1.国网湖北直流公司,湖北 宜昌 443005
2.华中科技大学 光学与电子信息学院,武汉 430074
粟梦涵(1995-),女,湖北建始人。工程师,硕士,主要研究方向为光通信与电力通信。
程孟凡,教授。E-mail:chengmf@hust.edu.cn
收稿:2025-10-19,
修回:2025-11-11,
纸质出版:2026-10-10
移动端阅览
粟梦涵,唐成喆,段施奇,等. 无源光网络物理层硬件身份识别方法与验证[J]. 光通信研究,2026(5): 86-90.
Su M H, Tang C Z, Duan S Q, et al. Physical-Layer Hardware Identity Recognition and Validation in Passive Optical Networks[J]. Study on Optical Communications, 2026(5): 86-90.
粟梦涵,唐成喆,段施奇,等. 无源光网络物理层硬件身份识别方法与验证[J]. 光通信研究,2026(5): 86-90. DOI: 10.13756/j.gtxyj.2026.250331.
Su M H, Tang C Z, Duan S Q, et al. Physical-Layer Hardware Identity Recognition and Validation in Passive Optical Networks[J]. Study on Optical Communications, 2026(5): 86-90. DOI: 10.13756/j.gtxyj.2026.250331.
目的
2
为实现无源光网络(PON)多用户场景下合法光网络单元(ONU)的高效身份识别,文章提出一种基于设备物理指纹的认证方案。
方法
2
利用信号中蕴含的硬件制造差异,设计了一种近似眼图的特征提取方法,并结合卷积神经网络(CNN)进行识别。同时采用小波变换特征提取方法作为对比。实验中,基于现场可编程门阵列(FPGA)板卡与光收发模块,分别搭建了8个ONU的少用户PON网络和32个ONU的多用户PON网络,通过高速示波器采集信号数据。两种特征提取方法均采用相同结构的CNN进行训练与识别。
结果
2
在少用户场景下,两种方法的平均识别准确率均接近100%,可实现对合法用户的高精度身份认证;在多用户场景下,小波变换方法的平均识别准确率显著下降至58%,而近似眼图方法仍保持98.45%的高准确率,且在计算复杂度上优于小波变换方法。
结论
2
基于设备物理指纹的认证方案能够有效识别PON网络中的合法ONU,其中近似眼图特征提取方法在多用户场景下表现出更高的识别精度与计算效率,适用于实际网络环境的高效身份认证。
Objective
2
To achieve efficient identification of legitimate Optical Network Units (ONU) in multi-user scenarios of Passive Optical Networks (PON)
this paper proposes an authentication scheme based on device physical fingerprints.
Methods
2
Exploiting hardware manufacturing differences hidden in the signals
a feature extraction method based on approximate eye diagrams was designed and combined with a Convolutional Neural Network (CNN) for recognition. For comparison
a wavelet transform-based feature extraction method was also employed. In the experiments
an 8-ONU small-scale PON network and a 32-ONU large-scale PON network were constructed using a Field-Programmable Gate Array (FPGA) board and optical transceiver modules. Signal data were captured with a high-speed oscilloscope
and both feature extraction methods were trained and tested using CNNs with identical structures.
Results
2
In the small-scale scenario
both methods achieved an average recognition accuracy close to 100%
enabling high-precision authentication of legitimate users. In the large-scale scenario
the wavelet transform method showed a significant drop in accuracy to 58%
whereas the approximate eye diagram method maintained a high average accuracy of 98.45% with lower computational complexity compared to the wavelet transform method.
Conclusion
2
The proposed authentication scheme based on device physical fingerprints can effectively identify legitimate ONUs in PON networks. The approximate eye diagram-based feature extraction method exhibits superior recognition accuracy and computational efficiency in multi-user scenarios
making it suitable for efficient authentication in practical network environments.
