A super-resolution beamforming method for autonomous underwater vehicle towed nest linear array based on Sparse Bayesian Learning

Compared with stationary sonar, Autonomous Underwater Vehicle (AUV) sonar has the advantages of strong maneuverability and low cost, which is more suitable for ocean intelligent unmanned systems. However, due to the limited size of the AUV platform, the aperture of the sonar hydrophone array is usually limited, and it is difficult to obtain high resolution by conventional beamforming (CBF) methods. In this paper, a super-resolution beamforming method based on Sparse Bayesian Learning (SBL) is introduced for the direction-of-arrival estimation (DOA) estimation of the AUV towed nest linear array. Based on the spatial sparsity of underwater acoustic radiation sources, the amplitudes of sources are approximated by using the hyperparameter variational method, and the super-resolution DOA estimation result is achieved by iterative solution of hyper-parameter and noise estimation. The simulation results show that the proposed method has higher angular resolution and lower sidelobe levels than the CBF when the array structure parameters are determined. In addition, it has a good performance of weak signal detection under strong interference background, which can be effectively applied to low-cost AUV sonar systems.