面向协作频谱感知的个性化差分隐私联邦学习方法

唐湘云; 康嘉文; 韩旭; 张焘; 刘寅秋; 孙庚; 焦雨涛

doi:10.11959/j.issn.1000-436x.2026067

Chinese

您当前的位置：

首页 >

文章列表页 >

面向协作频谱感知的个性化差分隐私联邦学习方法

学术论文 | 更新时间：2026-05-07

- 面向协作频谱感知的个性化差分隐私联邦学习方法
- Personalized differential privacy federated learning method for collaborative spectrum sensing
- 通信学报 2026年47卷第4期页码：97-112
- 作者机构：
  
  1.中央民族大学信息工程学院，北京 100081
  2.广东工业大学自动化学院，广东广州 510006
  3.北京交通大学网络空间安全学院，北京 100091
  4.新加坡南洋理工大学，新加坡 639798
  5.吉林大学计算机科学与技术学院，吉林长春 130012
  6.中国人民解放军陆军工程大学通信工程学院，江苏南京 210007
- 作者简介：
  
  [ "唐湘云（1994- ），女，湖南永州人，博士，中央民族大学副教授，主要研究方向为人工智能安全、数据安全、隐私保护等。" ]
  [ "康嘉文（1992- ），男，广东茂名人，博士，广东工业大学教授，主要研究方向为人工智能、信息安全等。" ]
  [ "韩旭（2002- ），女，广西河池人，中央民族大学硕士生，主要研究方向为联邦学习、隐私保护等。" ]
  [ "张焘（1994- ），男，安徽马鞍山人，博士，北京交通大学副教授，主要研究方向为区块链、联邦学习等。" ]
  [ "刘寅秋（1998- ），男，江苏徐州人，博士，新加坡南洋理工大学副研究员，主要研究方向为区块链安全、边缘智能等。" ]
  [ "孙庚（1990- ），男，吉林长春人，博士，吉林大学教授，主要研究方向为低空无线网络、移动边缘计算等。" ]
  [ "焦雨涛（1992- ），男，江苏南京人，博士，中国人民解放军陆军工程大学副教授，主要研究方向为智能短波通信、电磁频谱感知、无线联邦学习等。" ]
- 基金信息：
  
  国家自然科学基金资助项目(62572132;62572502;62571548);国家密码基金资助项目(2025NCSF02030);国家自然科学基金青年科学基金资助项目（C类）(62302539;62402029);北京市自然科学基金丰台联合基金资助项目(L251041)
- DOI：10.11959/j.issn.1000-436x.2026067
  中图分类号： TN92
- 收稿：2025-12-01，
  
  修回：2026-03-04，
  
  录用：2026-03-11，
  
  纸质出版：2026-04-20
- 稿件说明：
移动端阅览
唐湘云,康嘉文,韩旭等.面向协作频谱感知的个性化差分隐私联邦学习方法[J].通信学报,2026,47(04):97-112.

Tang Xiangyun,Kang Jiawen,Han Xu,et al.Personalized differential privacy federated learning method for collaborative spectrum sensing[J].Journal on Communications,2026,47(04):97-112.
唐湘云,康嘉文,韩旭等.面向协作频谱感知的个性化差分隐私联邦学习方法[J].通信学报,2026,47(04):97-112. DOI： 10.11959/j.issn.1000-436x.2026067.

Tang Xiangyun,Kang Jiawen,Han Xu,et al.Personalized differential privacy federated learning method for collaborative spectrum sensing[J].Journal on Communications,2026,47(04):97-112. DOI： 10.11959/j.issn.1000-436x.2026067.

摘要

针对协作频谱感知中数据非独立同分布（Non-IID）特性导致的模型性能下降问题，提出了一种融合个性化差分隐私与重平衡分簇策略的联邦学习方案（RebalFL）。该方案首先引入个性化差分隐私机制，允许数据设置差异化隐私预算，在保障隐私的同时减少噪声注入；其次，设计重平衡分簇策略，构建数据分布均衡的客户端簇，缓解模型漂移问题。实验结果表明，RebalFL在Non-IID场景下显著优于现有差分隐私方法，能有效提升频谱感知模型在隐私保护下的分类精度与鲁棒性。

Abstract

To address the degradation of model performance caused by data non-independent and identically distributed (Non-IID) characteristics in collaborative spectrum sensing

a federated learning scheme RebalFL was proposed

which integrated personalized differential privacy with a rebalancing clustering strategy. First

a personalized differential privacy mechanism that allowed heterogeneous privacy budgets for different data sources was introduced

thereby reducing noise injection while preserving privacy. Then

a rebalancing clustering strategy was designed to form client clusters with more balanced data distributions and mitigate model drift. Experimental results show that RebalFL outperforms existing differential privacy methods in Non-IID scenarios

substantially improving the classification accuracy and robustness of spectrum sensing models under privacy protection.

关键词

Keywords

references

王海涛 , 茆习文 , 张晨 , 等 . 基于天基干扰绘图的星地一体化系统频谱共享研究 [J ] . 通信学报 , 2024 , 45 ( 3 ): 155 - 165 .

