Widely Linear LQCMV Beamformer and Augmented Dual-Domain Adaptive Algorithm

A widely linear extension of the linearly and quadratically constrained minimum variance (LQCMV) beamformer is presented. By exploiting the pseudocovariance matrix which is complementary second order statistics for the ordinary covariance matrix, the widely linear LQCMV (WL-LQCMV) beamformer attains better performance when the received data is noncircular. Adaptive implementation of WL-LQCMV by the dual-domain adaptive algorithm (DDAA) brings a remarkable advantage of fast convergence. The key points of the convergence analysis of DDAA are elaborated. The simulation results are presented to show the efficacy of our approach.