A robust adaptive beamforming algorithm based on covariance matrix reconstruction

Considering that the sample covariance matrix contains desired signal and look direction mismatch in steering vector exist,the performance of the traditional adaptive beamformer degrads greatly,this paper proposes a robust adaptive beamforming algorithm based on covariance matrix reconstruction.It divides the whole spatial area into several non-overlapping regions,corresponding to the interference region and desired signal region,respectively.The standard Capon beamformer is used to estimate the interference power,then the improved interference covariance matrix can be constructed by integrating power inside the interference region.Secondly,the signal covariance matrix is reconstructed by the beamspace MUSIC spectral estimation method using the standard Capon beamformer which integrates the signal power inside the desired signal region,with its main eigenvector as the desired signal steering vector.Thus,it avoids the performance degradation of the adaptive beamforming caused by the desired signal and angle mismatch.Theory analysis and simulation have shown that algorithm has good performance in the case of training data contains desired signal and angle mismatch.