基于机器学习的光网络监测与优化方法

李鸿; 刘武; 罗鸣

doi:10.13756/j.gtxyj.2024.230046

您当前的位置：

首页 >

文章列表页 >

基于机器学习的光网络监测与优化方法

光通信系统与网络技术 | 更新时间：2024-05-30

- 基于机器学习的光网络监测与优化方法
- Optical Network Monitoring and Optimization Methods based on Machine Learning
- 光通信研究 2024年第3期页码：23004601-23004610
- 作者机构：
  
  a.中国信息通信科技集团有限公司　武汉邮电科学研究院有限公司，武汉　430074
  b.中国信息通信科技集团有限公司　光通信技术和网络全国重点实验室，武汉　430074
- 作者简介：
  
  李鸿（1996-），男，湖南平江人。硕士，主要研究方向为机器学习与光纤通信技术。
  刘武，正高级工程师。E-mail：liuwu@wri.com.cn
- 基金信息：
  
  国家重点研发计划资助项目(2022YFB2903303)
- DOI：10.13756/j.gtxyj.2024.230046
  中图分类号： TN929
- 纸质出版日期：2024-06-10，
  
  收稿日期：2023-08-08，
  
  修回日期：2023-10-20，
扫描看全文
李鸿，刘武，罗鸣.基于机器学习的光网络监测与优化方法[J].光通信研究，2024（3）：230046.

Li H, Liu W, Luo M. Optical Network Monitoring and Optimization Methods based on Machine Learning[J]. Study on Optical Communications, 2024（3）：230046.
李鸿，刘武，罗鸣.基于机器学习的光网络监测与优化方法[J].光通信研究，2024（3）：230046. DOI： 10.13756/j.gtxyj.2024.230046.

Li H, Liu W, Luo M. Optical Network Monitoring and Optimization Methods based on Machine Learning[J]. Study on Optical Communications, 2024（3）：230046. DOI： 10.13756/j.gtxyj.2024.230046.

摘要

近年来，许多新型调制和复用技术以及动态网络概念被提出，以适应不断提高的网络带宽和质量需求。网络控制平台向系统化、智能化趋势发展，要求网络管理者不断监测网络各项参数，时刻优化网络状态。然而，大范围部署额外的监测设备获取参数信息从成本控制的角度缺乏可行性，利用已知数据与特殊算法进行网络性能监测和优化是更优的选择。机器学习方法因为其足够准确和高效逐渐被学术界采纳用于完成上述任务。文章梳理了在光网络监测与优化任务中使用机器学习算法的不同应用场景，综述了该领域的研究成果，并提出了现存的基于机器学习的光网络监测与优化方法存在的问题及可能的进一步研究的方向。基于机器学习的光学性能监测包括光学损伤辨别、信道质量评估以及通道功率预测，基于机器学习的网络配置优化方法包括强化学习优化通道功率。进一步研究方向可以考虑加强与运营商的合作，使用真实的现场数据，不断获取数据动态训练模型，并使用迁移学习和数据增强等技术，以保证算法的鲁棒性与泛化能力。

Abstract

In recent years

many new modulation and multiplexing technologies and dynamic network concepts have been proposed to adapt to the ever-increasing network bandwidth and quality requirements. Network control platform has a systematic and intelli-gent development trend

which requires network managers to constantly monitor the parameters of the network and optimize the network state. However

it is not feasible to arrange additional monitoring equipment in a large range to obtain parameter informa-tion from the perspective of cost control. It is better to use known data and special algorithms to monitor and optimize network per-formance. Machine learning methods are increasingly adopted by the academic community because they are accurate and efficient enough to accomplish these tasks. This paper first reviews the different application scenarios of machine learning algorithms in op-tical network monitoring and optimization tasks. Then it reviews the research achievements in this field

and puts forward the ex-isting problems of machine learning-based optical network monitoring and optimization methods as well as the possible direction of future research. The optical performance monitoring based on machine learning includes failure identification

quality of transmis-sion estimation and channel power prediction. The network configuration optimization method based on machine learning includes reinforcement learning to optimize channel power. For future research direction

we believe that it is possible for researchers to use real data from network operators

newly collected data to dynamically train the model

and transfer learning and data enhancement techniques to ensure the robustness and generalization ability of the algorithm.

