Passive Cooperative Multi-Trajectory Tracking Using OFDM Waveforms

This paper investigates the passive multiple trajectories tracking with multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) signals transmitted by the base station (BS) in the rich scattering environment. In the proposed framework, multiple radar receivers track multiple trajectories cooperatively with the assistance of a fusion center. To cope with high dimensional and superimposed echoes received by each radar receiver, we propose a parameter estimation method based on Bayesian learning to obtain the coarse estimations of targets and clutter. Then, these estimations are utilized as inputs for the trajectory probability hypothesis density (TPHD) filter for trajectory tracking. Considering the potential for a target to generate multiple measurements in overlapping areas of sweeping beams and non-Poisson characteristics of the estimated clutter/false measurements, we derive the multi-detection TPHD (MD-TPHD) filter with non-Poisson clutter and its Gaussian mixture implementation, where the clutter is separated into a Poisson and a discrete distributed component. As the field of view (FoV) of each radar receiver is finite, we fuse the trajectories in a fusion center utilizing Arithmetic Average (AA) method. The fused TPHDs are then fed back to each radar receiver to assist more accurate tracking. Simulation results demonstrate the superiority of MD-TPHD filter assisted by trajectory fusion in the multi-detection and FoV limited scenarios.