Direct Multi-Target Tracking and Multisensor Fusion Using Antenna Arrays.
Marc Oispuu · 2009
Abstract: This paper investigates a direct Target Motion Analysis (TMA) estimator for the problem of calculating the states (i.e. source positions, velocities, etc.) of multi-ple sources from measurements made with multiple (fixed or moving) antenna arrays. We use the novel Subspace Data Fusion (SDF) approach and extend it to the multi-sensor case. In the SDF approach, subspaces are formed in a first pre-processing step from the raw antenna outputs. Then, the parameters of interest are estimated directly from a single cost function, which results from fusing all subspaces. This approach requires only a single low-dimensional optimization and completely circumvents the bearing data association problem inherent in traditional TMA approaches. We derive the Cramér-Rao Bound (CRB) for the direct multitarget tracking problem. In Monte Carlo simulations we find that the SDF estimator approaches the CRB and always per-forms better than or equal to the traditional TMA approach. We show that the state estimation accuracy can be improved by using multiple antenna arrays. 1