On the use of beamforming for estimation of spatially distributed signals

Abstract

In wireless communications, spatially distributed signals originate from local scattering around the transmitter. Spatially distributed signals cause conventional high-resolution methods to show degradation in performance, since the underlying data model is of full rank theoretically. In this paper, a relationship between the distribution of the spatial spread and the power azimuth spectrum is derived. Using this relationship, two simple and robust estimators based on the conventional beamformer and Capon's beamformer for estimation of the nominal direction and spatial spread are found. The proposed estimators show very good performance in numerical examples.