Multiple target tracking using sensor signal and target state models

Two approaches to track multiple, independently moving targets have been investigated. The first, the sensor signal model algorithm, uses the signals measured from the targets by an array of sensors to derive a set of equations which, when solved, provide an updated estimate of the angles the targets have with respect to the sensor array axis. The algorithm has been demonstrated via a simulation of tracking of two targets. The simulation showed that, under the proper conditions, the sensor model algorithm provides a good performance in tracking the targets. The second algorithm is based on a target state model and employs a nonlinear measurement equation which is composed of the target positions. The measurement equation is defined in a manner that does not require that each position measured be correctly associated with the corresponding target. The measurement equation is defined in terms of functionals which are symmetric with respect to permutations of the measured target positions. An extended Kalman filter is used to estimate the target state vector defined by the targets' positions and velocities. The target state algorithm has been further developed to include multiple sensor systems and false and missed detection phenomena. Simulations of tracking of two and six targets are used to demonstrate the performance of the algorithm.