An SVM-based Multiclass IDS for Multicast Routing Attacks in Mobile Ad Hoc Networks

Multicast routing protocols of Mobile Ad Hoc Networks are highly vulnerable to various internal stealthy attacks due to multicast group members. A cross-layer-based Intrusion Detection System exchanges features across layers to differentiate internal stealthy attackers from congested normal nodes of highly challenged networks. This proposed Intrusion Detection System uses the features of the MAC and the routing layers of two communicating nodes and their neighbors. Multiclass classification algorithms are efficient for MANET, where various internal attacking strategies exploit marginally different misbehaviors. In this work, well-known multiclass SVM techniques, such as a modified One-Vs-One and a One-Vs-All, are applied for the cross-layer-based of multicast communication of ad hoc networks. The proposed multiclass SVM-based, crosslayer IDS ensure high accuracy with minimum computation overhead to classify attackers in resource constrained networks like MANET.