李婕 , 罗鸣 , 李响 . 高速无源光网络技术路线及系统研究 [J ] . 光通信研究 , 2025 ( 5 ): 240099 .
Li J , Luo M , Li X . Technology Route and System Research of High-speed PON [J ] . Study on Optical Communications , 2025 ( 5 ): 240099 .
张佳玮 , 苏琛 , 纪越峰 . 工业PON中确定性网络传输技术研究 [J ] . 光通信研究 , 2024 ( 1 ): 230168 .
Zhang J W , Su C , Ji Y F . Research on Deterministic Network Transmission Technology in Industrial PON [J ] . Study on Optical Communications , 2024 ( 1 ): 230168 .
殷明珠 , 王玮 , 潘竞顺 , 等 . 下一代相干PON关键技术研究进展与展望 [J ] . 光通信研究 , 2024 ( 1 ): 230162 .
Yin M Z , Wang W , Pan J S , et al . Research Progress and Prospect of Key Techniques for Next-Generation Coherent PON [J ] . Study on Optical Communications , 2024 ( 1 ): 230162 .
Hall J , Barbeau M , Kranakis E . Detection of Transient in Radio Frequency Fingerprinting Using Signal Phase [J ] . Wireless Optical Communications , 2003 , 9 : 13 .
Xing Y , Hu A , Zhang J , et al . Design of a Channel Robust Radio Frequency Fingerprint Identification Scheme [J ] . IEEE Internet of Things Journal , 2023 , 10 ( 8 ): 6946 - 6959 .
Xie R , Xu W , Chen Y , et al . A Generalizable Model-and-Data Driven Approach for Open-Set RFF Authentication [J ] . IEEE Transactions on Information Forensics and Security , 2021 , 16 : 4435 - 4450 .
Chen J , Wong W K , Hamdaoui B . Unsupervised Contrastive Learning for Robust RF Device Fingerprinting under Time-Domain Shift [C ] // ICC 2024 IEEE International Conference on Communications . Denver, USA : IEEE , 2024 : 10622173 .
Li S , Cheng M , Chen Y , et al . Enhancing the Physical Layer Security of OFDM-PONs with Hardware Fingerprint Authentication: a Machine Learning Approach [J ] . Journal of Lightwave Technology , 2020 , 38 ( 12 ): 3238 - 3245 .
Fan C , Gong H , Cheng M , et al . Identify the Device Fingerprint of OFDM-PONs with a Noise-Model-Assisted CNN for Enhancing Security [J ] . IEEE Photonics Journal , 2021 , 13 ( 4 ): 8600104 .
Gong H , Cheng M , Shao W , et al . Machine Learning Assisted Hardware Fingerprint Identification for TDM-PON from Eye-Diagram [C ] // Asia Communications and Photonics Conference 2021 . Shanghai, China : Optica Publishing Group , 2021 : T2B.3 .
Huang D , Deng L , Yang Q , et al . Hardware Fingerprint Authentication based on Siamese Neural Networks in PON [J ] . IEEE Photonics Technology Letters , 2024 , 36 ( 7 ): 508 - 511 .
Kang L , Zhang L , Huang X , et al . Hardware Fingerprint Authentication in Optical Networks Assisted by Anomaly Detection [J ] . IEEE Photonics Technology Letters , 2022 , 34 ( 19 ): 1030 - 1033 .
Li Y , Hua N , Li J , et al . Optical Spectrum Feature Analysis and Recognition for Optical Network Security with Machine Learning [J ] . Optics Express , 2019 , 27 ( 17 ): 24808 .
Qiu Y , Peng X , Huang X , et al . Variational Autoencoder-Assisted Unsupervised Hardware Fingerprint Authentication in a Fiber Network [J ] . Optics Letters , 2024 , 49 ( 8 ): 2029 - 2032 .
0
浏览量
0
下载量
0
CSCD
0
CNKI被引量
关联资源
相关文章
相关作者
相关机构
鄂公网安备 42011202002092号