Wang H T , Mao X W , Zhang C , et al . Study of spectrum sharing in integrated satellite-terrestrial system based on space-based interference cartography [J ] . Journal on Communications , 2024 , 45 ( 3 ): 155 - 165 .

Zhang W S , Wang Y , Chen X , et al . Collaborative learning-based spectrum sensing under partial observations [J ] . IEEE Transactions on Cognitive Communications and Networking , 2024 , 10 ( 5 ): 1843 - 1855 .

Janu D , Singh K , Kumar S . Machine learning for cooperative spectrum sensing and sharing: a survey [J ] . Transactions on Emerging Telecommunications Technologies , 2022 , 33 : e4352 .

Hassan M U , Rehmani M H , Rehan M , et al . Differential privacy in cognitive radio networks: a comprehensive survey [J ] . Cognitive Computation , 2022 , 14 ( 2 ): 475 - 510 .

Bai L , Hu H B , Ye Q Q , et al . Membership inference attacks and defenses in federated learning: a survey [J ] . ACM Computing Surveys , 2025 , 57 ( 4 ): 1 - 35 .

Wasilewska M , Bogucka H , Poor H V . Secure federated learning for cognitive radio sensing [J ] . IEEE Communications Magazine , 2023 , 61 ( 3 ): 68 - 73 .

Yang H M , Ge M Y , Xue D Y , et al . Gradient leakage attacks in federated learning: research frontiers, taxonomy, and future directions [J ] . IEEE Network , 2024 , 38 ( 2 ): 247 - 254 .

Ouadrhiri A E , Abdelhadi A . Differential privacy for deep and federated learning: a survey [J ] . IEEE Access , 2022 , 10 : 22359 - 22380 .

Wu X , Zhang Y T , Shi M Y , et al . An adaptive federated learning scheme with differential privacy preserving [J ] . Future Generation Computer Systems , 2022 , 127 : 362 - 372 .

Bao T , Xu L , Zhu L H , et al . Successive point-of-interest recommendation with personalized local differential privacy [J ] . IEEE Transactions on Vehicular Technology , 2021 , 70 ( 10 ): 10477 - 10488 .

Costa L D S , Guimarães D A , Uchôa-Filho B F . On the signal-to-noise ratio wall of energy detection in spectrum sensing [J ] . IEEE Access , 2022 , 10 : 16499 - 16511 .

Arjoune Y , Kaabouch N . A comprehensive survey on spectrum sensing in cognitive radio networks: recent advances, new challenges, and future research directions [J ] . Sensors , 2019 , 19 ( 1 ): 126 .

Song Z H , Gao Y , Tafazolli R . A survey on spectrum sensing and learning technologies for 6G [J ] . IEICE Transactions on Communications , 2021 , 104 ( 10 ): 1207 - 1216 .

Charan C , Pandey R . Eigenvalue based double threshold spectrum sensing under noise uncertainty for cognitive radio [J ] . Optik , 2016 , 127 ( 15 ): 5968 - 5975 .

Nasser A , Hassan H A H , Chaaya J A , et al . Spectrum sensing for cognitive radio: recent advances and future challenge [J ] . Sensors , 2021 , 21 ( 7 ): 2408 .

Saber M , Rharras A E , Saadane R , et al . An optimized spectrum sensing implementation based on SVM, KNN and TREE algorithms [C ] // Proceedings of the 2019 15th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS) . Piscataway : IEEE Press , 2019 : 383 - 389 .

Liu C , Wang J , Liu X M , et al . Deep CM-CNN for spectrum sensing in cognitive radio [J ] . IEEE Journal on Selected Areas in Communications , 2019 , 37 ( 10 ): 2306 - 2321 .

Solanki S , Dehalwar V , Choudhary J , et al . Spectrum sensing in cognitive radio using CNN-RNN and transfer learning [J ] . IEEE Access , 2022 , 10 : 113482 - 113492 .

Li H N , Hu T H , Chen J X , et al . Privacy preserving algorithm for spectrum sensing in cognitive vehicle networks [J ] . Chinese Journal of Electronics , 2024 , 33 ( 1 ): 30 - 42 .

Kairouz P , McMahan H B . Advances and open problems in federated learning [J ] . Foundations and Trends in Machine Learning , 2021 , 14 ( 1/2 ): 1 - 210 .

Chen Z B , Xu Y Q , Wang H B , et al . Federated learning-based cooperative spectrum sensing in cognitive radio [J ] . IEEE Communications Letters , 2022 , 26 ( 2 ): 330 - 334 .

Song Y F , Chang H H , Liu L J . Federated dynamic spectrum access through multi-agent deep reinforcement learning [C ] // Proceedings of the GLOBECOM 2022-2022 IEEE Global Communications Conference . Piscataway : IEEE Press , 2022 : 3466 - 3471 .

Shi Y , Sagduyu Y E , Erpek T . Federated learning for distributed spectrum sensing in NextG communication networks [PP ] . V1 . ( 2022-04-06 ) [ 2025-12-01 ] . arXiv: arXiv. 2204.03027.

