Automatic robust adaptive beamforming via ridge regression using l1-norm approximation

In this paper the l₁-norm approximation, used to measure the noise level in the generalized sidelobe canceler reparameterization of the standard Capon beamformer, is adopted in the ridge regression problem to compute the DL level. The enhanced covariance matrix obtained by the new DL approach becomes less noise sensitive and more robust in small snapshot size. The performance improvement of the proposed approach over the current robust adaptive beamforming techniques developed is confirmed by simulation results.