Moving target beamforming based on Bayesian method

A recursive Bayesian approach to narrowband beamforming for a moving target is presented. The traditional Bayesian beamformer has a poor performance when dealing with moving targets because it can't use the observation signal recursively. To make use of motion information of the target, a motion model for the target is utilized to yield the predicted distribution of target's direction-of-arrival (DOA) for the new observation signal. The likelihood function of the observation signal is approximated based on the predicted DOA. The weight of minimum variance distortionless response (MVDR) beamformer for different directions in Bayesian beamformer is represented as the posteriori distribution that is computed by particle filter. Experiments show that the proposed method has an obvious improvement of output Signal-to-Interference-and-Noise-Ratio (SINR) under low Signal-to-Noise-Ratio (S-NR) compared with traditional method.