Energy Efficient Intrusion Detection Scheme Based on Bayesian Energy Prediction in WSN

Abstract

Wireless sensor network (WSN) has wide range of application like monitoring the environment, military, health application etc. Wireless sensor network has limited energy and resource, so challenging task in WSN is to design a network in such a way that maximize the lifetime of network. WSN are harmed or damaged by the Denial of Service (DoS) attack which destroy the network, resources and lose its energy rapidly. Various IDS used to detect malicious node in the network but they consume more energy to monitor malicious node, so decrease the network lifetime and throughput. It is important to form an energy efficient IDS which detect intruder accurately and consume less energy. In this paper, EEIDS (Energy Efficient Intrusion Detection Scheme) is proposed and designed, which detect malicious node based on energy consumption of nodes by comparing actual and predicted energy. The node with abnormal energy detected as malicious node. In EEIDS, Bayesian approach is used for energy prediction of sensor nodes, in which energy consumption of each sensor node is predicted using prior information and likelihood function also energy efficient approach used to reduce energy consumption of network. The simulation results show that EEIDS gives better network lifetime, throughput and energy consumption and effectively detect malicious node.