A Geometric Method for Blind Separation of Digital Signals with Finite Alphabets

Author

Xi Chen ; Leyman, A.R. ; Jun Fang

Author_Institution

Dept. of Electr. & Comput. Eng., Nat. Univ. of Singapore, Singapore

Volume

3

fYear

2007

fDate

15-20 April 2007

Abstract

We consider the problem of blind separation of discrete sources with finite alphabets. More specifically, multiple-amplitude-shift-keying (M-ASK) alphabet and complex quadrature amplitude modulation (QAM) alphabet are studied. The proposed separation method exploits the geometry of the received data constellation. The method relies on a finite-step noniterative algorithm, and therefore it is free from any convergence problem. Numerical simulation results are included to illustrate the performance of the proposed algorithm.

Keywords

amplitude shift keying; blind source separation; geometry; quadrature amplitude modulation; blind digital signal separation; complex quadrature amplitude modulation alphabet; finite alphabets; finite-step noniterative algorithm; geometric method; multiple-amplitude-shift-keying alphabet; Convergence; Equations; Geometry; Iterative algorithms; Iterative methods; MIMO; Matrix converters; Numerical simulation; Quadrature amplitude modulation; Sensor arrays; Blind separation; MIMO; finite alphabets;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on

Conference_Location

Honolulu, HI

ISSN

1520-6149

Print_ISBN

1-4244-0727-3

Type

conf

DOI

10.1109/ICASSP.2007.366836

Filename

4217866