A Hybrid Approach for Dynamic Intrusion Detection in Ad Hoc Networks Using Genetic Algorithm and Artificial Immune System

Mobile ad hoc network (MANET) is a self-created and self organized network of wireless mobile nodes. Due to special characteristics of these networks, security issue is a difficult taskto achieve. Hence, applying current intrusion detection techniques developed for fixed networks is not sufficient forMANETs. In this paper, we proposed an approach based on genetic algorithm (GA) and artificial immune system (AIS), called GAAIS, for dynamic intrusion detection in AODV-basedMANETs. GAAIS is able to adapting itself to network topology changes using two updating methods: partial and total. Eachnormal feature vector extracted from network traffic is represented by a hypersphere with fix radius. A set of sphericaldetector is generated using NicheMGA algorithm for covering the nonself space. Spherical detectors are used for detecting anomaly in network traffic. The performance of GAAIS isevaluated for detecting several types of routing attacks simulated using the NS2 simulator, such as Flooding, Blackhole, Neighbor,Rushing, and Wormhole. Experimental results show that GAAIS is more efficient in comparison with similar approaches.

کلیدواژه ها:

Ad hoc network ، Intrusion detection ، Routing attack ، Genetic algorithm ، Artificial immune system