An Improved Generalized Inverse Beamforming-Noise Source Localization Method Using Acoustic Vector Sensor Arrays

To resolve the radiated-noise sources of submerged submarines in shallow water, a new source localization method was developed based on generalized inverse beamforming (GIB). We chose a slender cylindrical shell to simulate the submarine, and a horizontal acoustic vector sensor array (AVSA) was applied to conduct source localization. Generally, the location of the horizontal array is close to the water bottom during measurements, and the acoustic reflection by the water bottom is the main factor degrading the performance of the source localization algorithms. To reduce or eliminate the water-bottom reflection, we proposed a new AVSA processing method, which is called combined signal processing (CSP). The proposed CSP method can form a unilateral directivity pattern that suppresses the image source signals without distorting the physical source signals. Subsequently, we introduced two iterative regularization matrices to the GIB algorithm, which contained a priori source information, and repeatedly penalized the iterative solution to improve the localization results. Numerical simulations and experiments were conducted to verify the proposed source localization method, the results of which suggest that it outperforms conventional source localization methods in terms of the dynamic range and spatial resolution of source maps, as well as the localization accuracy.