A Sensor Selection Method for Nonlinear Localization in the Presence of Sensor Errors

This paper considers the localization problem of a mobile source based on time difference of arrival (TDOA) measurements and angle of arrival (AOA) measurements in the presence of random noises in both measurements and sensor location. We propose a sensor selection mechanism which aims to choose a subset of sensors to implement the multi-sensor passive localization. We use the covariance of the improved unscented Kalman filter (UKF) as a cost function where the dynamic model is augmented by incorporating the sensor positions into the state vector. Correspondingly, the number of sigma points in the improved UKF is also enlarged. Although the proposed method requires higher computational complexity, the selected subset achieves a better estimation performance in comparison with that of the linearization method EKF or classical UKF method which ignores the sensor position uncertainties. And the CEO (cross entropy optimization) is employed to solve the resulting complex combinatorial optimization model. Simulation experiments are conducted and simulation results demonstrate the efficiency of the proposed sensor selection scheme for multi-sensor passive localization.