Off-Grid Underdetermined DOA Estimation of Quasi-stationary Signals via Sparse Bayesian Learning

In the context of sparse reconstruction framework, direction of arrival (DOA) estimation of quasi-stationary signals (QSS) usually has difficulties in practical situations where true DOAs are not on the discretized sampling grid. In order to solve such an off-grid DOA estimation problem, this paper proposes novel DOA estimation strategy based on off-grid sparse Bayesian learning method. By using the Khatri-Rao transform, the virtual array aperture of uniform circular array is extended, thus the proposed method have the ability to achieve underdetermined DOA estimation. Then, an expectation-maximization iteration method is developed to estimate DOAs of QSS based on the off-grid model from a Bayesian perspective. Compared with state-of-the-art techniques, the proposed method do not need estimate parameters in performing the algorithms and has better estimation precision. Numerical simulations demonstrate the validity of the proposed method.