Many computer users focus on the technologies that operate within a single Local Area Network (LAN). Of course, there must be a way for data to be transferred between LANs, so as to create the network of computer networks that is the Internet. A router is the key piece of hardware in the Internet, as it acts – strictly – as an interface between two computer networks.
Routers allow data to be transferred or routed between networks. They can do this in the most efficient way possible, to networks far removed from the LAN in which the data originated. When using the Transmission Control Protocol/Internet Protocol (TCP/IP), routers use IP addresses to determine the path to a destination. Devices such as switches and bridges, on the other hand, use media access control (MAC) addresses to determine the correct path to a destination. A router is known as a stateless device because it handles the destination address of the data that it routes rather than the data itself.
A router creates and maintains a table of all the available routes in the networks to which it is connected. When a router receives a packet, it first checks the destination IP address of the packet. It then uses the table to determine the most efficient, available routing path for the packet. The entries in the table can be inputted manually or dynamically maintained. If the router fails to find a good route, it may forward the packet to another router or drop the packet altogether. The way in which the router treats different packets can be specified by a network administrator.
Routers in different networks
Routers can be connected to several networks, and can route traffic to and from the networks to which they are connected. So for example, router A in the figure above, which is connected to a client PC, may belong to one network, whereas router B, connected to a server, belongs to another network. Data flows to and from the client PC via the intermediate routers, which belong to different networks, although some routers (1 and 2, or 3 and 4) belong to the same network. The exact path taken between routers A and B will depend on many factors, such as the amount of network traffic.
In fact, if a chunk of data at B is divided into several packets and sent to A, each packet may take a different route. The chunk of data can be reassembled from the packets at B and presented to the application that requires it. Data can only be routed to remote networks in this way if the protocol used to produce the data is a routable protocol. TCP/IP and Internetwork Packet Exchange/Sequenced Packet Exchange (IPX/SPX) for example, are routable protocols, whereas NetBios Extended User Interface (NetBEUI) is not.
A brouter is device that functions both as a network bridge and as a router. A brouter can route TCP/IP and IPX/SPX packets to remote networks, as a router can do, as these are routable protocols. However a brouter can also handle other traffic, such as NetBEUI packets, in the same way that a bridge would.