DocumentCode :
2946361
Title :
Blind separation of signals from piece-wise constant gain mixtures using independent component analysis
Author :
Pullela, Hari ; Rajan, P.K.
Author_Institution :
Dept. of Electr. & Comput. Eng., Tennessee Technol. Univ., Cookeville, TN, USA
fYear :
2004
fDate :
2004
Firstpage :
462
Lastpage :
466
Abstract :
An algorithm is developed to separate signals mixed with piecewise constant matrices. Independent component analysis technique is employed to separate the signals. Algorithms are presented to identify the lengths of the segments and remove order, scaling and sign ambiguities present in the independent component analysis of consecutive segments.
Keywords :
blind source separation; independent component analysis; matrix algebra; piecewise constant techniques; blind separation; correlation coefficient; independent component analysis; piecewise constant matrices; relative scale factor; Attenuation; Concatenated codes; Data mining; Electroencephalography; Independent component analysis; Microphones; Neural networks; Signal analysis; Signal processing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
System Theory, 2004. Proceedings of the Thirty-Sixth Southeastern Symposium on
ISSN :
0094-2898
Print_ISBN :
0-7803-8281-1
Type :
conf
DOI :
10.1109/SSST.2004.1295700
Filename :
1295700
Link To Document :
بازگشت