Improved Weight Clustering Algorithm for IDS in Mobile Ad hoc Network

An IDS is a defense system that detects hostile activities in a network and then tries to prevent such activities that may compromise system security. This paper studies on IDS architecture and clustering algorithm, and proposes an improved weight clustering algorithm for IDS in MANET with considering node credibility, link stability and fairness in cluster head competition. By using this algorithm, the whole network can be divided into some subclusters and carries out an effective intrusion detection mechanism, which performs distributed detection in the whole network and centralized decision-making at local node. In addition, it can reduce the computation load and the amount of bandwidth which is extremely important for MANET due to limited bandwidth and low communication speed in MAN-ET. The proposed algorithm is simulated via OPNET. Simulation results show that IWCA has better adaptability for MANET than other relative protocols, and be suitable for IDS in MANET.