Blind source separation method for multiple submarine feature signals in marine environment

Abstract

The invention discloses a blind source separation method for multiple submarine feature signals in a marine environment, which comprises the steps of collecting signal samples; performing clustering on mixed signal samples by adopting a clustering algorithm to obtain multiple clustering clusters based on a sample vector and a reference vector of a Hermitian angle in the time and frequency domain,and taking membership functions in the clustering algorithm to serve as masks; performing clustering verification based on the estimated cluster number; clustering the masks into Q mask clustering clusters based on a K-mean clustering algorithm; and performing signal separation through code estimation. The blind source separation method can solve a problem of blind source separation in an underdetermined condition, and the position of any mixed signal matrix or source is not required to be estimated in mask estimation, so that the complexity of signal separation is simplified, and the operation efficiency is improved.

Description

technical field [0001] The invention relates to the field of underwater acoustic signal processing, in particular to a method and system for blind source separation of characteristic signals of multiple submarines in a marine environment. Background technique [0002] There are many factors in the current blind separation algorithm. First, the blind separation algorithm is mainly used in the transient mixed form of the signal. There are relatively few blind separation algorithms in the signal convolution mixed form, which can be roughly divided into the following categories: 1) Based on the elimination of signal A signal separation network for cross-correlation between components that exploits feedback circuits to solve the problem. This algorithm has great defects in practical applications. When the scales of the signals differ greatly or the mixing matrix is ​​ill-conditioned, the separation effect is relatively poor, and even the signals cannot be separated. 2) Transform...

AI Technical Summary

Problems solved by technology

In addition, the above algorithm needs to estimate any mixed signal matrix or source position in the mask estimation, which makes the signal separation process relatively complicated and the operation efficiency is not high

Images