面向联邦原型学习的格陷门条件身份追溯机制

doi:10.19678/j.issn.1000-3428.0260364

摘要/Abstract

摘要： 近年来，为了克服传统联邦学习在面临设备系统异构和通信带宽受限时的瓶颈，联邦原型学习（FPL）通过聚合低维的特征原型而非庞大的高维模型参数，极大降低了通信开销。FPL的开放式分布式架构使其在实际部署中面临着安全威胁：缺乏严格准入机制的网络极易遭受外部攻击者的身份伪造；同时内部恶意节点可通过上传恶意构造的虚假原型实施投毒攻击，严重破坏全局特征空间。针对FPL在开放网络环境下易遭身份伪造，且现有统计学防御机制在非独立同分布场景下难以区分恶意扰动与良性异构特征导致诚实长尾节点被误剔除的问题，提出了一种面向FPL的格陷门可控身份追溯机制(FedLHT)。该机制将安全防线从概率性统计检测转化为确定性数学计算，构建了事前匿名、事中验证与事后追责的完整轻量级防御架构。在初始化阶段，基于格密码构造无证书签名体系，利用原像采样算法实现身份标识与格上短向量私钥的密码学绑定，实现了验证权与追溯权的物理解耦；在匿名准入阶段，通过注入格上误差生成隐匿真实身份的可追溯标签，并结合拒绝采样技术输出非交互式签名，在保障信息论匿名性的同时消除签名对私钥的统计依赖。在聚合验证阶段，将非线性群运算全面降级为格上矩阵-向量的线性乘加计算，实现了轻量级的批处理准入；在靶向追溯阶段，监管实体利用独占的格陷门主基及其短正交化性质求解有界距离解码问题，从异常误差项中靶向恢复恶意节点的真实身份，进而驱动全局模型精准剥离恶意历史贡献以实现自愈。该机制的安全性严格规约至平均情况下的带误差学习（LWE）问题与小整数解（SIS）问题，从根本上赋予了系统抵抗未来大规模量子计算机攻击的后量子安全属性。在模型自愈层面充分利用了特征原型向量的线性叠加同态特性。锁定恶意身份后服务器仅需在全局原型池中执行低开销的向量减法，即可在常数级时间内抹除恶意节点的历史贡献，实现了训练过程的无缝衔接。与隐私保护联邦学习投毒防御机制(ShieldFL)及基于信任引导的鲁棒联邦学习机制(FLTrust)等方案相比，全局模型准确率在CIFAR-10数据集上维持在83.65%以上；在数据异构程度系数为1.0的严苛Non-IID环境下，对诚实长尾数据的保留率高达98.75%，实现了零误判响应。此外，在40个并发客户端规模下，客户端局部耗时仅1.22毫秒，服务器端身份验证总开销低至6.42毫秒，有效克服了传统双线性配对群签名方案的计算瓶颈，适用于资源受限的移动物联网场景。

Abstract: Federated Prototype Learning (FPL) significantly reduces communication overhead but faces critical security risks in open networks, including identity forgery and poisoning attacks. Existing statistical defenses struggle to distinguish malicious perturbations from benign heterogeneous features under non-independent and identically distributed (Non-IID) settings, often wrongfully eliminating honest long-tail clients. To address these vulnerabilities, this paper proposes a lattice trapdoor-based controllable identity tracing scheme for FPL, termed Federated Lattice Hidden Trapdoor (FedLHT). Shifting from probabilistic statistical detection to deterministic mathematical computation, FedLHT establishes a lightweight framework for pre-training anonymity, real-time authentication, and post-attack accountability. It constructs a certificateless signature system via lattice cryptography, utilizing preimage sampling to bind client identities with private keys. Traceable tags and rejection sampling guarantee information-theoretic anonymity. During aggregation, nonlinear group operations are linearized over lattices for lightweight batch verification. For targeted tracing, regulators utilize the lattice trapdoor master basis to solve the bounded distance decoding problem, accurately recovering malicious identities from abnormal noise. FedLHT's security is rigorously reduced to average-case Learning With Errors (LWE) and Short Integer Solution (SIS) problems, ensuring post-quantum security. Furthermore, utilizing the linear additive homomorphism of prototypes, the server seamlessly erases malicious historical contributions via lightweight vector subtraction for rapid model self-recovery. Compared with state-of-the-art defenses, such as ShieldFL (Mitigating Poisoning Attacks in Privacy-Preserving Federated Learning) and FLTrust (Byzantine-robust Federated Learning via Trust Bootstrapping), the proposed mechanism maintains a global model accuracy above 84.65% on the CIFAR-10 dataset. Furthermore, in a severe Non-IID setting with a heterogeneity coefficient of 1.0, it achieves a 98.75% retention rate for honest long-tail data, realizing zero-misjudgment response. For 40 concurrent clients, local computing and server-side authentication overheads are merely 1.22 ms and 6.42 ms, effectively overcoming computational bottlenecks of traditional pairing-based schemes for resource-constrained Internet of Things (IoT) scenarios.

