CLAIDS: Cellular Learning Automata Based Approach for Anomaly Nodes Detection in Clustered Mobile Ad Hoc Networks
Amirhosein Fathinavid and Maryam Ansari
Security is hard to achieve due to nodes’ dynamic nature in mobile ad hoc Networks. Routing protocols for MANETs are designed based on the assumption that all participating nodes are fully cooperative. In this paper, we design a misbehavior detection system for clustered wireless mobile ad hoc networks in forwarding packets with unknown traffic parameters. Therefore, to design such a protocol, we encounter three important problems, cluster formation, detection of misbehaving nodes, replacing well-behaving nodes instead of misbehaving nodes in routing phase. In this paper, we propose a five-phase algorithm to detect the malicious nodes based on Cellular learning automata. In our protocol, by the proposed clustering algorithm, the wireless nodes are grouped into clusters. Then, by the proposed method, each cluster-head can detect misbehaving nodes based on behavior and energy level of each node in packet forwarding in the clusters themselves. Finally, by the detection of misbehaving nodes, each cluster member can replace a well-behaving node of its neighbors instead of misbehaving node in existing path. We implemented the system in network simulator “GloMoSim” and MATLAB and we evaluated the performance of our proposed method by performing. Simulation results show that the system has good detection capabilities in finding malicious nodes in network.
Keywords: Mobile ad hoc networks; intrusion detection; cellular learning automata; GloMoSim simulator.