A blind source separation sorting algorithm for sonar echo signal based on frequency-point amplitude correlation

Due to reflection, delay and other factors, sonar echo signal usually has convolution mixing form. Frequency domain blind source separation can effectively solve the convolution mixed signal. But blind source separation must solve the sorting problem. According to the characteristics of large amplitude correlation of adjacent frequency points of echo signal. In this paper, proposing a frequency domain blind source separation sorting algorithm, which calculates and arranges the amplitude correlation of adjacent frequency points. Finally sorted the separated signals in frequency domain. And conducting blind source separation experiments on simulated target echo signals and sonar collected data. Experimental results showed the effectiveness of the proposed algorithm.