On a sparse component analysis approach to blind source separation

Abstract

Blind source separation has found applications in various areas including biomedical signal processing and genomic signal processing. Often, blind source separation is performed via independent component analysis (ICA) under the assumption of mutual independence among source signals. However, in bio-signal and genomic signal processing, the assumption of independence is often untrue, and the performance of the ICA approach is not so good. Much effort has been devoted to searching alternative approaches to blind source separation without the independence assumption. In this paper we present a sparse component analysis method, which exploits the sparseness of the source signals and makes the separated signals as sparse as possible according to a properly defined sparsity function, to reliably extract source signals from their mixtures. Some related theoretical and practical issues are investigated, with support and validation by simulation results. © Springer-Verlag Berlin Heidelberg 2006.