A Novel Intrusion Detection Algorithm: An AODV Routing Protocol Case Study

Mobile ad-hoc network (MANET) is a collection of movable nodes capable of self-routing, constraining energy and decentralized handling of nodes. It faces many challenges due to uncertainty of network topology i.e. security and congestion. In this paper we propose a novel algorithm for intrusion detection against attacks such as probing, Denial-of-service (DoS), vampire and User-To-Root (U2R) in a MANET environment. The attack detection has been carried out using a profile (behavior) analysis and a confusion matrix (True positives, True negatives, False positives, False negatives). The performance of a standard Ad hoc On-Demand Distance Vector (AODV) routing protocol has been reported for all 4 types of attack in a network simulator-2 (ns-2) environment. To the best of authors' knowledge, this is the first paper reporting a novel intrusion detection algorithm using behavior analysis for an AODV protocol in a MANET environment.