[Oral Presentation]Approximate Joint Diagonalization for ARMA Dependent Source Separation - THE ELEVENTH IEEE SENSOR ARRAY AND MULTICHANNEL SIGNAL PROCESSING WORKSHOP (IEEE SAM2020)

Approximate Joint Diagonalization for ARMA Dependent Source Separation

ID：119 Submission ID：193 View Protection：ATTENDEE Updated Time：2020-08-05 10:17:28 Hits：346 Oral Presentation

Start Time：2020-06-08 15:00 （Asia/Shanghai）

Duration：20min

Session：[S] Special Session » [SS14] Dependent Source Separation

Video No Permission

Tips: The file permissions under this presentation are only for participants. You have not logged in yet and cannot view it temporarily.

Abstract

In this paper, an Approximate Joint Diagonalization (AJD) approach is proposed to separate dependent source signals. The diagonal structure of the Auto Regressive Moving Average (ARMA) matrix coefficients moves the problem from Blind Source Separation (BSS) to AJD one. The identified matrix coefficients of the observed signal are jointly diagonalized to achieve the mixture matrix identification. Simulation results are provided to illustrate the effectiveness of the proposed approach.

Keywords

Source Separation; Auto regressive moving average signal; Dependent sources; Approximate joint diagonalization

Speaker

Saliha Meziani

Ecole Militaire Polytechnique, Algeria

Submission Author

Saliha Meziani Ecole Militaire Polytechnique, Algeria

Adel Belouchrani Ecole Nationale Polythechnique, Algiers, Algeria

Karim Abed-Meraim University of Orleans & PRISME Lab., France