Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent Communications

We propose a cost-effective and memory-aware end-to-end learning scheme utilizing bi-directional gated recurrent unit (bi-GRU) for geometric constellation shaping (GCS) under the first-order regular perturbation (FRP) auxiliary channel. The performance of the proposed system has been numerically ver...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhiyang Liu, Xiaoyu Liu, Shilin Xiao, Weiying Yang, Weisheng Hu
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Photonics Journal
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10365158/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832590399460343808
author Zhiyang Liu
Xiaoyu Liu
Shilin Xiao
Weiying Yang
Weisheng Hu
author_facet Zhiyang Liu
Xiaoyu Liu
Shilin Xiao
Weiying Yang
Weisheng Hu
author_sort Zhiyang Liu
collection DOAJ
description We propose a cost-effective and memory-aware end-to-end learning scheme utilizing bi-directional gated recurrent unit (bi-GRU) for geometric constellation shaping (GCS) under the first-order regular perturbation (FRP) auxiliary channel. The performance of the proposed system has been numerically verified at a 32 GBd 5-channel wavelength division multiplexing (WDM) 64 quadrature amplitude modulation (QAM) 800 km optical coherent communication system. Results show that the proposed bi-GRU based GCS scheme can achieve a performance gain over square 64QAM in mutual information (MI) with 0.12 bits/symbol and a Q-factor gain of 0.4 dB at optimal launched optical power. When transmission distance is extended to 1280 km, a generalized mutual information (GMI) gain of 0.136 bits/symbol is observed. Additionally, compared with the bi-directional long short-term memory (bi-LSTM) based GCS, the proposed bi-GRU scheme has lower computation complexity with similar system performance.
format Article
id doaj-art-733247933d7a45a581764e78c22be6bf
institution Kabale University
issn 1943-0655
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Photonics Journal
spelling doaj-art-733247933d7a45a581764e78c22be6bf2025-01-24T00:00:23ZengIEEEIEEE Photonics Journal1943-06552024-01-0116111010.1109/JPHOT.2023.334418410365158Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent CommunicationsZhiyang Liu0https://orcid.org/0000-0003-1990-028XXiaoyu Liu1https://orcid.org/0000-0002-8054-423XShilin Xiao2https://orcid.org/0000-0003-0887-782XWeiying Yang3https://orcid.org/0009-0002-1844-0867Weisheng Hu4https://orcid.org/0000-0002-6168-2688State Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, ChinaState Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, ChinaState Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, ChinaState Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, ChinaState Key Laboratory of Advanced Optical Communication Systems and Networks, Shanghai Jiao Tong University, Shanghai, ChinaWe propose a cost-effective and memory-aware end-to-end learning scheme utilizing bi-directional gated recurrent unit (bi-GRU) for geometric constellation shaping (GCS) under the first-order regular perturbation (FRP) auxiliary channel. The performance of the proposed system has been numerically verified at a 32 GBd 5-channel wavelength division multiplexing (WDM) 64 quadrature amplitude modulation (QAM) 800 km optical coherent communication system. Results show that the proposed bi-GRU based GCS scheme can achieve a performance gain over square 64QAM in mutual information (MI) with 0.12 bits/symbol and a Q-factor gain of 0.4 dB at optimal launched optical power. When transmission distance is extended to 1280 km, a generalized mutual information (GMI) gain of 0.136 bits/symbol is observed. Additionally, compared with the bi-directional long short-term memory (bi-LSTM) based GCS, the proposed bi-GRU scheme has lower computation complexity with similar system performance.https://ieeexplore.ieee.org/document/10365158/Bi-directional gated recurrent unitend-to-end learninggeometric constellation shaping
spellingShingle Zhiyang Liu
Xiaoyu Liu
Shilin Xiao
Weiying Yang
Weisheng Hu
Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent Communications
IEEE Photonics Journal
Bi-directional gated recurrent unit
end-to-end learning
geometric constellation shaping
title Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent Communications
title_full Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent Communications
title_fullStr Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent Communications
title_full_unstemmed Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent Communications
title_short Bi-GRU Enhanced Cost-Effective Memory-Aware End-to-End Learning for Geometric Constellation Shaping in Optical Coherent Communications
title_sort bi gru enhanced cost effective memory aware end to end learning for geometric constellation shaping in optical coherent communications
topic Bi-directional gated recurrent unit
end-to-end learning
geometric constellation shaping
url https://ieeexplore.ieee.org/document/10365158/
work_keys_str_mv AT zhiyangliu bigruenhancedcosteffectivememoryawareendtoendlearningforgeometricconstellationshapinginopticalcoherentcommunications
AT xiaoyuliu bigruenhancedcosteffectivememoryawareendtoendlearningforgeometricconstellationshapinginopticalcoherentcommunications
AT shilinxiao bigruenhancedcosteffectivememoryawareendtoendlearningforgeometricconstellationshapinginopticalcoherentcommunications
AT weiyingyang bigruenhancedcosteffectivememoryawareendtoendlearningforgeometricconstellationshapinginopticalcoherentcommunications
AT weishenghu bigruenhancedcosteffectivememoryawareendtoendlearningforgeometricconstellationshapinginopticalcoherentcommunications