关键词

机器学习光学性能监测光网络优化神经网络强化学习

Keywords

machine learningoptical performance monitoringoptical network optimizationneural networkreinforcement learning

references

Musumeci F, Rottondi C, Nag A, et al. An Overview on Application of Machine Learning Techniques in Optical Networks[J]. IEEE Communications Surveys & Tutorials, 2018, 21(2): 1383-1408.

Saif W S, Esmail M A, Ragheb A M, et al. Machine Learning Techniques for Optical Performance Monitoring and Modulation Format Identification: A Survey[J]. IEEE Communications Surveys & Tutorials, 2020, 23(4): 2839-2882.

Skoog R A, Banwell T C, Gannett J W, et al. Automatic Identification of Impairments Using Support Vector Machine Pattern Classification on Eye Diagrams[J]. IEEE Photonics Technology Letters, 2006, 18(22): 2398-2400.

Anderson T B, Dods S D, Clarke K, et al. Multi-Impairment Monitoring for Photonic Networks[C]//ECOC 2007. Berlin, Germany: VDE, 2007: 1-4.

Jargon J A, Wu X, Willner A E. Optical Performance Monitoring Using Artificial Neural Networks Trained with Eye-Diagram Parameters[J]. IEEE Photonics Technology Letters, 2009, 21(1): 54-56.

Wu X, Jargon J A, Skoog R A, et al. Applications of Artificial Neural Networks in Optical Performance Monitoring[J]. Journal of Lightwave Technology, 2009, 27(16): 3580-3589.

Wu X, Jargon J A, Paraschis L, et al. ANN-based Optical Performance Monitoring of QPSK Signals Using Parameters Derived from Balanced-Detected Asynchronous Diagrams[J]. IEEE Photonics Technology Letters, 2010, 22(4): 248-250.

Ruiz M, Fresi F, Vela A P, et al. Service-Triggered Failure Identification/Localization Through Monitoring of Multiple Parameters[C]//ECOC 2016. Dusseldorf, Germany: VDE, 2016: 1136-1138.

Vela A P, Ruiz M, Fresi F, et al. Early Pre-FEC BER Degradation Detection to Meet Committed QoS[C]//OFC 2017. Los Angeles, CA, USA: IEEE, 2017: W4F.3.

Shahkarami S, Musumeci F, Cugini F, et al. Machine-Learning-based Soft-Failure Detection and Identifica-tion in Optical Networks[C]//OFC 2018. San Diego, CA, USA: IEEE, 2018: M3A.5.

Lun H, Liu X, Cai M, et al. GAN based Soft Failure Detection and Identification for Long-haul Coherent Transmission Systems[C]//OFC 2021. San Francis-co, CA, USA: IEEE, 2021: Th4J.2.

Jiménez T, Aguado J C, de Miguel I, et al. A Cognitive Quality of Transmission Estimator for Core Optical Networks[J]. Journal of Lightwave Technology, 2013, 31(6): 942-951.

Barletta L, Giusti A, Rottondi C, et al. QoT Estimation for Unestablished Lighpaths Using Machine Learning[C]//OFC 2017. Los Angeles, CA, USA: IEEE, 2017: Th1J.1.

Wass J, Thrane J, Piels M, et al. Gaussian Process Regression for WDM System Performance Prediction[C]//OFC 2017. Los Angeles, CA, USA: IEEE, 2017: Tu3D.7.

Lonardi M, Pesic J, Zami T, et al. The Perks of Using Machine Learning for QoT Estimation with Uncertain Network Parameters[C]//Photonic Networks and Devices 2020. Washington, DC, USA: Optica Publishing Group, 2020: NeM3B.2.

Pesic J, Lonardi M, Rossi N, et al. How Uncertainty on the Fiber Span Lengths Influences QoT Estimation Using Machine Learning in WDM Networks[C]//OFC 2020. San Diego, CA, USA: IEEE, 2020: Th3D.5.

Morette N, de Jauregui Ruiz I F, Pointurier Y. Leveraging ML-based QoT Tool Parameter Feeding for Ac-curate WDM Network Performance Prediction[C]//OFC 2021. San Francisco, CA, USA: IEEE, 2021: Th4J.4.

