A Dummy Packet-Based Hybrid Security Framework for Mitigating Routing Misbehavior in Multi-Hop Wireless Networks

Wireless networks have emerged as a key enabling technology, expanding rapidly and offers numerous potential applications. The security issues have to be adequately addressed to realize the potential of multi-hop cooperative wireless network scenarios. The open nature of the wireless network is the opportunity for invaders launching various attacks with minimum effort owing to the multi-hop scenario of routing protocols. Routing protocols work based on the assumption that intermediate nodes are cooperating and well-behaving. Even in the presence of strong authentication mechanisms, detecting the continuous and selective packet dropping attack is a challenging process. This paper focuses on the design of a secure routing framework, and extend it to the routing protocols of various multi-hop wireless networks vulnerable to such attacks. The proposed hybrid security framework combats the routing misbehavior attacks in the presence of a wide range of malicious nodes. The proposed framework extends the hybrid security model adaptable to various multi-hop wireless networks with flexible routing overhead. The framework introduces a dummy packet based acknowledgment scheme that inserts dummy packets in the real payload traffic and masks the dummy traffic sequence through the dynamic traffic pattern. It optimizes the dummy packet generation based on the packet drop experienced and minimized the dummy traffic to balance the routing security and overhead. It confirms the presence of malicious nodes based on the dummy packet dropping and relies on the trust mechanism to eliminate the misbehaving nodes in the critical path. The use of subjective and fuzzy trust model validates the accuracy of uncertain evidence and contextual factors in the trust. The effectiveness of the framework is realized by applying it on various routing protocols in wireless networks. The performance evaluation confirms excellent packet delivery of the proposed hybrid framework over various networks in a highly vulnerable environment.