AN ASYMPTOTICALLY OPTIMAL APPROXIMATION OF THE CONDITIONAL MEAN CHANNEL ESTIMATOR BASED ON GAUSSIAN MIXTURE MODELS

Michael Koller, Benedikt Fesl, Nurettin Turan, Wolfgang Utschick

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

7 Zitate (Scopus)

Abstract

This paper investigates a channel estimator based on Gaussian mixture models (GMMs). We fit a GMM to given channel samples to obtain an analytic probability density function (PDF) which approximates the true channel PDF. Then, a conditional mean estimator (CME) corresponding to this approximating PDF is computed in closed form and used as an approximation of the optimal CME based on the true channel PDF. This optimal estimator cannot be calculated analytically because the true channel PDF is generally not available. To motivate the GMM-based estimator, we show that it converges to the optimal CME as the number of GMM components is increased. In numerical experiments, a reasonable number of GMM components already shows promising estimation results.

OriginalspracheEnglisch
Titel2022 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2022 - Proceedings
Herausgeber (Verlag)Institute of Electrical and Electronics Engineers Inc.
Seiten5268-5272
Seitenumfang5
ISBN (elektronisch)9781665405409
DOIs
PublikationsstatusVeröffentlicht - 2022
Veranstaltung2022 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2022 - Hybrid, Singapur
Dauer: 22 Mai 202227 Mai 2022

Publikationsreihe

NameICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Band2022-May
ISSN (Print)1520-6149

Konferenz

Konferenz2022 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2022
Land/GebietSingapur
OrtHybrid
Zeitraum22/05/2227/05/22

Fingerprint

Untersuchen Sie die Forschungsthemen von „AN ASYMPTOTICALLY OPTIMAL APPROXIMATION OF THE CONDITIONAL MEAN CHANNEL ESTIMATOR BASED ON GAUSSIAN MIXTURE MODELS“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren