Desigen and Simulation of a Routing Algorithm in Mobile Ad-Hoc Networks based on Artificial Neural Networks

Abstract
Wireless Ad Hoc networks are among different wireless networks which are extensively studied in the past few years. Ad Hoc networks consist of several mobile nodes with alternate locations. Hence, such a network, has to employ resilient routing algorithms to manage node movement and correct routing in the presence of broken links and outage of availability seamless to users of any two communicating nodes.
In this paper different routing strategies based on neural network techniques are studied. Then a new algorithm based in shortest path method and Hopfield neural network is proposed. The proposed algorithm is then simulated using NS-2 network simulator. Numerical results show that the new algorithms can outperform traditional DSR and AODV and has shorter traffic overhead.

Keywords: Ad-hoc Networks, Routing overhead, end-to-end delay, NS-2 network simulator