Energy efficient malicious node detection scheme in wireless networks

Abstract

Detecting malicious nodes that causes packet dropping attacks with low routing overhead is important for the security of Ad hoc networks and also for high efficiency of the system. But the current malicious node detection mechanisms can detect collaborative attack such as blackhole attack but has high routing overhead and high end to end delay which affects the efficiency of the system. In this paper we design a Dynamic Source routing (DSR) based energy efficient malicious node detection scheme Extended Cooperative Bait Detection Scheme (ECBDS). This method is robust against collaborative attacks such as blackhole and grayhole and also reduces the routing overhead and end to end delay by using Dynamic Probabilistic Flooding (DPF) protocol. In this ECBDS scheme, the proactive and reactive defense architectures are integrated and random cooperation is done with the adjacent node. To bait malicious nodes, the address of the adjacent node is used as a bait destination address to send a reply message (RREP). The rebroadcast probability is determined to reduce the flooding of bait packets thereby reducing the routing overhead and malicious nodes are detected using a reverse tracing technique thereby prevents and ensures security. In order to improve the energy efficiency of the proposed detection scheme, the backbone node is included in the network. It combines stable routes, battery power and signal strength of the nodes to achieve the QoS parameters.