Robust adaptive beamforming algorithm based on neural network

Adaptive beamforming is known to have resolution and interference rejection capability when the array steering vector is precisely known. However, the performance of adaptive beamforming techniques may degrade severely in the presence of mismatches between the assumed array response and the true array response. In this paper, we propose a novel neural network approach to robust adaptive beamforming. The proposed algorithm is based on explicit modeling of uncertainty in the desired signal array response and a three-layer radial basis function neural network (RBFNN), which belongs to the class of diagonal loading approaches. In the proposed algorithm, the computation of the optimum weight vector is viewed as a mapping problem, which can be modeled using a RBFNN trained with input/output pairs. The proposed algorithm has nearly optimal performance under good conditions, provides the robustness against the signal steering vector mismatches and makes the mean output array SINR consistently close to the optimal one. Simulation results validate substantial performance improvements relative to other adaptive beamforming methods.