Intrusion detection system for mitigating attacks using energy monitoring in wireless sensor networks

Abstract

In wireless sensor network WSN, the existing literature works predicts the energy consumption using operational state transition model. In this method, each state should be known to the sink, in order to predict accurately. Also, in some cases, if the residual energy estimation is not based on the operational states of a sensor node, it may not reflect the exact amount of energy consumed by each node. In order to overcome these issues, in this paper, we propose to design an intrusion detection system to detect and mitigate the attacks using energy monitoring technique. In this technique, the malicious attacks are detected by monitoring the node's energy. The monitoring nodes predicts the energy using hidden Markov model HMM and compares it with the actual energy detected through an energy query technique. The detected attacks are classified based on the energy threshold levels. The attacks are then mitigated by establishing the attack-free route between source and destination, respectively. By simulation results, we show that the proposed technique increases the efficiency, reduces overhead and energy consumption.