Intrusion Detection in Mobile Ad-Hoc Networks using Bayesian Game Methodology

Abstract

The dynamic and distributed nature of MANETs make them vulnerable to various types of attacks like black hole attack, traffic distortion, IP spoofing, DoS attack etc. Malicious nodes can launch attacks against other normal nodes and deteriorate the overall performance of the entire network [1�3]. Unlike in wired networks, there are no fixed checkpoints like router and switches in MANETs, where the Intrusion Detection System (IDS) can be deployed .However, due to limited wireless communication range and node mobility, nodes in MANET must cooperate with each other to provide networking services among themselves. Therefore, each node in a MANET acts both as a host and a router. Present Intrusion Detection Systems (IDSs) for MANETs require continuous monitoring which leads to rapid depletion of a node�s battery life. To avoid this issue we propose a system to prevent intrusion in MANET using Bayesian model based MAC Identification from multiple nodes in network. Using such system we can provide lightweight burden to nodes hence improving energy efficiency. Simulated results shows improvement in estimated delay and average bits transfer parameters