Multi-wavelength optical information processing with deep reinforcement learning

Abstract Multi-wavelength optical information processing systems are commonly utilized in optical neural networks and broadband signal processing. However, their effectiveness is often compromised by frequency-selective responses caused by fabrication, transmission, and environmental factors. To mit...

Full description

Saved in:
Bibliographic Details
Main Authors: Qiuquan Yan, Hao Ouyang, Zilong Tao, Meili Shen, Shiyin Du, Jun Zhang, Hengzhu Liu, Hao Hao, Tian Jiang
Format: Article
Language:English
Published: Nature Publishing Group 2025-04-01
Series:Light: Science & Applications
Online Access:https://doi.org/10.1038/s41377-025-01846-6
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850181229369884672
author Qiuquan Yan
Hao Ouyang
Zilong Tao
Meili Shen
Shiyin Du
Jun Zhang
Hengzhu Liu
Hao Hao
Tian Jiang
author_facet Qiuquan Yan
Hao Ouyang
Zilong Tao
Meili Shen
Shiyin Du
Jun Zhang
Hengzhu Liu
Hao Hao
Tian Jiang
author_sort Qiuquan Yan
collection DOAJ
description Abstract Multi-wavelength optical information processing systems are commonly utilized in optical neural networks and broadband signal processing. However, their effectiveness is often compromised by frequency-selective responses caused by fabrication, transmission, and environmental factors. To mitigate these issues, this study introduces a deep reinforcement learning calibration (DRC) method inspired by the deep deterministic policy gradient training strategy. This method continuously and autonomously learns from the system, effectively accumulating experiential knowledge for calibration strategies and demonstrating superior adaptability compared to traditional methods. In systems based on dispersion compensating fiber, micro-ring resonator array, and Mach-Zehnder interferometer array that use multi-wavelength optical carriers as the light source, the DRC method enables the completion of the corresponding signal processing functions within 21 iterations. This method provides efficient and accurate control, making it suitable for applications such as optical convolution computation acceleration, microwave photonic signal processing, and optical network routing.
format Article
id doaj-art-6c6ee048087142db8c4c9e3dfa5a2fcc
institution OA Journals
issn 2047-7538
language English
publishDate 2025-04-01
publisher Nature Publishing Group
record_format Article
series Light: Science & Applications
spelling doaj-art-6c6ee048087142db8c4c9e3dfa5a2fcc2025-08-20T02:17:57ZengNature Publishing GroupLight: Science & Applications2047-75382025-04-0114111210.1038/s41377-025-01846-6Multi-wavelength optical information processing with deep reinforcement learningQiuquan Yan0Hao Ouyang1Zilong Tao2Meili Shen3Shiyin Du4Jun Zhang5Hengzhu Liu6Hao Hao7Tian Jiang8College of Computer Science and Technology, National University of Defense TechnologyInstitute for Quantum Science and Technology, College of Science, National University of Defense TechnologyCollege of Computer Science and Technology, National University of Defense TechnologyNational Innovation Institute of Defense Technology, Academy of Military Science PLACollege of Computer Science and Technology, National University of Defense TechnologyNational Innovation Institute of Defense Technology, Academy of Military Science PLACollege of Computer Science and Technology, National University of Defense TechnologyInstitute for Quantum Science and Technology, College of Science, National University of Defense TechnologyInstitute for Quantum Science and Technology, College of Science, National University of Defense TechnologyAbstract Multi-wavelength optical information processing systems are commonly utilized in optical neural networks and broadband signal processing. However, their effectiveness is often compromised by frequency-selective responses caused by fabrication, transmission, and environmental factors. To mitigate these issues, this study introduces a deep reinforcement learning calibration (DRC) method inspired by the deep deterministic policy gradient training strategy. This method continuously and autonomously learns from the system, effectively accumulating experiential knowledge for calibration strategies and demonstrating superior adaptability compared to traditional methods. In systems based on dispersion compensating fiber, micro-ring resonator array, and Mach-Zehnder interferometer array that use multi-wavelength optical carriers as the light source, the DRC method enables the completion of the corresponding signal processing functions within 21 iterations. This method provides efficient and accurate control, making it suitable for applications such as optical convolution computation acceleration, microwave photonic signal processing, and optical network routing.https://doi.org/10.1038/s41377-025-01846-6
spellingShingle Qiuquan Yan
Hao Ouyang
Zilong Tao
Meili Shen
Shiyin Du
Jun Zhang
Hengzhu Liu
Hao Hao
Tian Jiang
Multi-wavelength optical information processing with deep reinforcement learning
Light: Science & Applications
title Multi-wavelength optical information processing with deep reinforcement learning
title_full Multi-wavelength optical information processing with deep reinforcement learning
title_fullStr Multi-wavelength optical information processing with deep reinforcement learning
title_full_unstemmed Multi-wavelength optical information processing with deep reinforcement learning
title_short Multi-wavelength optical information processing with deep reinforcement learning
title_sort multi wavelength optical information processing with deep reinforcement learning
url https://doi.org/10.1038/s41377-025-01846-6
work_keys_str_mv AT qiuquanyan multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT haoouyang multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT zilongtao multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT meilishen multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT shiyindu multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT junzhang multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT hengzhuliu multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT haohao multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning
AT tianjiang multiwavelengthopticalinformationprocessingwithdeepreinforcementlearning