A game-theoretic approach for combating node compromise attack in Wireless Sensor Network

Wireless Sensor Network(WSN), due to their popularity have become common in many domains like defense, traffic management, health applications, environment monitoring etc. With these wide range of applications and increasing popularity of WSN, that carry sensitive information are vulnerable to various threats and attacks. In this paper, an intelligent Intruder Detection System(IDS) is proposed that uses sophisticated data fusion technique, involving the use Neural Network as decision making tool and game theory model for the interaction between victim and attacker node. In specific Bayesian game theory is used in formulating the scenario of the intruder, in IDS. These combined effects make IDS robust to combat against the packets attempting to hack some sensitive information residing in the legitimate node. This paper presents the study and combined implementation of both game theory concept and Neural Network to function as an intelligent IDS, which is trained to produce optimal output strategy with varying parameters of intruder. The obtained results, shows that the defence mechanism can be considered as one of the solution concept towards security concerns of WSN.