Decorrelated state estimation for distributed tracking using multiple sensors in cluttered environments

This paper extends a previously developed process of constructing decorrelated state estimates for distributed fusion architectures where each processor uses multiple sensors to maintain target tracks in cluttered environments. This construction removes correlation with previous intermediate updates from current intermediate state updates at each sensor stage of a sequential filter to form the decorrelated sequences. These sequences are in turn used as measurements in the joint probabilistic data association (JPDA) filter at the global processor. The algorithm to construct the decorrelated sequences is presented and applied to a proposed distributed fusion system where each local processor receives measurements from multiple sensors.