Tow-Ship Interference Suppression Based on Blind Source Separation for Passive Sonar

We propose a new blind spatial correlation subtraction (BSCS) method combined with conventional beam former (CBF), blind source separation (BSS) and minimum variance distortion less response (MVDR) algorithm for tow-ship interference suppression and targets´ bearing estimation. In BSCS, firstly, CBF is utilized as a tow-ship noise estimator and the BSS based on frequency domain is used to extract acoustic contact components from the array data. Then, two steps that correlation and spectrum subtraction are performed between the CBF´s beams and the BSS´s outputs. Finally, for the sake of bearing estimation, the BSCS method maps the subtracted BSS´s output back to observation space, based on which classic MVDR method is used to obtain the interested targets´ bearings. The towed array simulation data and sea trial result reveal that the tow-ship interference suppression and weak targets detection capability of the proposed BSCS are superior to the classic MVDR method.