Robust Adaptive Beamforming against Array Calibration Errors

Adaptive beamforming methods degrade in the presence of model mismatch. In this paper, we develop a modified interference covariance matrix reconstruction based beamformer that is robust against large array calibration errors. The calibration errors can come from the element position errors, and/or amplitude and phase errors, etc.. The proposed method is based on the fact that the sample covariance matrix can approximate the interference covariance matrix properly when the desired signal is small, and a reconstructed covariance matrix based on the Capon spectral will be better than the sample covariance matrix when the desired signal is large. A weighted summation of two covariance matrices in references is used to reconstruct the interference covariance matrix. Moreover, a computationally efficient convex optimization-based algorithm is used to estimate the mismatch of the steering vector associated with the desired signal. Several simulation cases are applied to show the superiority of the proposed method over other robust adaptive beamformers.