面向医疗数据的隐私保护联邦学习架构

doi:10.19665/j.issn1001-2400.20230202

摘要/Abstract

摘要：

联邦学习作为一种新兴的神经网络训练模型,因其可以在保护用户数据隐私的前提下进行模型训练而受到广泛关注。然而,由于攻击者可以从共享梯度中跟踪和提取参与者的隐私,因此联邦学习仍然面临各种安全和隐私威胁。针对医疗数据在联邦学习过程中面临的隐私泄露问题,基于Paillier同态加密技术提出一种保护隐私的医疗数据联邦学习架构。首先,采用Paillier加密技术对客户端的共享训练模型进行加密,确保训练模型的安全性和隐私性,同时设计了零知识证明身份认证模块确保参与训练成员身份的可信性;其次,在服务器端通过构造消息确认机制将掉线或无响应用户暂时剔除,减少了服务器等待时间,降低了通信开销。实验结果表明,所提机制在实现隐私保护的同时,具有较高的模型准确率,较低的通信时延,并具有一定的可扩展性。

关键词: 联邦学习, 隐私保护技术, 同态加密, 医学图像

Abstract:

As an emerging training model with neural networks,federated learning has received widespread attention due to its ability to carry out model training on the premise of protecting user data privacy.However,since adversaries can track and derive participants’ privacy from the shared gradients,federated learning is still exposed to various security and privacy threats.Aiming at the privacy leakage problem of medical data in the process of federated learning,a secure and privacy-preserving medical data federated learning architecture is proposed based on Paillier homomorphic encryption technology (HEFLPS).First,the shared training model of the client is encrypted with Paillier homomorphic encryption technology to ensure the security and privacy of the training model,and a zero-knowledge proof identity authentication module is designed to ensure the credibility of the training members;second,the disconnected or unresponsive users are temporarily eliminated by constructing a message confirmation mechanism on the server side,which reduces the waiting time of the server and reduces the communication cost.Experimental results show that the proposed mechanism has high model accuracy,low communication delay and a certain scalability while achieving privacy protection.

Key words: federated learning, privacy-preserving techniques, homomorphic encryption, medical image

中图分类号:

TN393

王波,李洪涛,王洁,郭一娜. 面向医疗数据的隐私保护联邦学习架构[J]. 西安电子科技大学学报, 2023, 50(5): 166-177.

WANG Bo,LI Hongtao,WANG Jie,GUO Yina. Federated learning scheme for privacy-preserving of medical data[J]. Journal of Xidian University, 2023, 50(5): 166-177.

