Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Blind Source SeparationDependent Component Analysis Using Time-Frequency Analysis

Blind Source Separation: Dependent Component Analysis Using Time-Frequency Analysis [Sparsity is an important property shared by many kinds of signals in numerous practical applications. These signals are sparse to some extent in different representation domains, such as time domain, frequency domain or time-frequency domain. In recent years, sparsity has been widely exploited to solve the problem of underdetermined blind source separation (UBSS), where the number of sources exceeds that of the observed mixtures. In fact, the sparsity assumption can also be satisfied by some dependent source signals. For these signals, it is possible to find a number of areas in some representation domains, where the source signals are not active, that is, signals are sparse in theses areas. The sparsity property provides a possibility for the blind separation of dependent sources. In this chapter, the sparsity of dependent sources in the time-frequency (TF) domain will be exploited to achieve blind source separation, where time-frequency analysis (TFA) will be used as a powerful tool for dependent component analysis (DCA). We will also show that for those non-sparse signals whose auto-source points and cross-source points do not overlap in the TF plane, they can be separated by using TFA if the underdetermined mixing system is known.] http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png

Blind Source SeparationDependent Component Analysis Using Time-Frequency Analysis

Download PDF
23 pages

Loading next page...
 
/lp/springer-journals/blind-source-separation-dependent-component-analysis-using-time-RDt87twIZL
Publisher
Springer Singapore
Copyright
© The Author(s) 2015
ISBN
978-981-287-226-5
Pages
49 –71
DOI
10.1007/978-981-287-227-2_3
Publisher site
See Chapter on Publisher Site

Abstract

[Sparsity is an important property shared by many kinds of signals in numerous practical applications. These signals are sparse to some extent in different representation domains, such as time domain, frequency domain or time-frequency domain. In recent years, sparsity has been widely exploited to solve the problem of underdetermined blind source separation (UBSS), where the number of sources exceeds that of the observed mixtures. In fact, the sparsity assumption can also be satisfied by some dependent source signals. For these signals, it is possible to find a number of areas in some representation domains, where the source signals are not active, that is, signals are sparse in theses areas. The sparsity property provides a possibility for the blind separation of dependent sources. In this chapter, the sparsity of dependent sources in the time-frequency (TF) domain will be exploited to achieve blind source separation, where time-frequency analysis (TFA) will be used as a powerful tool for dependent component analysis (DCA). We will also show that for those non-sparse signals whose auto-source points and cross-source points do not overlap in the TF plane, they can be separated by using TFA if the underdetermined mixing system is known.]

Published: Sep 17, 2014

Keywords: Time-frequency analysis; Spatial time-frequency distribution; Quadratic TFD

There are no references for this article.