Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities

Rasmus Thomas Jones; Tobias A. Eriksson; Metodi Plamenov Yankov; Darko Zibar

doi:10.1109/ECOC.2018.8535453

Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities

Rasmus Thomas Jones
, Tobias A. Eriksson
, Metodi Plamenov Yankov
, Darko Zibar

Centre of Excellence for Silicon Photonics for Optical Communications

Japan National Institute of Information and Communications Technology

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

834 Downloads (Orbit)

Abstract

A new geometric shaping method is proposed, leveraging unsupervised machinelearning to optimize the constellation design. The learned constellationmitigates nonlinear effects with gains up to 0.13 bit/4D when trained with asimplified fiber channel model.

Original language	English
Title of host publication	Proceedings of the 44rd European Conference and Exhibition on Optical Communications (ECOC 2018)
Number of pages	3
Publisher	IEEE
Publication date	2018
ISBN (Print)	9781538648629
DOIs	https://doi.org/10.1109/ECOC.2018.8535453
Publication status	Published - 2018
Event	44th European Conference on Optical Communication - Fiera Roma, Rome, Italy Duration: 23 Sept 2018 → 27 Sept 2018

Conference

Conference	44th European Conference on Optical Communication
Location	Fiera Roma
Country/Territory	Italy
City	Rome
Period	23/09/2018 → 27/09/2018

Access to Document

10.1109/ECOC.2018.8535453

ECOC2018Accepted author manuscript, 250 KB

OpenUrl availability

Check availability in DTU Findit

Cite this

@inproceedings{017f862b5225491da8474504f5bb1bc7,

title = "Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities",

abstract = "A new geometric shaping method is proposed, leveraging unsupervised machinelearning to optimize the constellation design. The learned constellationmitigates nonlinear effects with gains up to 0.13 bit/4D when trained with asimplified fiber channel model.",

author = "Jones, \{Rasmus Thomas\} and Eriksson, \{Tobias A.\} and Yankov, \{Metodi Plamenov\} and Darko Zibar",

year = "2018",

doi = "10.1109/ECOC.2018.8535453",

language = "English",

isbn = "9781538648629 ",

booktitle = "Proceedings of the 44rd European Conference and Exhibition on Optical Communications (ECOC 2018)",

publisher = "IEEE",

address = "United States",

note = "44th European Conference on Optical Communication, ECOC 2018 ; Conference date: 23-09-2018 Through 27-09-2018",

}

Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities. / Jones, Rasmus Thomas; Eriksson, Tobias A.; Yankov, Metodi Plamenov et al.
Proceedings of the 44rd European Conference and Exhibition on Optical Communications (ECOC 2018). IEEE, 2018.

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities

AU - Jones, Rasmus Thomas

AU - Eriksson, Tobias A.

AU - Yankov, Metodi Plamenov

AU - Zibar, Darko

PY - 2018

Y1 - 2018

N2 - A new geometric shaping method is proposed, leveraging unsupervised machinelearning to optimize the constellation design. The learned constellationmitigates nonlinear effects with gains up to 0.13 bit/4D when trained with asimplified fiber channel model.

AB - A new geometric shaping method is proposed, leveraging unsupervised machinelearning to optimize the constellation design. The learned constellationmitigates nonlinear effects with gains up to 0.13 bit/4D when trained with asimplified fiber channel model.

U2 - 10.1109/ECOC.2018.8535453

DO - 10.1109/ECOC.2018.8535453

M3 - Article in proceedings

SN - 9781538648629

BT - Proceedings of the 44rd European Conference and Exhibition on Optical Communications (ECOC 2018)

PB - IEEE

T2 - 44th European Conference on Optical Communication

Y2 - 23 September 2018 through 27 September 2018

ER -

Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities

Abstract

Conference

Access to Document

OpenUrl availability

Fingerprint

Cite this