An alternative to diagonal loading for implementation of a white noise array gain constrained robust beamformer

Diagonal loading is one of the most popular methods of robust adaptive beamforming, and the solution to many different problems aimed at producing beamformers which are robust to finite samples effects or/and steering vector errors. Among the latter, constraining the white noise array gain (WNAG) is a meaningful approach. However, relating the loading level to the desired WNAG is not straightforward. In this communication, using a generalized sidelobe canceler structure of the beamformer, we prove that the WNAG constraint can be encoded directly in the beamformer, and the latter can be obtained in a rather simple way from a specific eigenvector and without going through the diagonal loading step. (C) 2018 Elsevier B.V. All rights reserved.