In the existing wireless technology, routing support for mobile hosts is presently being understood as "Mobile IP" technology. Mobile IP is used to support mobile host roaming, where a mobile host connects to the network through various means other than its fixed-address domain space. On the other hand, the aim of a mobile ad-hoc network is to advance mobility into the chain of autonomous wireless domains, where a set of nodes themselves establishes the network routing structure in an ad-hoc manner.

A mobile ad-hoc network is defined as a collection of mobile nodes each combining the functionality of a router and a host. These mobile hosts are attached to a number of wireless mediums and are free to organize themselves. These hosts are equipped with wireless transmitters and receivers. Designing routing protocols for mobile ad-hoc networks has taken approaches similar to routing protocols used for conventional wired networks. Most of these protocols are based on table-driven or on-demand routing techniques.

The main task of any routing protocol is to aid communication within the network. The primary objective of a routing protocol is to establish correct routing between a source and a destination node to deliver messages in a timely fashion. In a mobile ad-hoc network environment, it may be necessary for one mobile host to assist other hosts in forwarding a packet to its destination. In a mobile ad-hoc network, to establish an active connection, the end node as well as the intermediate nodes can be mobile.

A node's mobility poses several scalability problems for mobile ad-hoc network protocols. When network topology changes, control messages are broadcast frequently between nodes, so that new routes can be formed and propagated throughout the network. It's expected that when nodes change topology there might be a short period where lots of control messages will propagate the network to broadcast the new destination paths. Minimizing control messages is necessary because of the additional load they place on the wireless links. Therefore, the techniques to reduce control messages must create a balance between, on the one hand, minimum amounts of messages by keeping network state information in each node, and on the other hand, flooding the network each time the topology changes. This could also lead to very slow network convergence where nodes contain either incomplete or expiry views of the network topology.