Robust blind source separation and dispersing algorithms

Pando Georgiev, Andrzej Cichocki

Research output: Contribution to journalConference articlepeer-review


We show that statistically independent source signals can be separated simultaneously, if for some time delays p they have nonzero cumulants cusi(p) = cu{si(k), si(k), si(k-p), si(k-p)}. If the sources have distinct cumulant functions, then the separation is possible with another procedure, which could be more effective for large scale problems. In both cases the problem of blind source separation can be converted to a symmetric eigenvalue problem of a generalized cumulant matrices, which are not sensitive to Gaussian noise. We propose new algorithms, based on the non-smooth optimization theory, which disperse the eigenvalues of these generalized cumulant matrices. We propose new orthogonalization procedure for the mixing matrix, which is robust to additive Gaussian noise.

Original languageEnglish
Pages (from-to)I/997-I/1000
JournalICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Publication statusPublished - 2002
Externally publishedYes
Event2002 IEEE International Conference on Acustics, Speech, and Signal Processing - Orlando, FL, United States
Duration: 13 May 200217 May 2002


Dive into the research topics of 'Robust blind source separation and dispersing algorithms'. Together they form a unique fingerprint.

Cite this