Distributed Fusion of Local Probability Data Association Filters in Multi-Sensor Environment

The problem of data association for target tracking in a multi-sensor cluttered environment is discussed. The probabilistic data association filter (PDAF) is useful to obtain proper estimate of state in this environment. We propose two distributed algorithms for PDAF to acquire high accuracy system and reduce computation burden caused by clutter. The distributed process and its modified fusion algorithm for the PDAF is introduced, such as the optimal fusion formula (OFF) and covariance intersection (CI). The OFF is optimal in view of each local sensor and it has the great accuracy among the distributed fusion algorithms. On the other hands, the Cl has weighted convex combination without cross-covariance, so it has the advantage of fastness. Finally, the simulation results show that the proposed algorithms have advantages over robustness and lower computation burden.