An Energy Aware Trust Based Intrusion Detection System with Adaptive Acknowledgement for Wireless Sensor Network

Abstract

Today modern and innovative applications for health care environment based on Wireless Sensor Network (WSN) are being developed in the commercial sector. The emerging sensor networks are rapidly becoming a fundamental part of medical solutions due to increase in accessibility for healthcare professionals/patients and ultimately leading to reduced healthcare costs. Health watch is becoming popular, however securing data becomes critical as privacy issues are a major concern in wireless systems. Finding threats and blocking them without affecting the network is essential without increasing the processing overheads and energy. In this paper, a Trust Based Adaptive Acknowledgment (TRAACK) Intrusion-Detection System for WSN based on number of active successful deliveries, and Kalman filter to predict node trust is proposed. In TRAACK the entire route based trust strength for multi hop sensor networks is reverted to nodes acting as Security Agent. Based on trust value of entire route, Acknowledgement is initiated on select packets to decrease control overhead. It is observed that the packet delivery ratio for the proposed TRAACK method improves even when malicious nodes are present, the proposed technique TRAACK is able to detect malicious nodes and avoid them in the route discovery process. Simulations show improved performance in the presence of malicious nodes without compromise in energy.