Li T , Sahu A K , Zaheer M , et al . Federated optimization in heterogeneous networks [J ] . Proceedings of Machine Learning and Systems , 2020 , 2 : 429 - 450 .

Wang J Y , Liu Q H , Liang H , et al . Tackling the objective inconsistency problem in heterogeneous federated optimization [J ] . Advances in Neural Information Processing Systems , 2020 , 33 : 7611 - 7623 .

李梦倩 , 田有亮 , 张军鹏 , 等 . 基于零集中差分隐私的联邦学习激励方案 [J ] . 通信学报 , 2025 , 46 ( 1 ): 79 - 92 .

Li M Q , Tian Y L , Zhang J P , et al . Incentive scheme for federated learning based on zero-concentrated differential privacy [J ] . Journal on Communications , 2025 , 46 ( 1 ): 79 - 92 .

Fu J , Hong Y , Ling X P , et al . Differentially private federated learning: a systematic review [PP ] . V4 . ( 2025-10-02 ) [ 2025-12-01 ] . arXiv: arXiv. 2405.08299.

Fu J , Ye Q Q , Hu H B , et al . DPSUR: accelerating differentially private stochastic gradient descent using selective update and release [PP ] . V2 . ( 2023-11-29 ) [ 2025-12-01 ] . arXiv: arXiv. 2311.14056.

Liu J X , Lou J , Xiong L , et al . Projected federated averaging with heterogeneous differential privacy [J ] . Proceedings of the VLDB Endowment , 2021 , 15 ( 4 ): 828 - 840 .

Liu J X , Lou J , Xiong L , et al . Cross-silo federated learning with record-level personalized differential privacy [C ] // Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security . New York : ACM Press , 2024 : 303 - 317 .

Jorgensen Z , Yu T , Cormode G . Conservative or liberal? Personalized differential privacy [C ] // Proceedings of the 2015 IEEE 31st International Conference on Data Engineering . Piscataway : IEEE Press , 2015 : 1023 - 1034 .

Boenisch F , Dziedzic A , Mühl C , et al . Have it your way: individualized privacy assignment for DP-SGD [C ] // Proceedings of the Advances in Neural Information Processing Systems 36 . Massachusetts : MIT Press , 2023 : 19073 - 19103 .

Aygül M A , Çırpan H A , Arslan H . Machine learning-based spectrum occupancy prediction: a comprehensive survey [J ] . Frontiers in Communications and Networks , 2025 , 6 : 1482698 .

Zhao Y , Chen J J . A survey on differential privacy for unstructured data content [J ] . ACM Computing Surveys , 2022 , 54 ( 10 s): 1 - 28 .

Niu B , Chen Y H , Wang B Y , et al . AdaPDP: adaptive personalized differential privacy [C ] // Proceedings of the IEEE INFOCOM 2021-IEEE Conference on Computer Communications . Piscataway : IEEE Press , 2021 : 1 - 10 .

Liu Z X , Li T , Smith V , et al . Enhancing the privacy of federated learning with sketching [PP ] . V1 . ( 2019-11-05 ) [ 2025-12-01 ] . arXiv: arXiv. 1911 . 01812 .

Hayes J , Melis L , Danezis G , et al . LOGAN: membership inference attacks against generative models [J ] . Proceedings on Privacy Enhancing Technologies , 2019 , 2019( 1 ): 133 - 152 .

Wang Z B , Song M K , Zhang Z F , et al . Beyond inferring class representatives: user-level privacy leakage from federated learning [C ] // Proceedings of the IEEE Conference on Computer Communications . Piscataway : IEEE Press , 2019 : 2512 - 2520 .

Mironov I , Talwar K , Zhang L . Rényi differential privacy of the sampled Gaussian mechanism [PP ] . V1 . ( 2019-08-28 ) [ 2025-12-01 ] . arXiv: arXiv. 1908 . 10530 .

Mironov I . Rényi differential privacy [C ] // Proceedings of the 2017 IEEE 30th Computer Security Foundations Symposium (CSF) . Piscataway : IEEE Press , 2017 : 263 - 275 .

Arjoune Y , Kaabouch N . On spectrum sensing, a machine learning method for cognitive radio systems [C ] // Proceedings of the 2019 IEEE International Conference on Electro Information Technology (EIT) . Piscataway : IEEE Press , 2019 : 333 - 338 .

McMahan B , Moore E , Ramage D , et al . Communication-efficient learning of deep networks from decentralized data [C ] // Artificial Intelligence and Statistics . New York : PMLR , 2017 : 1273 - 1282 .

Geyer R C , Klein T , Nabi M . Differentially private federated learning: a client level perspective [PP ] . V2 . ( 2018-03-01 ) [ 2025-12-01 ] . arXiv: arXiv. 1712 . 07557 .

浏览量

116

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于多方计算的安全拜占庭弹性联邦学习

面向Non-IID数据的拜占庭鲁棒联邦学习

面向隐私保护的非聚合式数据共享综述

ISAC网络中基于特征信息的联邦学习动态压缩方法