Sparse Bayesian Learning Underwater Source Localization Method Based on Matched Beam

In recent years, compressed sensing algorithms have received great attention in underwater passive source localization, but the classic sparse Bayesian learning localization method does not take into account the situation of environmental mismatch. To solve this problem, this paper proposes sparse Bayesian learning underwater source localization method based on matched beam (MBP-SBL). The method converts the pressure data received by the array into the beam domain through beamforming technology, and improves its robustness and resolution through beam filtering and sparse iteration in the beam domain. Simulation data and sea trial data in north Elba island are used to verify the proposed method. The results show that, compared with the classical Bartlett, MBP, MV-EPC and MFP-SBL methods, the method proposed in this paper is more tolerant to environmental mismatch while ensuring the localization result of high resolution and low side lobe.