Robust adaptive beamforming in the presence of strong desired signal and DOA mismatch

The performances of most previous beamformers degrade dramatically when the desired signal (DS) of high signal-to-noise ratio (SNR) is present in the training data with direction-of-arrival (DOA) mismatch. In this paper, two new robust adaptive beamformers are presented to achieve robustness over a large range of SNR. The basic ideas of the presented beamformers are based on the interference-plus-noise covariance matrix reconstruction and two different steering vector estimation methods. Compared with the previous steering vector estimation method, they have good performance while have less computational complexity and don´t need any specific optimization software. Simulation results demonstrate that the performances of the proposed beamformers are close to the optimal values over a large range of SNR.