Nonlinear Estimation Fusion Using Recursive Filtering Approach in Distributed Passive Sensor Networks

A recursive algorithm of nonlinear estimation fusion is presented to distributed passive sensor networks that use multiple maneuverable aircrafts with onboard direction finder to survey the certain area. The main issue addressed in this research is to construct the hierarchical architecture which consists of a group of local processors and a global processor to improve tracking accuracy. The tracking function is performed in both Reference Cartesian Coordinates (RCC) and Modified Spherical Coordinates (MSC). The state estimate is produced by each local processor that processes angle-only measurements using the MSC Extended Kalman Filter (EKF). In the global processor, the RCC Recursive Filter (RF) based on the weighted least squares estimate is developed by utilizing the EKF error covariance matrices, transformed from the MSC to the RCC, to obtain the RF gain for combining the output estimates of local processors. In comparison with passive tracking algorithms RF and EKF, the three typical scenarios, emitter location, target motion analysis and maneuvering target tracking, are investigated through simulations. The results show that the performance of proposed filter has dramatically improved.