Improved Root Sparse Bayesian Learning for DOA Estimation in Non-uniform Noise

Abstract

The vigorous development of sparse signal reconstruction (SSR) technology provides a new idea for realizing direction-of-arrival (DOA) estimation. This paper proposes an improved root sparse Bayesian learning algorithm to solve the problem of poor estimation accuracy of traditional DOA estimation algorithms based on SSR technology under off-grid error and non-uniform noise. The improved algorithm not only achieves accurate estimation of the non-uniform noise through a small number of iterations but also uses the expectation-maximization (EM) algorithm to iteratively refine the discrete sampling grid, which shows that the calculation of updating the grid points can be realized by the root of a particular polynomial. The simulation proves that the algorithm has excellent estimation performance under the coarse grid and non-uniform noise.