Design of a Cross Layer Intrusion Detection System for Mobile Ad Hoc Networks to Mitigate Black Hole Attack

Abstract

Despite their advantages and ease of deployment, mobile wireless networks (MANETs) have security weaknesses due to their limited resources and mobile nature. They are vulnerable to many attacks, including the black hole attack, which has become difficult to detect accurately by traditional systems that rely on a single layer to identify malicious nodes. For example, the packet drop rate criterion at the network layer is not sufficient to identify malicious nodes, because sometimes packet drops are due to network conditions and not malicious action. To meet these challenges, it has become necessary to design a more accurate system to detect this type of attack by exploiting information extracted from several layers of the OSI model. This paper proposes the design of a cross layer intrusion detection system to mitigate black hole attacks in mobile wireless networks. The proposed approach consists of two levels. At the first level, data extracted from multiple layers is used as input to the fuzzy logic system to determine the nature of a particular node (malicious node, normal node or suspicious node). At the second level, if a node is suspicious, Dempster-Shafer theory will be applied to verify that node. The proposed approach, evaluated using the NS2 simulator, proves effective in detecting and isolating the malicious node.