Resolution enhanced Newtonized orthogonal matching pursuit solver for compressive beamforming

Abstract

Compared with conventional solvers for compressive beamforming (CB), Newtonized orthogonal matching pursuit (NOMP) can alleviate the basis mismatch problem. However, it needs to improve the performance of identifying closely-spaced sources. This paper proposes a resolution enhanced NOMP (RE-NOMP) solver for CB, which consists of three steps: initial source coordinate estimation, optimization of initial source coordinate estimation, and Newtonized refinement. Simulations and experiments demonstrate that RE-NOMP solver-based CB not only can effectively alleviate the basis mismatch problem but also achieves higher spatial resolution than the previous NOMP solver-based CB.