Abstract:
We propose a method for multi-sensor management, based on Rényi information gain, for tracking a maneuvering target. An interacting multiple model-cubature Kalman filter (IMM-CKF) algorithm is applied to obtain the maneuvering target's location estimation and the Rényi information gain of each sensor for the Gaussian non-linear system. A multi-sensor management model is established to select sensors according to their maximal system effectiveness, based on the estimated location of the maneuvering target. The simulation results show that, when compare with the Shannon entropy-based method, our proposed method achieves greater tracking accuracy and makes more effective use of sensor resources.