Source number estimation based on wavelet packet and singular value decomposition

In the applications of blind source separation, source number estimation is a hotspot and difficulty on the condition of multi-channel observation number less than independent source number. Here an algorithm of source number estimation based on wavelet packet and singular value decomposition is pro...

Full description

Saved in:
Bibliographic Details
Main Authors: Wenfeng Wu, Xunjia Su, Xiaohu Chen, Genliang Hou
Format: Conference Proceeding
Language:English
Subjects:
Computational modeling
Covariance matrix
Dominate eigenvalue
Eigenvalues and eigenfunctions
Estimation
Matrix decomposition
Noise
Numerical models
Singular value decomposition
Source number estimation
Wavelet packet analysis
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the applications of blind source separation, source number estimation is a hotspot and difficulty on the condition of multi-channel observation number less than independent source number. Here an algorithm of source number estimation based on wavelet packet and singular value decomposition is proposed. Especially it combines mechanism analysis, wavelet packet analysis and singular value decomposition. Computer simulation shows this method is satisfied to estimate independent source number on the condition of multi-channel observation number less than independent source number.
DOI:10.1109/ICICISYS.2010.5658439