Sensor placement algorithms for target localization in sensor networks

This paper formulates and solves the sensor placement problem for efficient target localization in a sensor network. We illustrate our approach through a missile defense problem where the objective is the efficient localization of the missile using multiple sensors. We develop a mathematical framework for the localization of the missile using multiple sensors based on Cramer-Rao lower bound (CRLB) analysis. We derive the fisher information matrix to facilitate the evaluation of estimation accuracy. We develop an evolutionary algorithm for obtaining the sensor placements. Simulation results show that the evolutionary algorithm outperforms a greedy sensor placement algorithm and obtains sensor placements with very low estimation error.