Integrated Clustering based Intrusion Detection Model (IC-IDM) in Mobile Ad-hoc Networks

Abstract

Mobile Ad-hoc Networks (MANET) has become a significant and more exhilarating technology in recent days because of its rapid propagation of wireless devices. And also, MANETs are more susceptible to attacks due to its infrastructure less and dynamic nature. That is, the networks are not having centralized management and monitoring point. Intrusion Detection Systems (IDS) give auditing and monitoring abilities to the network that can provide local security to each node and also provides way to attain determined hopes on other nodes. Enhancement of the prior works on anomaly detection, Integrated Clustering Based Intrusion Detection Model (ID-IDM) combines the clustering procedure of networking with the IDS working structure to detect the intrusions with minimal computational overhead and minimal power consumptions. On considering that the nodes in MANET are power and battery constrained, an honest monitoring node must be avail to monitor and response against the intrusions periodically. For that, a novel procedure is also incorporated for evaluating the Node Honesty Rates (NHR) in the network. The nodes within the clusters are needed to monitor the overall network operations and additionally, node revocation procedure has also been included for the better performance of the proposed model. Integrated clustering is involved in detecting intrusions collectively since a single node in the network can neither perceive the attacks by the malicious node alone nor the malicious node has been revoked from the network. Hence, the IC-IDM results in précised intrusion detection with low processing and memory overhead irrespective of network traffic and nodes mobility. The performance of the affirmed model is validated using NS2 simulations based on clustering overhead, packet delivery ratio, energy utilization and power consumption.