基于多源域迁移学习的光纤非线性损伤补偿

陈志轩; 蔡炬; 张洪波; 张敏; 万峰; 杜杰; 刘娇; 张倩武

doi:10.13756/j.gtxyj.2025.230147

您当前的位置：

首页 >

文章列表页 >

基于多源域迁移学习的光纤非线性损伤补偿

光通信系统与网络技术 | 更新时间：2025-02-17

- 基于多源域迁移学习的光纤非线性损伤补偿
- Fiber Nonlinear Impairment Compensation based on Multi-source Domain Transfer Learning
- 光通信研究 2025年第1期
- 作者机构：
  
  1.成都信息工程大学　通信工程学院，成都　 610225
  2.上海大学　特种光纤与光接入网省部共建重点实验室，上海　 200072
- 作者简介：
  
  陈志轩（1999－），男，四川成都人。硕士，主要研究方向为高速光纤通信和深度学习。
  张洪波，讲师。E-mail：zhanghb@cuit.edu.cn
- 基金信息：
  
  四川省科技计划资助项目(2021YFG0149);上海市科委重点实验室资助项目(SKLSFO2019-06);高等学校学科创新引智计划资助项目(111;D20031)
- DOI：10.13756/j.gtxyj.2025.230147
  中图分类号： TN929
- 纸质出版日期：2025-02-10，
  
  收稿日期：2023-09-19，
  
  修回日期：2023-09-27，
- 稿件说明：
移动端阅览
陈志轩，蔡炬，张洪波，等. 基于多源域迁移学习的光纤非线性损伤补偿[J]. 光通信研究，2025(1): 230147.

Chen Z X, Cai J, Zhang H B, et al. Fiber Nonlinear Impairment Compensation based on Multi-source Domain Transfer Learning[J]. Study on Optical Communications, 2025(1): 230147.
陈志轩，蔡炬，张洪波，等. 基于多源域迁移学习的光纤非线性损伤补偿[J]. 光通信研究，2025(1): 230147. DOI： 10.13756/j.gtxyj.2025.230147.

Chen Z X, Cai J, Zhang H B, et al. Fiber Nonlinear Impairment Compensation based on Multi-source Domain Transfer Learning[J]. Study on Optical Communications, 2025(1): 230147. DOI： 10.13756/j.gtxyj.2025.230147.

摘要

【目的】

近年来基于神经网络（NN）的均衡器在光纤非线性损伤补偿中被广泛使用，但在实际应用中，需要消耗大量资源对NN进行再训练以适应新环境下的光通信系统。迁移学习通过将初始系统（源域）训练的NN模型中的一部分参数应用于新环境（目标域）下的NN模型，仅需使用少量的训练数据即可实现目标域模型的快速重构，但是该方法需要在所有源域中找到最佳源域进行迁移以获得良好的性能，且当目标域变化后需重新寻找最佳源域，这会消耗大量的训练资源。

【方法】

为解决此问题，文章提出了一种基于多源域迁移学习的方法，该方法将卷积神经网络（CNN）和双向长短期记忆（BiLSTM）作为均衡器，通过特定源域训练和多源域训练两个过程交替更新网络参数，然后针对新环境下的光通信系统进行微调，只需使用少量的初始训练数据即可适应传输系统的变化，且无需寻找最佳源域即可实现良好的性能。

【结果】

文章在5通道50-GBaud波分复用（WDM）双偏振16阶正交幅度调制（DP-16QAM）的光传输系统中进行了仿真验证，综合数值仿真结果表明，仅需使用35%的目标域数据进行多源域迁移学习，即可具有比再训练方法更好的性能。同时，与再训练和单源域迁移学习相比，多源域迁移学习的

因子分别提高了0.82和0.18 dB。

【结论】

因此，多源域迁移学习方案更适合于实际的光通信系统。

Abstract

【Objective】

In recent years

equalizer based on Neural Network (NN) has been widely used in optical fiber nonlinear impairment compensation. However

in practical application

it needs to consume a lot of resources to retrain NN to adapt to optical communication system in new environment. Transfer learning applies some parameters of the NN model trained by the initial system (source domain) to the NN model in the new environment (target domain). Only a small amount of training data is needed to achieve rapid reconstruction of the target domain model. However

this method needs to find the best source domain in all source domains for migration to obtain good performance. When the target domain changes

it is necessary to find the best source domain again

which will consume a lot of training resources. This work suggests a solution based on multi-source domain transfer learning to solve this issue.

【Methods】

This method employs Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) as equalizers. It alternately updates network parameters through two processes: specific source domain training and multi-source domain training. Subsequently

the optical communication system in the new environment is fine-tuned

allowing it to adapt to changes in the transmission system using only a small amount of initial training data. Moreover

good performance can be achieved without the need to search for the optimal source domain.

【Results】

A 5-channel 50-GBaud Wavelength Division Multiplexing (WDM) Dual-Polarization 16-order Quadrature Amplitude Modulation (DP-16QAM) optical transmission system is simulated to verify the effectiveness of the proposed method. The numerical simulation results show that the multi-source domain transfer learning outperforms the retraining method when using just 35% of the target domain data. Meanwhile

the

-factor of multi-source domain transfer learning are improved by 0.82 and 0.18 dB

respectively

in compared with retraining and single source domain transfer learning.

【Conclusion】

Therefore

the multi-source transfer learning scheme is suitable for practical optical communication systems.

关键词

Keywords

references

Hui R . Introduction to Fiber-optic Communications [M ] . Amsterdam : Academic Press , 2019 .

