The subject of IP Mobility generally comes up first in wireless environments where end users devices could be connected to the network in any number of places. However, the problem is more general that this. A commuter train, for example, might offer its riders Internet access through either wired or wireless connections in the car. The train could then access the network through a series of radio transmitters located along the tracks, defaulting to a satellite connection whenever there is no local radio transmitter. This is an example of a mobile network, as opposed to a single mobile node.
The simplest solution to this problem is to use DHCP and allow the roaming devices to automatically obtain new and locally relevant IP addresses for whatever section of network they are connected to, as we discussed in Chapter 20.
The trouble with the DHCP approach, particularly with mobile nodes, is that it requires all of the applications on the roaming devices to drop all of their network connections, obtain a new address and potentially new server and domain information, and then re-establish communications. This is not always convenient.
A mobile network could, in principle, buffer packets from client devices while obtaining an address for the new network by DHCP. But the time required to get a new DHCP address and resynchronize routing tables is often far too long for seamless operation.
The strongest case for IP Mobility comes from mobile devices such as third-generation cellular ...