Intelligent black hole detection in mobile AdHoc networks

Abstract

Security is a critical and challenging issue in MANET due to its open-nature characteristics such as: mobility, wireless communications, self-organizing and dynamic topology. MANETs are commonly the target of black hole attacks. These are launched by malicious nodes that join the network to sabotage and drain it of its resources. Black hole nodes intercept exchanged data packets and simply drop them. The black hole node uses vulnerabilities in the routing protocol of MANETS to declare itself as the closest relay node to any destination. This work proposed two detection protocols based on the collected dataset, namely: the BDD-AODV and Hybrid protocols. Both protocols were built on top of the original AODV. The BDD-AODV protocol depends on the features collected for the prevention and detection of black hole attack techniques. On the other hand, the Hybrid protocol is a combination of both the MI-AODV and the proposed BDD-AODV protocols. Extensive simulation experiments were conducted to evaluate the performance of the proposed algorithms. Simulation results show that the proposed protocols improved the detection and prevention of black hole nodes, and hence, the network achieved a higher packet delivery ratio, lower dropped packets ratio, and lower overhead. However, this improvement led to a slight increase in the end-to-end delay.