基于对偶学习的高效可见光通信非线性均衡器

宋锐; 邓春丽; 李志良; 刘宏立

doi:null

您当前的位置：

首页 >

文章列表页 >

基于对偶学习的高效可见光通信非线性均衡器

光信息技术 | 更新时间：2026-01-22

- 基于对偶学习的高效可见光通信非线性均衡器
- Efficient nonllinear equalizer for visible light communication based on dual learning
- 光学技术 2026年52卷第1期页码：68-74
- 作者机构：
  
  1.张家界航空工业职业技术学院航空电气工程系，湖南张家界 427000
  2.湖南大学电气与信息工程学院，湖南长沙 410082
- 作者简介：
  
  宋锐（1977— ），男，副教授，从事半导体照明和可见光通信研究。zjjsongrui@163.com
- 基金信息：
  
  湖南省自然科学基金资助项目(2022JJ60088);湖南省教育厅科学研究资助项目(24C1329)
- DOI：
  中图分类号： TN929
- 收稿：2025-06-17，
  
  修回：2025-09-15，
  
  纸质出版：2026-01-20
- 稿件说明：
移动端阅览
宋锐,邓春丽,李志良等.基于对偶学习的高效可见光通信非线性均衡器[J].光学技术,2026,52(01):68-74.

SONG Rui,DENG Chunli,LI Zhiliang,et al.Efficient nonllinear equalizer for visible light communication based on dual learning[J].OPTICAL TECHNIQUE,2026,52(01):68-74.
宋锐,邓春丽,李志良等.基于对偶学习的高效可见光通信非线性均衡器[J].光学技术,2026,52(01):68-74. DOI：

SONG Rui,DENG Chunli,LI Zhiliang,et al.Efficient nonllinear equalizer for visible light communication based on dual learning[J].OPTICAL TECHNIQUE,2026,52(01):68-74. DOI：

摘要

针对室内可见光通信（VLC）中由LED动态非线性和信道衰落所引起的信号失真问题，提出一种基于对偶学习的非线性均衡方法。通过搭建PAM4调制VLC仿真系统，结合LED非线性模型与NLOS信道以模拟失真特性，设计了基于对偶神经网络结构的均衡器，利用正向均衡与反向映射网络实现双向约束，并借助对偶一致性损失提升在有限训练样本下的均衡性能。实验结果表明，方法显著降低了系统误码率，在相同训练符号数量下，误码率较传统均衡方法最多可降低2个数量级；与传统神经网络均衡器相比，达到FEC阈值所需训练序列长度最大可减少62.5%，能够在显著提升VLC系统传输可靠性的同时，大幅降低均衡器训练开销。

Abstract

To address signal distortion caused by LED dynamic nonlinearity and channel fading in indoor visible light communication （VLC） systems

a nonlinear equalization method based on dual learning is proposed. A PAM4-modulated VLC simulation system is constructed

incorporating an LED nonlinearity model and a non-line-of-sight （NLOS） channel to emulate distortion characteristics. A dual neural network structure is designed

which employs a forward equalization network and a backward mapping network to achieve bidirectional constraints

and utilizes a dual consistency loss to enhance equalization performance under limited training samples. Experimental results show that the proposed method significantly improves the bit error rate （BER） performance of the system. Under the same number of training symbols

the BER can be reduced by up to two orders of magnitude compared to conventional equalization methods. Compared to traditional neural network equalizers

the required training sequence length to achieve the FEC threshold is reduced by up to 62.5%

effectively enhancing the transmission reliability of the VLC system while substantially reducing the training overhead.

关键词

Keywords

references

马瑞 . 基于自注意力神经网络的室内可见光定位方法 [J ] . 光学技术 , 2025 , 51 （ 1 ）: 108 — 115 .

Ma Rui . Indoor visible light positioning method based on self-attention neural network [J ] . Optical Technique , 2025 , 51 （ 1 ）: 108 — 115 .

刘杰徽 , 何英 , 胡锋 , 等 . 基于PIN的水下可见光双向语音通信试验研究 [J ] . 光通信研究 , 2022 ,（ 3 ）: 65 — 71 .

Liu Jiehui , He Ying , Hu Feng , et al . Experimental research on underwater visible light two-way voice Communication based on PIN [J ] . Study on Optical Communications , 2022 ,（ 3 ）: 65 — 71 .

贾科军 , 雷蕾 , 蔺莹 , 等 . 基于记忆多项式的Volterra混合均衡器设计与研究 [J ] . 光通信研究 , 2024 （ 3 ）: 40 — 48 .

