Direct Position Determination of Moving Sources Based on Delay and Doppler

In this paper, we address the problem of direct position determination (DPD) for moving sources. Different from traditional two-step localization methods where the intermediate time difference of arrival (TDOA) and frequency difference of arrival (FDOA) parameters are estimated and used for the subsequent moving source localization, in this paper the locations and velocities of moving sources are estimated directly from the received signals. To address the difficulty of estimating the high-dimensional unknown parameters simultaneously, a new multiple particle filter based DPD method is proposed, where the position and velocity are updated alternately using separate local particle filters. The proposed method requires smaller particle numbers compared to the classic particle filters, and its convergence is proved both theoretically and numerically. Moreover, the Cramer Rao lower bound for the considered moving source DPD is developed. Simulation results corroborate the effectiveness of the proposed method in terms of computational efficiency and estimation accuracy.