An algorithm for multisource beamforming and multitarget tracking

A new algorithm for simultaneous robust multisource beamforming and adaptive multitarget tracking is proposed. Self-robustness to locations errors or variations is introduced by a source-subspace-based tracking procedure of steering vectors in the array manifold. This LMS-type procedure is generalized from a former work we developed in the single source case. Two beamforming structures are actually proposed. The first is adaptive and optimal for uncorrelated sources and correlated noise. The second is conventional and optimal for correlated sources and uncorrelated white noise. The proposed algorithm and MUSIC show an identical asymptotic variance in localization for immobile sources, whereas for the mobile case, the proposed algorithm is highly advantageous. Then, it is shown that the additional use of some kinematic parameters (i.e., speed, acceleration, etc.) inferred from the reconstructed trajectories improves the tracking performance and overcomes some of the problems of crossing targets. The efficiency of multitarget tracking and the robustness of multisource beamforming are proved and then confirmed by simulation. The number of sources can be initialized and tracked by a marginal proposed procedure. The beamforming performance is shown to be optimal as the single source case. Finally, the algorithm has a very low order of arithmetic complexity.