End-to-End Learning of Geometrical Shaping Maximizing Generalized Mutual Information

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Gümüs, Kadir, Alvarado, Alex, Chen, Bin, Häger, Christian, Agrell, Erik
Natura: Preprint
Pubblicazione: 2019
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914649863094272
author Gümüs, Kadir
Alvarado, Alex
Chen, Bin
Häger, Christian
Agrell, Erik
author_facet Gümüs, Kadir
Alvarado, Alex
Chen, Bin
Häger, Christian
Agrell, Erik
contents GMI-based end-to-end learning is shown to be highly nonconvex. We apply gradient descent initialized with Gray-labeled APSK constellations directly to the constellation coordinates. State-of-the-art constellations in 2D and 4D are found providing reach increases up to 26\% w.r.t. to QAM.
format Preprint
id arxiv_https___arxiv_org_abs_1912_05638
institution arXiv
publishDate 2019
record_format arxiv
spellingShingle End-to-End Learning of Geometrical Shaping Maximizing Generalized Mutual Information
Gümüs, Kadir
Alvarado, Alex
Chen, Bin
Häger, Christian
Agrell, Erik
Signal Processing
Artificial Intelligence
Information Theory
Machine Learning
GMI-based end-to-end learning is shown to be highly nonconvex. We apply gradient descent initialized with Gray-labeled APSK constellations directly to the constellation coordinates. State-of-the-art constellations in 2D and 4D are found providing reach increases up to 26\% w.r.t. to QAM.
title End-to-End Learning of Geometrical Shaping Maximizing Generalized Mutual Information
topic Signal Processing
Artificial Intelligence
Information Theory
Machine Learning
url https://arxiv.org/abs/1912.05638