图/表 10

图1

图2

图3

图4

图5

图6

图7

图8

表3

图9

参考文献 28

[1]	CUI J, ZHU H, DENG H, et al. FeARH:Federated Machine Learningwith Anonymous Random Hybridization on Electronic Medical Records[J]. Journal of Biomedical Informatics, 2021, 117:103735. doi: 10.1016/j.jbi.2021.103735
[2]	QIAN F, ZHANG A. The Valueof Federated Learning During and Post-COVID-19[J]. International Journal for Quality in Health Care, 2021(1):1.
[3]	李雪莲, 张夏川, 高军涛, 等. 支持属性和代理重加密的区块链数据共享方案[J]. 西安电子科技大学学报, 2022, 49(1):1-16.
	LI Xuelian, ZHANG Xiachuan, GAO Juntao, et al. Blockchain Data Sharing Scheme Supporting Attribute and Proxy Re-Encryption[J]. Journal of Xidian University, 2022, 49(1):1-16.
[4]	SADILEK A, LIU L, NGUYEN D, et al. Privacy-First Health Researchwith Federated Learning[J]. NPJ Digital Medicine, 2021, 4(1):1-8. doi: 10.1038/s41746-020-00373-5
[5]	YIN L, FENG J, XUN H, et al. A Privacy-Preserving Federated Learning for Multiparty Data Sharing in Social Iots[J]. IEEE Transactions on Network Science and Engineering, 2021, 8(3):2706-2718. doi: 10.1109/TNSE.2021.3074185
[6]	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 Yixuan, CHEN Hong, LIU Yuhan, et al. Privacy-Preserving Techniques in Federated Learning[J]. Journal of Software, 2022, 33(3):1057-1092.
[7]	李尤慧子, 殷昱煜, 高洪皓, 等. 面向隐私保护的非聚合式数据共享综述[J]. 通信学报, 2021, 42(6):195-212. doi: 10.11959/j.issn.1000-436x.2021120
	LI Youhuizi, YIN Yuyu, GAO Honghao, et al. Survey on Privacy Protection in Non-Aggregated Data Sharing[J] Journal of Communication, 2021, 42(6):195-212. doi: 10.11959/j.issn.1000-436x.2021120
[8]	ZERKA F, BARAKAT S, WALSH S, et al. Systematic Review of Privacy-Preserving Distributed Machine Learning from Federated Databases in Health Care[J]. JCO Clinical Cancer Informatics, 2020, 4(4):184-200.
[9]	MOTHUKURI V, PARIZI R M, POURIYEH S, et al. A Survey on Security and Privacy of Federated Learning[J]. Future Generation Computer Systems, 2021, 115:619-640. doi: 10.1016/j.future.2020.10.007
[10]	KAIROUZ P, MCMAHAN H B, AVENT B, et al. Advances and Open Problems in Federated Learning[M]. Hanover: Now Foundations and Trends, 2021.
[11]	ZHANG J, CHEN B, CHENG X, et al. Poisongan:Generative Poisoning Attacks Against Federated Learning in Edge Computing Systems[J]. IEEE Internet of Things Journal, 2020, 8(5):3310-3322. doi: 10.1109/JIoT.6488907
[12]	LI Q, WEN Z, WU Z, et al. A Surveyon Federated Learning Systems:Vision,Hype and Reality for Data Privacy and Protection[J]. IEEE Transactions on Knowledge and Data Engineering, 2021, 35(4):3347-3366. doi: 10.1109/TKDE.2021.3124599
[13]	李瑞琪, 贾春福, 王雅飞. 基于NTRU的多密钥同态代理重加密方案及其应用[J]. 通信学报, 2021, 42(3):11-22. doi: 10.11959/j.issn.1000-436x.2021023
	LI Ruiqi, JIA Chunfu, WANG Yafei. Multi-Key Homomorphic Proxy Re-Encryption Scheme Based on NTRU and Its Application[J]. Journal of Communication, 2021, 42(3):11-22. doi: 10.11959/j.issn.1000-436x.2021023
[14]	徐花, 田有亮. 差分隐私下的权重社交网络隐私保护[J]. 西安电子科技大学学报, 2022, 49(1):17-25.
	XU Hua, TIAN Youliang.Protection of Privacy of the Weighted Social Network Under Differential Privacy[J]. Journal of Xidian University, 2022, 49(1):17-25.
[15]	FANG C, GUO Y, WANG N, et al. Highly Efficient Federated Learning with Strong Privacy Preservation in Cloud Computing[J]. Computers & Security, 2020, 96:101889. doi: 10.1016/j.cose.2020.101889
[16]	KU H, SUSILO W, ZHANG Y, et al. Privacy-Preserving Federated Learningin Medical Diagnosis with Homomorphic Re-Encryption[J]. Computer Standards & Interfaces, 2022, 80:103583. doi: 10.1016/j.csi.2021.103583
[17]	JIANG B, LI J, WANG H, et al. Privacy-Preserving Federated Learning for Industrial Edge Computing via Hybrid Differential Privacy and Adaptive Compression[J]. IEEE Transactions on Industrial Informatics, 2023, 19(2):1136-1144. doi: 10.1109/TII.2021.3131175
[18]	晏燕, 董卓越, 徐飞, 等. 一种Hilbert编码的本地化位置隐私保护方法[J]. 西安电子科技大学学报, 2023, 50(2):147-160.
	YAN Yan, DONG Zhuoyue, XU Fei, et al. Localized Location Privacy Protection Method Using the Hilbert Encoding[J]. Journal of Xidian University, 2023, 50(2):147-160.
[19]	LI T, SAHU A K, TALWALKA R, et al. Federated Learning:Challenges,Methods,and Future Directions[J]. IEEE Signal Processing Magazine, 2020, 37(3):50-60.
[20]	ZHOU Y, YE Q, LV J. Communication-Efficient Federated Learning with Compensated Overlap-Fedavg[J]. IEEE Transactions on Parallel and Distributed Systems, 2021, 33(1):192-205. doi: 10.1109/TPDS.2021.3090331
[21]	JIA B, ZHANG X, LIU J, et al. Blockchain-Enabled Federated Learning Data Protection Aggregation Scheme with Differential Privacy and Homomorphic Encryption in IoT[J]. IEEE Transactions on Industrial Informatics, 2021, 18(6):4049-4058. doi: 10.1109/TII.2021.3085960
[22]	KUMARI K A, SHARMA A, CHAKRABORTY C, et al. Preserving Health Care Data Security and Privacy Using Carmichael’s Theorem-Based Homomorphic Encryption and Modified Enhanced Homomorphic Encryption Schemes in Edge Computing Systems[J]. Big Data, 2022, 10(1):1-17. doi: 10.1089/big.2021.0012
[23]	SUN J, LI A, WANG B, et al. Soteria:Provable Defense Against Privacy Leakage in Federated Learning from Representation Perspective[C]//In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway:IEEE, 2021:9311-9319.
[24]	LI J, KUANG X, LIN S, et al. Privacy Preservation for Machine Learning Training and Classification Based on Homomorphic Encryption Schemes[J]. Information Sciences, 2020, 526:166-179. doi: 10.1016/j.ins.2020.03.041
[25]	DWIVEDI A D, SINGH R, GHOSH U, et al. Privacy Preserving Authentication System Based on Non-Interactive Zero Knowledge Proof Suitable for Internet of Things[J]. Journal of Ambient Intelligence and Humanized Computing, 2022, 13:4639-4649. doi: 10.1007/s12652-021-03459-4
[26]	MAJOR W, BUCHANAN W, AHMAD J, et al. An Authentication Protocol Based on Chaos and Zero Knowledge Proof[J]. Nonlinear Dynamics 2020, 99(4):3065-3087. doi: 10.1007/s11071-020-05463-3
[27]	ZHANG C, LI S, XIA J, et al. {BatchCrypt}:Efficient Homomorphic Encryption for {Cross-Silo} Federated Learning[C]//2020 USENIX Annual Technical Conference (USENIX ATC 20).Berkeley:USENIX, 2020:493-506.
[28]	PARK J, HAN D J, CHOI M, et al. Sageflow:Robust Federated Learning Against Both Stragglers and Adversaries[J]. Advances in Neural Information Processing Systems, 2021, 34:840-851.

	客户端数量	5	6	7	8	9	10
COVID-19	未使用文中框架总时长	49 959.60	59 949.46	69 939.37	79 929.28	89 919.19	99 910.10
	文中框架总时长	49 659.60	59 619.90	69 577.70	79 522.50	89 464.60	99 465.10
	差值	300.00	329.56	361.67	406.78	454.59	445.00
MNIST	未使用文中框架总时长	255.60	305.70	355.80	405.96	456.10	506.20
	文中框架总时长	215.63	255.84	295.25	333.37	374.58	417.80
	差值	39.97	49.86	60.55	72.59	81.52	88.40