Sparse Bayesian DOA Estimation Based on Subspace Model

A sparse bayesian learning (SBL) DOA estimation algorithm based on a subspace model is proposed to improve the accuracy of the direction of arrival (DOA) estimation under coherent source conditions. Firstly, the subspace smoothing technique is combined with the weighted subspace fitting technique. This subspace model is then applied to the sparse Bayesian algorithm and the discrete grid points are considered as dynamic parameters and the grid point positions are updated iteratively using the Expectation-Maximization algorithm. Compared with traditional sparse algorithms, the proposed algorithm has significant advantages in terms of root mean square error and computational complexity, and has strong robustness to the estimation error of the number of signal sources.