Morette N, de Jauregui Ruiz I F, Hafermann H, et al. On the Robustness of a ML-based Method for QoT Tool Parameter Refinement in Partially Loaded Networks[C]//OFC 2022. San Diego, CA, USA: IEEE, 2022: M3F.1.

Karandin O, Ferrari A, Musumeci F, et al. Low-Margin Optical-Network Design with Multiple Physical-Layer Parameter Uncertainties[C]//ECOC 2022. Ba-sel, Switzerland: IEEE, 2022: Mo3B.2.

Yang X, Ferrari A, Morette N, et al. QoT Estimation Improvement with Inputs Refinement Tool for C+L Networks[C]//OFC 2023. San Diego, CA, USA: IEEE, 2023: W4G.6.

Huang Y S, Samoud W, Gutterman C L, et al. A Machine Learning Approach for Dynamic Optical Channel Add/Drop Strategies that Minimize EDFA Power Excursions[C]//ECOC 2016. Dusseldorf, Germany: VDE, 2016: 268-270.

Gutterman C L, Mo W, Zhu S, et al. Neural Network based Wavelength Assignment in Optical Switching[C]//Big-DAMA'17.CA, USA: ACM, 2017: 37-42.

Zhu S X, Gutterman C L, Mo W, et al. Machine Learning based Prediction of Erbium-Doped Fiber WDM Line Amplifier Gain Spectra[C]//ECOC 2018. Rome, Italy: IEEE, 2018: Mo3E.

Ionescu M, Ghazisaeidi A, Renaudier J, et al. Design Optimisation of Power-Efficient Submarine Line Through Machine Learning[C]//CLEO 2020. San Jose, CA, USA: IEEE, 2020: STh4M.5.

Freire-Hermelo M, Sengupta D, Lavignotte A, et al. Dynamic Power Predistortion Implementation with Reinforcement Learning for Excursion-Free Amplified Optical Systems[C]//ONDM 2020. Castelldefels, Barcelona, Spain: IFIP Open Digital Library, 2020: PS.

Rottondi C, Barletta L, Giusti A, et al. Machine-Learning Method for Quality of Transmission Prediction of Unestablished Lightpaths[J]. Journal of Optical Communications and Networking, 2018, 10(2): A286-A297.

Huang Y S, Gutterman C L, Samadi P, et al. Dynamic Mitigation of EDFA Power Excursions with Machine Learning[J]. Optics Express, 2017, 25(3): 2245-2258.

Ishii K, Kurumida J, Namiki S. Wavelength Assignment Dependency of AGC EDFA Gain Offset Under Dynamic Optical Circuit Switching[C]//OFC 2014. San Francisco, CA, USA: IEEE, 2014: W3E.4.

Jiang Z P, Lin J, Hu H. Machine Learning based ED-FA Channel In-band Gain Ripple Modeling[C]//OFC 2022. San Diego, CA, USA: IEEE, 2022: W4I.2.

Wang Z H, Akinrintoyo E, Kilper D, et al. Optical Signal Spectrum Prediction Using Machine Learning and In-line Channel Monitors in a Multi-span ROADM System[C]//ECOC 2022. Basel, Switzerland: IEEE, 2022: We3B.5.

Ionescu M, Renaudier J, Ghazisaeidi A, et al. Optimization of Power Efficient Spatial Division Multiplexed Submarine Cables Using Adaptive Transponders and Machine Learning[J]. Journal of Lightwave Technology, 2022, 40(6): 1597-1604.

Freire-Hermelo M, Sengupta D, Lavignotte A, et al. Reinforcement Learning for Compensating Power Excursions in Amplified WDM Systems[J]. Journal of Lightwave Technology, 2021, 39(21): 6805-6813.

Abadi M, Barham P, Chen J, et al. Tensorflow: A System for Large-Scale Machine Learning[C]//OSDI 16. Savannah, GA, USA: USENIX, 2016: 265-283.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于GCNet的毫米波波束选择算法

深度学习在光纤通信系统损伤补偿中的应用

基于神经网络的编码压扩联合峰均比降低技术

光纤分布式传感系统在变电站维护中的应用

迁移学习辅助基于机器学习的传输质量评估