Zhang S , Yaman F , Nakamura K , et al . Field and Lab Experimental Demonstration of Nonlinear Impairment Compensation Using Neural Networks [J ] . Nature Communications , 2019 , 10 ( 1 ): 3033 .

Zhang F , Fang X , Chen X . Neural Network-based Fi-ber Nonlinearity Mitigation in High-speed Coherent Optical Transmission Systems [C ] // 2022 Optical Fiber Communications Conference and Exhibition (OFC) . San Diego, CA, USA : IEEE , 2022 : 9748180 .

Yadav G S , Chuang C Y , Feng K M , et al . Reducing Computation Complexity by Using Elastic Net Regularization based Pruned Volterra Equalization in a 80 Gbit/s PAM-4 Signal for Inter-data Center Interconnects [J ] . Optics Express , 2020 , 28 ( 26 ): 38539 － 38552 .

Schädler M , Böcherer G , Pachnicke S . Soft-Demap-ping for Short Reach Optical Communication: a Comparison of Deep Neural Networks and Volterra Series [J ] . Journal of Lightwave Technology , 2021 , 39 ( 10 ): 3095 － 3105 .

Ip E , Kahn J M . Compensation of Dispersion and Nonlinear Impairments Using Digital Backpropagation [J ] . Journal of Lightwave Technology , 2008 , 26 ( 20 ): 3416 － 3425 .

Gonçalves J , Martins C S , Guiomar F P , et al . Nonlinear Compensation with DBP Aided by a Memory Polynomial [J ] . Optics Express , 2016 , 24 ( 26 ): 30309 － 30316 .

蒙建宇 , 张洪波 , 张敏 , 等 . 基于IPCA-DNN算法的光纤非线性损伤补偿 [J ] . 光学学报 , 2021 , 41 ( 24 ): 2406002 .

Meng J Y , Zhang H B , Zhang M , et al . Fiber Nonlinear Impairments Compensation based on IPCA-DNN Algorithm [J ] . Acta Optica Sinica , 2021 , 41 ( 24 ): 2406002 .

Ming H , Chen X , Fang X , et al . Ultralow Complexity Long Short-term Memory Network for Fiber Nonlinearity Mitigation in Coherent Optical Communication Systems [J ] . Journal of Lightwave Technology , 2022 , 40 ( 8 ): 2427 － 2434 .

陈志轩 , 蔡炬 , 张洪波 , 等 . 基于LDBP的色散和光纤非线性损伤补偿 [J ] . 光通信研究 , 2024 ( 4 ): 230047 .

Chen Z X , Cai J , Zhang H B , et al . Chromatic Dispersion and Fiber Nonlinear Impairment Compensation based on LDBP [J ] . Study on Optical Communica-tions , 2024 ( 4 ): 230047 .

Tan C , Sun F , Kong T , et al . A Survey on Deep Transfer Learning [C ] // 27th International Conference on Artificial Neural Networks (ICANN 2018) . Rhodes, Greece : Springer , 2018 : 270 － 279 .

Zhang J , Xia L , Zhu M , et al . Fast Remodeling for Nonlinear Distortion Mitigation based on Transfer Learning [J ] . Optics Letters , 2019 , 44 ( 17 ): 4243 － 4246 .

Cheng Y , Zhang W , Fu S , et al . Transfer Learning Simplified Multi-task Deep Neural Network for PDM-64QAM Optical Performance Monitoring [J ] . Optics Express , 2020 , 28 ( 5 ): 7607 － 7617 .

Xu Z , Sun C , Ji T , et al . Feedforward and Recurrent Neural Network-based Transfer Learning for Nonlin-ear Equalization in Short-reach Optical Links [J ] . Journal of Lightwave Technology , 2021 , 39 ( 2 ): 475 － 480 .

Zhang W , Jin T , Xu T , et al . Nonlinear Mitigation with TL-NN-NLC in Coherent Optical Fiber Communications [C ] // 2020 Asia Communications and Pho-tonics Conference (ACP) and International Conference on Information Photonics and Optical Communications (IPOC) . Beijing, China : IEEE , 2020 : 9365253 .

Freire P J , Abode D , Prilepsky J E , et al . Transfer Learning for Neural Networks-based Equalizers in Coherent Optical Systems [J ] . Journal of Lightwave Technology , 2021 , 39 ( 21 ): 6733 － 6745 .

Nichol A , Achiam J , Schulman J . On First-order Me-ta-learning Algorithms [EB/OL ] . ( 2018-10-22 ) [ 2023-09-19 ] . http://arxiv.org/abs/1803.02999 http://arxiv.org/abs/1803.02999

Finn C , Abbeel P , Levine S . Model-agnostic Meta-learning for Fast Adaptation of Deep Networks [EB/OL ] . ( 2017-07-18 ) [ 2023-09-19 ] . https://arxiv.org/abs/1703.03400 https://arxiv.org/abs/1703.03400 .

Yang Y , Gao F , Zhong Z , et al . Deep Transfer Learning-based Downlink Channel Prediction for FDD Massive MIMO Systems [J ] . IEEE Transactions on Communications , 2020 , 68 ( 12 ): 7485 － 7497 .

浏览量

下载量

CSCD

CNKI被引量

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于LDBP的色散和光纤非线性损伤补偿

基于CNN-CBAM的虚假数据注入攻击辨识研究

光通信系统全生命周期能效探讨

混合连续—离散变量量子密钥分发

迁移学习辅助基于机器学习的传输质量评估