张晗瑞, 张淑芬, 刘新蕊, 姜志, 王炯炯. 面向联邦原型学习的格陷门条件身份追溯机制[J]. 计算机工程, doi: 10.19678/j.issn.1000-3428.0260364.

ZHANG Hanrui, ZHANG Shufen, LIU Xinrui, JIANG Zhi , WANG Jiongjiong. Lattice Trapdoor-based Controllable Identity Traceability Mechanism for Federated Prototype Learning[J]. Computer Engineering, doi: 10.19678/j.issn.1000-3428.0260364.

参考文献

[1] Mcmahan B, Moore E, Ramage D, et al. Communication-efficient learning of deep networks from decentralized data [C]//Proceedings of the 20th International Conference on Artificial Intelligence and Statistics. Fort Lauderdale: PMLR, 2017: 1273-1282.
[2] 王瑞锦,王金波,张凤荔,等.联邦原型学习的特征图中毒攻击和双重防御机制[J].软件学报, 2025, 36(03): 1355-1374. (Wang R J, Wang J B, Zhang F L, et al. Feature map poisoning attacks and dual defense mechanisms for federated prototype learning [J]. Journal of Software, 2025, 36(03): 1355-1374.)
[3] Tang L T, Chen Z N, Zhang L F, et al. Research progress of privacy issues in federated learning [J]. Journal of Software, 2023, 34(1): 197-229.
[4] 刘艺璇,陈红,刘宇涵,等.联邦学习中的隐私保护技术[J]. 软件学报, 2022, 33(3): 1057-1092. Liu y x, Chen h, Liu y h, et al. Privacy-preserving techniques in federated learning [J]. Journal of Software, 2022, 33(3): 1057-1092 (in Chinese).
[5] 欧璇璇,侯文涵,胡信豪,等.一种双向防御的联邦学习框架[J/OL].计算机工程,1-14[2026-03-20].https://doi.org/10.19678/j.issn.1000-3428.0253081. (Ou Xuanxuan, Hou Wenhan, Hu Xinhao, et al. A federated learning framework with bidirectional defense[J/OL].Computer Engineering,1-14[2026-03-20].https://doi.org/10.19678/j.issn.1000-3428.0253081.)
[6] Wang R J, Lai J S, Zhang Z Y, et al. Privacy-preserving federated learning for Internet of Medical Things under edge computing [J]. IEEE Journal of Biomedical and Health Informatics, 2023, 27(2): 854-865.
[7] Wei K, Li J, Ding M, et al. User-level privacy-preserving federated learning: analysis and performance optimization [J]. IEEE Transactions on Mobile Computing, 2022, 21(9): 3388-3401.
[8] 云健,张雪怡.面向区块链联邦学习的模型质量评估与验证机制优化[J/OL].计算机工程,1-17[2026-03-20].https://doi.org/10.19678/j.issn.1000-3428.2521002.(Yun Jian, Zhang Xueyi. Optimization of model quality evaluation and verification mechanism for blockchain-based federated learning[J/OL]. Computer Engineering,1-17[2026-03-20].https://doi.org/10.19678/j.issn.1000-3428.2521002.)
[9] Tan Y, Long G D, Liu L, et al. FedProto: federated prototype learning across heterogeneous clients [C]// Proceedings of the 36th AAAI Conference on Artificial Intelligence. New York: AAAI Press, 2022: 8432-8440.
[10] 陈庆礼,郭渊博,方晨.面向数据异构的聚类联邦学习算法[J].计算机应用, 2025, 45(4): 1086-1094. (Chen Qing Li, Guo Yuan Bo, Fang Chen. Clustered federated learning algorithm for data heterogeneity [J]. Journal of Computer Applications, 2025, 45(4): 1086-1094.)
[11] SHEIKHI S, KOSTAKOS P, LOVEN L. Hybrid reputation aggregation: a robust defense mechanism for adversarial federated learning in 5G and edge network environments [J/OL]. arXiv preprint arXiv:2509.18044, 2025.
[12] Yin D, Chen Y, Ramchandran K, et al. Byzantine-robust distributed learning: Towards optimal statistical rates [C]// Proceedings of the 35th International Conference on Machine Learning. PMLR, 2018: 5650-5659.
[13] Xie C, Koyejo S, Gupta I. Zeno: Distributed stochastic gradient descent with suspicion-based fault-tolerance [C]// Proceedings of the 36th International Conference on Machine Learning. PMLR, 2019: 6893-6901.
[14] Cao X, Jia J, Zhang Z, et al. FLTrust: Byzantine-robust federated learning via trust bootstrapping [C]// Proceedings of the 28th Annual Network and Distributed System Security Symposium (NDSS). 2021.
