Robust Null-Steering Beamforming Design under Steering Vector Mismatches

This paper is concerned with a robust null-steering beamforming design in the presence of unknown steering vector errors. First, in the scenario with a single interference, a convex cost function is constructed to develop a sub-filter, by optimizing the worst-case signal-to-scattered-interference-ratio gain (SSIRG). Next, a composite filter is deduced from the spatial convolution of multiple sub-filters, which is applicable for multi- interference scenarios. Finally, the obtained filter coefficients are further optimized to suppress the diffuse noise and the white noise. As a result, the proposed method not only is robust against the steering vector mismatches but also obtains a good trade-off among metrics including the SSIRG, the white noise gain, and the directivity factor. The validity of the proposed method is confirmed by numerical experiments.