Robust adaptive beamforming using interference covariance matrix reconstruction

Abstract

The performance of adaptive beamforming degrades severely when the strong desired signal is present in training snapshots with model mismatch. A robust adaptive beamforming is proposed using interference covariance matrix reconstruction in this paper. In the proposed method, the eigenvalue and eigenvector of desired signal is determined by calculating the correlation coefficients between eigenvectors of sample covariance matrix and the presumed array steering vector. Subsequently, the covariance matrix is reconstructed after removing the desired signal component from signal subspace. Finally, the average noise power is computed by estimating the noise subspace dimensions indirectly, and added to the reconstructed matrix in order to prevent the matrix from being singular. Compared with the conventional robust adaptive beamforming methods, the proposed method has improved performance and less computational complexity. Simulation results demonstrate the robustness and effectiveness of the proposed method.