[15] Kang J, Xiong Z, Niyato D, et al. Incentive mechanism for reliable federated learning: A joint optimization approach to combining reputation and contract theory [J]. IEEE Internet of Things Journal, 2019, 6(5): 3232-3246.
[16] Zhang J, Zhang H, Wang G, et al. Defending against poisoning attacks in federated prototype learning on non-IID data [C]// Proceedings of the Wireless Artificial Intelligent Computing Systems and Applications (WASA). 2024: 123-135.
[17] Cui J, Zhang X, Zhong H, et al. Efficient privacy-preserving authentication with traceability for VANETs [J]. IEEE Transactions on Vehicular Technology, 2023, 72(5): 6543-6555.
[18] Yang Y, Wang Z, Liu G. A lightweight privacy-preserving authentication scheme based on ring signature for mobile edge computing [J]. IEEE Systems Journal, 2023, 17(1): 1234-1245.
[19] Xu G, Li H, Liu S, et al. VerifyNet: Secure and verifiable federated learning [J]. IEEE Transactions on Information Forensics and Security, 2020, 15: 911-926.
[20] Micciancio D, Regev O. Worst-case to average-case reductions based on Gaussian measures [J]. SIAM Journal on Computing, 2004, 37(1): 267-302.
[21] Camenisch J, Stadler M. Efficient group signature schemes for large groups [C]// Proceedings of the 17th Annual International Cryptology Conference (CRYPTO). Springer, 1997: 410-424.
[22] Gentry C, Peikert C, Vaikuntanathan V. Trapdoors for hard lattices and new cryptographic constructions [C]// Proceedings of the 40th Annual ACM Symposium on Theory of Computing (STOC). ACM, 2008: 197-206.
[23] Fiat A, Shamir A. How to prove yourself: Practical solutions to identification and signature problems [C]// Conference on the Theory and Application of Cryptographic Techniques (CRYPTO). Springer, 1986: 186-194.
[24] Pointcheval D, Stern J. Security arguments for digital signatures and blind signatures [J]. Journal of Cryptology, 2000, 13(3): 361-396.
[25] Lyubashevsky V. Lattice signatures without trapdoors [C]// Annual International Conference on the Theory and Applications of Cryptographic Techniques (EUROCRYPT). Springer, 2012: 738-755.
[26] Krizhevsky A, Hinton G. Learning multiple layers of features from tiny images [R]. Technical report, University of Toronto, 2009.
[27] Ma Z, Ma J, Miao Y, et al. ShieldFL: Mitigating poisoning attacks in privacy-preserving federated learning [J]. IEEE Transactions on Information Forensics and Security, 2022, 17: 1639-1654.
[28] Zhang l, Li j, Yang y. Message linkable group signature with information binding and efficient revocation for privacy-preserving announcement in VANETs [J]. IEEE Transactions on Dependable and Secure Computing, 2024, 21(6): 5667-5680.
[29] Pu l, Gu j, Lin c, et al. FedLG: lightweight generic certificateless authentication for trustworthy federated learning in VANETs [J]. IEEE Transactions on Information Forensics and Security, 2025.
[30] Zhou m, Lin c, Xu s, et al. Sphinx: certificateless conditional privacy-preserving authentication with secure transmission for VANETs [J]. IEEE Transactions on Dependable and Secure Computing, 2025, 22(6): 7189-7202.
[31] Briggs C, Fan Z, Andras P. Federated learning with hierarchical clustering of local updates [C]// Proceedings of the 2020 International Joint Conference on Neural Networks (IJCNN). IEEE, 2020: 1-8.
[32] Hsu T M H, Qi H, Brown M. Measuring the effects of non-identical data distribution for federated visual classification [J/OL]. arXiv preprint arXiv:1909.06335, 2019.

选择文件类型/文献管理软件名称

选择包含的内容