Optimal Eigenspace-Based Widely Linear Beamformer for Noncircular Signals

This paper introduces an eigenspace-based widely linear (WL) beam former for noncircular signals. The conventional minimum variance distortion less response (MVDR) beam former becomes sub optimal when the signals are noncircular, and the optimal WL MVDR beam former has been proposed to gain the optimal outputs when the number of training samples is infinite. However, its performance may degrate for finite number of training samples. To circumvent this issue, the eigenspace-based WL beam former is proposed in this paper, which utilizes the eigenstructure of the correlation matrix to enhance the performance of the WL MVDR beam former. Moreover, it is proved that the eigenspace-based beam former not only can perform better under finite snapshots but also can handle a number of interferences greater than or equal to the number of sensors.