Robust Adaptive Beamforming Based on Optimized Phase Response Constraint

It is well known that the performance of the minimum variance distortionless response (MVDR) beamformer is very sensitive to covariance matrix error and steering vector mismatch. In this paper, an extension of linear constraint minimum variance (LCMV) method that is robust against these non-ideal factors is proposed. This method imposes three-point linear constraints within the mainlobe region of beampattern such that the magnitude response in the mainlobe is unity while the corresponding phase response is optimally determined. The mainlobe of the proposed beampattern is well-maintained in terms of distortionless. Numerical results demonstrate that this method has excellent signal-to-interference-plus-noise ratio performance compared with the traditional methods in the presence of covariance matrix and steering vector errors.