Decorrelated state estimation for distributed tracking of interacting targets in cluttered environments

This paper develops a process to construct decorrelated state estimates when tracking in cluttered environments using a distributed fusion architecture. This construction removes correlation with previous state estimates from the current updates in order to use the updates as measurements in the Kalman filter at the global processor. The effects of correlation with other interacting targets are also investigated for multiple target tracking. The algorithm to construct the decorrelated sequences is presented and applied on a particular distributed architecture where each processor receives measurements from only one sensor.