Energy efficient learning solution for intrusion detection in Wireless Sensor Networks

Abstract

The protocols in Wireless Sensor Networks (WSN) have a unique requirement for being of low complexity and energy-efficient. Due to their possible deployment in remote locations for civil, educational, scientific and military purposes, security, which includes intrusion detection and intrusion prevention, is of utmost importance. In this paper, we propose a simple, low complexity and energy-aware protocol for intrusion detection in WSN. The protocol is self-learning and distributed in nature. The distributed nature avoids all other nodes being sacrificed when a single node is compromised. The protocol juxtaposes the concept of stochastic learning automata on packet sampling mechanism to achieve an energy aware intrusion detection system. We have rigorously evaluated the performance of our proposed solution by performing a variety of experiments and have found our solution approach to be promising.