Une méthode MUSIC adaptée aux grands réseaux de capteurs


Por: Vallet, P, Loubaton, P, Mestre, X

Publicada: 1 ene 2016
Categoría: Electrical and electronic engineering

Resumen:
In the field of DoA estimation in array processing, we summarize recent results on an improved MUSIC method (referred to as G-MUSIC), developed in the context where the number of samples N of the observed signal is large, and of the same order of magnitude than the number of sensors. Using results from random matrix theory, we provide a statistical performance analysis of the G-MUSIC method, in terms of consistency, mean square error, and asymptotic normality, in the context of unknown deterministic source signals. A comparison with the classical MUSIC method is provided in a first scenario with widespace DoA, and in a second scenario with closely spaced DoA. © 2016 Lavoisier.

Filiaciones:
Vallet, P:
 Univ Bordeaux, CNRS, Bordeaux INP, Lab Integrat Mat Syst, 351 Cours Liberat, F-33405 Talence, France

Loubaton, P:
 Univ Paris Est, MLV, CNRS, Lab Informat Gaspard Monge, 5 Bd Descartes, F-77454 Marne La Vallee, France

Mestre, X:
 CTTC, Av Carl Friedrich Gauss, Barcelona 08860, Spain
ISSN: 19585608
Editorial
Lavoisier, 1041 RUE DES RESIDENCES, GRENOBLE, 38040, FRANCE, Francia
Tipo de documento: Article
Volumen: 33 Número: 2-3
Páginas: 223-248
WOS Id: 000382602400004
imagen Green Submitted, All Open Access; Green

MÉTRICAS