Direction of arrival estimation under Class A modelled noise in shallow water using variational Bayesian inference method

The shallow water noise shows obvious impulsive property, which greatly degrades the direction of arrival (DOA) performance due to the conventional design concept based on the Gaussian assumption. In this paper, DOA estimator in presence of impulsive noise utilising variational Bayesian inference is proposed. The Middleton's Class A noise model is considered as a typical underwater noise model to analyse the performance of DOA estimation. The DOA estimation problem is modelled as the sparse signal recovery problem, and the hierarchical Bayesian learning framework is formulated by considering the common sparsity of signal and the element-wise sparsity of the impulsive noise. The variational Bayesian inference realises the posterior estimation of signal and impulsive noise components. To mitigate the basis mismatches, the root sparse Bayesian learning method is applied to refine the steering vectors. Simulations verify the advantages of the proposed DOA method in terms of spatial resolution, root mean square error, accuracy, and robustness compared with the state-of-the-art benchmarks in the presence of Middleton's Class A noise.