An Algorithm for Fixed Single Observer Passive Tracking Based on Iterative Grid Search

The passive location and tracking of a moving target emitter with fixed single observer is a new research direction in the field of electronic warfare. The passive tracking algorithm is one of the key techniques. It is of great importance to develop a robust and fast tracking algorithm in the fixed single observer passive location and tracking system because of its inherent disadvantage of weak observability and large initial error. The process of passive tracking of a moving target with fixed single observer is a typical nonlinear filter process. Extended kalman filter is a traditional method for the nonlinear filter problem. Extended kalman filter processes the first order linearization for nonlinear model in the estimation which included linearization error. Based on the extended kalman filter, many improved algorithms are proposed to reduce or avoid the effect of the linearization error. But the extended kalman filter and its improved algorithms depend on the choice of the initialization state, and need compute the Jacobian matrix no exception and did not comb-out the effect of the linearization error. Therefore, an algorithm for fixed single observer passive tracking based on iterative grid search is proposed by using spatial-frequency domain information such as angle, angular velocity and Doppler frequency rate-of-change in this paper. This algorithm is developed from grid search but has less computation in same estimation resolution. Compared with usual algorithms used in passive tracking such as extended kalman filter, modified covariance extended kalman filter and unscented kalman filter, the proposed algorithm is not affected by state initialization error and linearization error. If the choice of error threshold and grid search time is in reason, the last estimation would approach the global minimum mean square error. The correctness as well as validity of the algorithm is also showed through numerical simulation and experiment results.