Location estimation of an intruder in wireless ad hoc networks

Abstract

Security within wireless ad hoc networks is hard to achieve due to the vulnerabilities in the data transmission process and the continuously changing network topology. An intrusion detection system is therefore essential to exclude any malicious nodes from the network. Furthermore, localization of the intruder helps in avoiding further attacks from the same user on the system. This paper proposes the application of a genetic algorithm to estimate the location of such malicious nodes. Simulation results show that with nine anchor nodes in a 1 km by 1 km area, the algorithm is capable of estimating the intruder’s position within a radius of 45 meters in 60% of the cases. Additionally, the accuracy of the algorithm improves as the number of anchor nodes increases.