Semi-Blind Source Separation for Memoryless Volterra Channels in UWB and its Uniqueness

Author

Petrochilos, Nicolas ; Witrisal, Klaus

Author_Institution

CReSTIC, Reims Univ.

fYear

2006

fDate

12-14 July 2006

Firstpage

566

Lastpage

570

Abstract

We study in this article the output of an autocorrelation receiver for a delay-hopped transmitted-reference ultra-wideband (UWB) communications systems with multiple access. Next to a linear multi-input multi-output (MIMO) channel, this receiver generates quadratic channels consisting of cross-product between original sources. Hence the received signal is the output of a multi-user (MU) MIMO Volterra channel, which can be ensured to be memoryless. We propose a novel blind algorithm to perform source separation and a semi-blind version of it. After demonstrating its effectiveness, we discuss the uniqueness problem for linear-quadratic channels with binary input

Keywords

MIMO systems; blind source separation; matrix algebra; multiuser channels; radio receivers; ultra wideband communication; MIMO channel; UWB communications systems; autocorrelation receiver; delay-hopped transmitted-reference ultra-wideband; linear multiinput multioutput; linear-quadratic channels; memoryless Volterra channels; multiuser channels; semi-blind source separation; Autocorrelation; Blind source separation; Data models; Delay; MIMO; Phase estimation; Sampling methods; Source separation; Ultra wideband communication; Ultra wideband technology;

fLanguage

English

Publisher

ieee

Conference_Titel

Sensor Array and Multichannel Processing, 2006. Fourth IEEE Workshop on

Conference_Location

Waltham, MA

Print_ISBN

1-4244-0308-1

Type

conf

DOI

10.1109/SAM.2006.1706197

Filename

1706197