Introduction to routers
In a network you can only transfer packets from one network to another when the network portions of the two IP addresses match. When the network portions of the IP addresses do not match, a hub or a switch will not forward packets from one network to another. A router is required to forward these packets from network to network. This might sound like a drawback or a limitation, but this is advantage as it allows network administrators to segment networks and create smaller collision and broadcast domains, thus making networks perform more efficiently as there are less collisions and re-transmissions.
A router is required to forward packets from one network to another, a router is a device that forwards data from one network to another network, commonly from a LAN to another LAN or perhaps via a Wide Area Network.
A router can be described as a gateway (quite often the local router port is known as the ‘default gateway’ as it is thought of being a gateway out of a current network to other networks.)
In order to know where it has to route packets, routers can either be told (by a network administrator) how the network is configured or they can find out for themselves. Routers have the ability to talk to neighbouring routers to get a picture of the network and they can determine what they think is the best path through a network. When routers talk to each other they do so by a ROUTING Prototcol. Common routing protocols are RIP OSPF and EIGRP (The advantages and disadvantages of each routing protocol will be covered later in this unit)
As routers deal with ip addresses, they operate at layer 3 of the OSI model.