ICAR, a tool for Blind Source Separation using Fourth Order Statistics only

Abstract : The problem of blind separation of overdetermined mixtures of sources, that is, with fewer sources than (or as many sources as) sensors, is addressed in this paper. A new method, named ICAR (Independent Component Analysis using Redundancies in the quadricovariance), is proposed in order to process complex data. This method, without any whitening operation, only exploits some redundancies of a particular quadricovariance matrix of the data. Computer simulations demonstrate that ICAR offers in general good results and even outperforms classical methods in several situations: ICAR ~(i) succeeds in separating sources with low signal to noise ratios, ~(ii) does not require sources with different SO or/and FO spectral densities, ~(iii) is asymptotically not affected by the presence of a Gaussian noise with unknown spatial correlation, (iv) is not sensitive to an over estimation of the number of sources.
Liste complète des métadonnées

Cited literature [33 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00743890
Contributor : Pierre Comon <>
Submitted on : Sunday, October 21, 2012 - 5:09:59 PM
Last modification on : Monday, November 5, 2018 - 3:52:01 PM
Document(s) archivé(s) le : Tuesday, January 22, 2013 - 3:39:13 AM

File

Albera_ICAR_hal.pdf
Files produced by the author(s)

Identifiers

Citation

Laurent Albera, Anne Férreol, Pascal Chevalier, Pierre Comon. ICAR, a tool for Blind Source Separation using Fourth Order Statistics only. IEEE Transactions on Signal Processing, Institute of Electrical and Electronics Engineers, 2005, 53 (10), pp.3633-3643. ⟨10.1109/TSP.2005.855089⟩. ⟨hal-00743890⟩

Share

Metrics

Record views

696

Files downloads

397