Self organization algorithm for unattended acoustic sensor networks in ground target tracking

In this paper we develop a self organization algorithm for unattended acoustic sensor networks used for ground target tracking. Instead of using sensor nodes with acoustic sensor arrays, the algorithm uses sensor nodes with a single acoustic sensor, which will greatly improve the flexibility of sensor network and reduce the cost and size of nodes. The self organization algorithm selects optimum group of sensor nodes to form a localization sensor group that minimize the localization error. The localization sensor group is dynamically updated along the tracking process to match the dynamics of the target. A Kalman filter based tracking method is proposed for improved performance. This combination will enable a highly accurate tracking using low cost, small size acoustic sensors. The simulation results confirm the effectiveness of the algorithm.