Title :
Blind Source Separation Using Generalized Singular Value Decomposition
Author_Institution :
Sch. of Autom., Guangdong Univ. of Technol., Guangzhou, China
Abstract :
A novel algorithm for blind separation of instantaneous linear mixtures of source signals is proposed. The proposed algorithm is based on the generalized singular value decomposition of a matrix pencil. We use a set of linear combinations of time-delayed correlation matrices to estimate the mixing matrix. Simulation results show that the proposed algorithm has better performance in accuracy comparing with TLS-ESPRIT algorithm.
Keywords :
blind source separation; correlation methods; singular value decomposition; blind source separation; matrix pencil; second-order statistics; singular value decomposition; time-delayed correlation matrices; Additive noise; Automation; Blind source separation; Eigenvalues and eigenfunctions; Higher order statistics; Information science; Matrix decomposition; Singular value decomposition; Source separation; Symmetric matrices;
Conference_Titel :
Information Science and Engineering (ICISE), 2009 1st International Conference on
Conference_Location :
Nanjing
Print_ISBN :
978-1-4244-4909-5
DOI :
10.1109/ICISE.2009.364
