, is broken into two components, one representing the network identifier, the other the specific node on that network. A router keeps track of what networks are attached to each of its ports, and passes appropriate traffic to each network. In addition, a router can be programmed to filter certain types of network traffic, while passing others. This might be traffic to or from a certain address, or all traffic of a certain type.

Routers are capable of learning the routes to each network, and storing them in a routing table. A routing table can also be updated manually. Routers also send each other periodic messages about changes in routing information, called RIP (routing information protocol) messages. Routers also keep track of data relating to the shortest and least expensive paths between networks, which are also stored in the routing table and used to make routing decisions.

How is it used?
A router is used to segment and control network traffic. It offers some of the same functionality as a bridge (see Bridge), but whereas a bridge passes or restricts traffic between segments of the same network, a router controls traffic between two or more networks. A router also offers much more flexibility than a bridge, since it can control traffic based not only on its destination address, but also its source address, the type of message, or the relative efficiency of multiple paths.

Another advantage of routers is that they can offer more than one path through the network for a given message, so that they provide a backup path in case of failure or congestion on the primary route. Routers may be set up so that one port provides a primary, or usual path for traffic to a certain network, while a second port is designated as a backup path. This type of redundant path design is called a mesh topology.

Finally, routers, unlike bridges, recognize and react to different types of high-level protocols. However, some protocols are not routable, so the router must be capable of bridging these protocols.

Where do I get more information?
Contact your local CBV Office.