A New BSS Method of Single-Channel Mixture Signal Based on ISBF and Wavelet

Authors:
Xiefeng Cheng;Yewei Tao;Yufeng Guo;Xuejun Zhang
Affiliations:
Nanjing University of Posts & Telecommunications, Nanjing, P.R. China 210003 and Jinan University, Jinan, P.R. China 250100;Nanjing University of Posts & Telecommunications, Nanjing, P.R. China 210003 and Jinan University, Jinan, P.R. China 250100;Nanjing University of Posts & Telecommunications, Nanjing, P.R. China 210003;Nanjing University of Posts & Telecommunications, Nanjing, P.R. China 210003
Venue:
ISNN '08 Proceedings of the 5th international symposium on Neural Networks: Advances in Neural Networks, Part II
Year:
2008

Citing 3
Cited 0

Independent component analysis, a new concept?

Signal Processing - Special issue on higher order statistics
Wavelet regression for random or irregular design

Computational Statistics & Data Analysis
Independent component analysis: algorithms and applications

Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

A new BSS method based on independent sub-band function components (ISBF) and wavelet to separate single-channel mixture signal in noise was studied. Through obtaining sub-band functions with independent component characteristic in the time domain, 6-20 sub-band function by ICA were employed as the preparation knowledge for the blind source separation (BSS). By combining the independent sub-band function components (ISBF) into the single-channel mixture signal, a separation modeling of single-channel mixture signal was built based on ISBF. And the separation mathematics model of the single-channel signal in noise is investigated. The wavelet transform was used to eliminate the noise as well. Two simulation samples performed to verify the availability of the proposed methods. The results show that the methods played good role in one sensor source BSS, and had a capability to extract the sound signal feature.