Analysis and Online Realization of the CCA Approach for Blind Source Separation

Author

Liu, Wei ; Mandic, Danilo P. ; Cichocki, Andrzej

Author_Institution

Univ. of Sheffield, Sheffield

Volume

18

Issue

5

fYear

2007

Firstpage

1505

Lastpage

1510

Abstract

A critical analysis of the canonical correlation analysis (CCA) approach in blind source separation (BSS) is provided. It is proved that by maximizing the autocorrelation functions of the recovered signals we can separate the source signals successfully. It is further shown that the CCA approach represents the same class of generalized eigenvalue decomposition (GEVD) problems as the matrix pencil method. Finally, online realizations of the CCA approach are discussed with a linear-predictor-based algorithm studied as an example.

Keywords

blind source separation; correlation methods; eigenvalues and eigenfunctions; matrix algebra; autocorrelation functions; blind source separation; canonical correlation analysis; generalized eigenvalue decomposition problems; linear-predictor-based algorithm; matrix pencil method; recovered signals; Autocorrelation; Blind source separation; Covariance matrix; Eigenvalues and eigenfunctions; Higher order statistics; Matrix decomposition; Signal processing; Signal processing algorithms; Source separation; Statistical analysis; Blind source separation (BSS); canonical correlation analysis (CCA); linear predictor; matrix pencil; second-order statistics (SOS); Algorithms; Artificial Intelligence; Computer Simulation; Models, Statistical; Pattern Recognition, Automated; Regression Analysis;

fLanguage

English

Journal_Title

Neural Networks, IEEE Transactions on

Publisher

ieee

ISSN

1045-9227

Type

jour

DOI

10.1109/TNN.2007.894017

Filename

4298121