Robust blind beamforming using neural network

Many blind beamforming algorithms, such as C-CAB, use cyclostationarity to estimate the steering vector and adaptively obtain the LCMV optimum solution. However, LCMV methods are sensitive to the mismatch caused by the uncalibration array or estimate error. After discussion of this mismatch, a robust blind beamforming algorithm is presented in the paper. Implemented as a neural network, this algorithm reduces computational complexity for real-time use. Computer simulations verify the analysis.