A Beamspace-Based Sparse Estimation Method for Array Signal

Abstract

In this paper, the problem of direction of arrival (DOA) estimation with sparse methods for array processing is concerned with the observation domain aspect, and an estimation method named beamspace-based sparse (BSE) is proposed. In BSE method, the beam space energy of the array signal is observed and modeled as the weighted sum of the signal energy of each azimuth beam pattern sequences of the conventional beamforming (CBF). BSE constructs a solution architecture for joint -norm minimization and quadratic constraint linear programming (QCLP) of noise power. Based on the estimation of noise background power under Gaussian noise conditions, a parameter selection method is derived, which can be quickly solved by the convex programming method. BSE has higher azimuth resolution and a lower false alarm rate when compared to sparse estimation methods based on other observation domains. It also performs well in coherent environments.