Efficient Estimation for Sensor Biases and Target States in the Presence of Sensor Position Errors

The measurement biases of sensors during the target tracking process, if not accounted for, can significantly deteriorate the localization accuracy. This article focuses on the problem of target tracking using range and azimuth measurements obtained by asynchronous sensors when there are measurement biases and sensor position errors. A new online sequential estimation method is proposed by formulating the target states, sensor measurement biases, and the sensor positions as an augmented state vector. The filtering concepts and the minimum mean square error (MMSE) criterion are employed to perform real-time processing, which can provide simultaneous estimates for the augmented state vector. Comparing with the existing batch processing methods, the computational complexity is reduced by applying real-time estimations. Simulation validates the effectiveness of the proposed method in achieving the Bayesian Cramér-Rao lower bound (BCRLB) performance under the distance-dependent measurement noise.