An Agent Based Intrusion Detection System for Wireless Network with Artificial Immune System (AIS) and Negative Clone Selection

Intrusion in Wireless network differs from IP network in a sense that wireless intrusion is both of packet level as well as signal level. Hence a wireless intrusion signature may be as simple as say a changed MAC address or jamming signal to as complicated as session hijacking. Therefore merely managing and cross verifying the patterns from an intrusion source are difficult in such a network. Beside the difficulty of detecting the intrusion at different layers, the network credential varies from node to node due to factors like mobility, congestion, node failure and so on. Hence conventional techniques for intrusion detection fail to prevail in wireless networks. Therefore in this work we device a unique agent based technique to gather information from various nodes and use this information with an evolutionary artificial immune system to detect the intrusion and prevent the same via bypassing or delaying the transmission over the intrusive paths. Simulation results show that the overhead of running AIS system does not vary and is consistent for topological changes. The system also proves that the proposed system is well suited for intrusion detection and prevention in wireless network.