AN EFFICIENT ATTACK DETECTION SYSTEM FOR MOBILE AD-HOC NETWORK

A mobile ad hoc network (MANET) is a wireless net-work that does not rely on any fixed infrastructure (i.e., routing facilities, such as wired networks and access points), and whose nodes must coordinate among themselves to de-termine connectivity and routing. The traditional way of protecting networks is not directly applicable to MANETs. Many conventional security solutions are ineffective and inefficient for the highly dynamic and resource-constrained environments where MANET use might be expected. Since prevention techniques are never enough, intrusion detection systems (IDSs), which monitor system activities and detect intrusions, are generally used to complement other security mechanisms. How to detect intrusions effectively and effi-ciently on this highly dynamic, distributed and resource-constrained environment is a challenging research problem. In this paper, we investigate the use of evolutionary compu-tation techniques for synthesizing intrusion detection pro-grams on MANETs. We evolve programs to detect the fol-lowing attacks against MANETs: dropping attacks and pow-er consumption attack. The proposed system is a novel ar-chitecture that uses knowledge-based intrusion detection techniques to detect the attacks that an adversary can per-form against the routing fabric of mobile ad hoc networks. Moreover, the system is designed to take countermeasures to minimize the effectiveness of an attack and keep the perfor-mance of the network within acceptable limits. The novelty of the system lies in the usage of timed finite state machines that enable the real-time detection of attacks. Our system does not introduce any changes to the underlying routing protocol and operates as an intermediate component between the network traffic and the routing protocol. The system was developed and tested to operate in AODV-enabled networks. Our experimental results compare with normal AODV, un-der attack AODV and the results is more efficient than exist-ing works.