Signal Reconstruction Robust Beamforming Method via Presence of Desired Signal in Training Data

DU Zi-bing,WANG Lu,YANG Kun-de
DOI: https://doi.org/10.3969/j.issn.1006-9348.2013.04.053
2013-01-01
Abstract:Study robust beamforming problem.If the training data contain desired signals and the steering vector has errors,signal cancellation of MVDR will occur,as a result,this will cause performance degradation of MVDR beamforming.For this issue,a novel robust MVDR algorithm was proposed in the symmetric interference field.Firstly,the desired signal component was obtained by eliminating the real part of data covariance,and then desired signal covariance matrix was reconstructed and subtracted from the data covariance matrix.Finally,the weights of MVDR beamformer were computed using the remaining part which only contains interference and noise covariance information.Computer simulation results show that the method eliminates the adverse effect of desired signal and improves the robustness of MVDR beamformer against system errors.
What problem does this paper attempt to address?