Jia Kejun , Lei Lei , Lan Ying , et al . Design and study of memory polynomial-based Volterra hybrid equalizer [J ] . Study on Optical Communications , 2024 （ 3 ）: 40 — 48 .

Deng X , Mardanikorani S , Wu Y , et al . Mitigating LED nonlinearity to enhance visible light communications [J ] . IEEE Transactions on Communications , 2018 , 66 （ 11 ）: 5593 — 5607 .

Lu H , Jin J , Wang J . Alleviation of LED nonlinearity impact in visible light communication using companding and predistortion [J ] . IET Communications , 2019 , 13 （ 7 ）: 818 — 821 .

Turan B , Narmanlioglu O , Koc O N , et al . Measurement based non-line-of-sight vehicular visible light communication channel characterization [J ] . IEEE Transactions on Vehicular Technology , 2022 , 71 （ 9 ）: 10110 — 10114 .

Lin B , Yang J , Nie K , et al . NLOS visible light positioning and communication based on LoRa modulation [J ] . Optics Express , 2024 , 32 （ 14 ）: 24128 — 24143 .

Wang L , Wang X , Kang J , et al . A 75-Mb/s RGB PAM-4 visible light communication transceiver system with pre-and post-equalization [J ] . Journal of Lightwave Technology , 2021 , 39 （ 5 ）: 1381 — 1390 .

Zhang G , Hong X , Fei C , et al . Sparsity-aware nonlinear equalization with greedy algorithms for LED-based visible light communication systems [J ] . Journal of Lightwave Technology , 2019 , 37 （ 20 ）: 5273 — 5281 .

成院波 , 查海朋 , 保贤 , 等 . LED非线性幅度分解多项式预失真 [J ] . 光通信研究 , 2024 ,（ 4 ）: 108 — 114 .

Cheng Yuanbo , Zha Haipeng , Bao Xian , et al . Amplitude factorization polynomial pre-distortion for nonlinear LED [J ] . Study on Optical Communications , 2024 ,（ 4 ）: 108 — 114 .

Miao P , Chen G , Cumanan K , et al . Deep hybrid neural network-based channel equalization in visible light communication [J ] . IEEE Communications Letters , 2022 , 26 （ 7 ）: 1593 — 1597 .

Zhou Y , Wei Y , Hu F , et al . Comparison of nonlinear equalizers for high-speed visible light communication utilizing silicon substrate phosphorescent white LED [J ] . Optics Express , 2020 , 28 （ 2 ）: 2302 — 2316 .

贺锋涛 , 王乐莹 , 王晓波 , 等 . 基于改进的AdaBoost无线光通信信号检测算法 [J ] . 激光技术 , 2023 , 47 （ 5 ）: 659 — 665 .

He Fengtao , Wang Leying , Wang Xiaobo , et al . Signal detection algorithm of wireless optical communication based on the improved AdaBoost [J ] . Laser Technology , 2023 , 47 （ 5 ）: 659 — 665 .

Zou P , Zhao Y , Hu F , et al . Enhanced performance of MIMO multi-branch hybrid neural network in single receiver MIMO visible light communication system [J ] . Optics Express , 2020 , 28 （ 19 ）: 28017 — 28032 .

Haigh P A , Ghassemlooy Z , Rajbhandari S , et al . Visible light communications: 170 Mb/s using an artificial neural network equalizer in a low bandwidth white light configuration [J ] . Journal of Lightwave Technology , 2014 , 32 （ 9 ）: 1807 — 1813 .

Chen H , Liu X , Fan Z , et al . Dual-Learning based neural networks for short-reach optical communications [C ] // 2021 Conference on Lasers and Electro-Optics（CLEO） . IEEE , 2021 : 1 — 2 .

Ulkar M G , Baykas T , Pusane A E . VLCnet: Deep learning based end-to-end visible light communication system [J ] . Journal of Lightwave Technology , 2020 , 38 （ 21 ）: 5937 — 5948 .

Stepniak G , Siuzdak J . Influence of lighting LED design parameters on the dynamic nonlinear response [J ] . Journal of Lightwave Technology , 2021 , 40 （ 4 ）: 954 — 960 .

Lin B , Lai Q , Ghassemlooy Z , et al . A machine learning based signal demodulator in NOMA-VLC [J ] . Journal of Lightwave Technology , 2021 , 39 （ 10 ）: 3081 — 3087 .

Lu X , Lu C , Yu W , et al . Memory-controlled deep LSTM neural network post-equalizer used in high-speed PAM VLC system [J ] . Optics Express , 2019 , 27 （ 5 ）: 7822 — 7833 .

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

暂无数据