A New Passive Sonar Bearing Estimation Algorithm Combined with Blind Source Separation

Abstract

A new direction finding algorithm was proposed to improve the performance of bearing estimation of passive sonar system. In this paper, the proposed algorithm has three steps, firstly conventional beam former (CBF) is used to detect the targets' coarse bearings, secondly blind source separation (BSS) is used to make the interested beams of the targets become statistic independent, finally the proposed algorithm uses multiple signal classification (MUSIC) algorithm to locate the targets' precision directions based on the array-space data which reconstructed by BSS's output and CBF's coarse bearing information. Simulations result show that the proposed algorithm outperforms the beam-space MUSIC method and classic MUSIC algorithm in the aspect of angular resolution and bearing estimation accuracy. The results of sea trial show that the proposed algorithm can achieve better performance in weak target detection over classic MUSIC and beam-space MUSIC algorithms.