基于激励机制的联邦学习优化算法

doi:10.11959/j.issn.1000-436x.2023095

通信学报 ›› 2023, Vol. 44 ›› Issue (5): 169-180.doi: 10.11959/j.issn.1000-436x.2023095

基于激励机制的联邦学习优化算法

田有亮¹^,²^,³, 吴柿红¹^,², 李沓¹^,², 王林冬¹^,², 周骅⁴

¹ 贵州大学公共大数据国家重点实验室，贵州贵阳 550025
² 贵州大学计算机科学与技术学院，贵州贵阳 550025
³ 贵州大学密码学与数据安全研究所，贵州贵阳 550025
⁴ 贵州大学大数据与信息工程学院，贵州贵阳 550025

修回日期:2023-04-06 出版日期:2023-05-25 发布日期:2023-05-01
作者简介:田有亮（1982- ），男，贵州盘州人，博士，贵州大学教授、博士生导师，主要研究方向为博弈论、密码学与安全协议、大数据隐私保护
吴柿红（1995- ），女，贵州福泉人，贵州大学硕士生，主要研究方向为联邦学习、密码学等
李沓（1998- ），男，贵州盘州人，贵州大学博士生，主要研究方向为密码学与区块链技术
王林冬（1997- ），男，浙江杭州人，贵州大学硕士生，主要研究方向为数字水印、信息安全等
周骅（1978- ），男，江苏无锡人，博士，贵州大学副教授、硕士生导师，主要研究方向为物联网安全、硬件安全机制、电路与系统
基金资助:
国家重点研发计划基金资助项目(2021YFB3101100);国家自然科学基金资助项目(U1836205);国家自然科学基金资助项目(62272123);贵州省高层次创新型人才基金资助项目([2020]6008);贵阳市科技计划基金资助项目([2021]1-5);贵阳市科技计划基金资助项目([2022]2-4);贵州省科技计划基金资助项目([2020]5017);贵州省科技计划基金资助项目([2022]065);贵州大学人才引进基金资助项目([2015]-53)

Federated learning optimization algorithm based on incentive mechanism

Youliang TIAN¹^,²^,³, Shihong WU¹^,², Ta LI¹^,², Lindong WANG¹^,², Hua ZHOU⁴

¹ State Key Laboratory of Public Big Data, Guizhou University, Guiyang 550025, China
² College of Computer Science and Technology, Guizhou University, Guiyang 550025, China
³ Institute of Cryptography ＆Data Security, Guizhou University, Guiyang 550025, China
⁴ College of Big Data and Information Engineering, Guizhou University, Guiyang 550025, China

Revised:2023-04-06 Online:2023-05-25 Published:2023-05-01
Supported by:
The Key Research and Development Program of China(2021YFB3101100);The National Natural Science Foundation of China(U1836205);The National Natural Science Foundation of China(62272123);Project of High-Level Innovative Talents of Guizhou Province([2020]6008);Science and Technology Program of Guiyang([2021]1-5);Science and Technology Program of Guiyang([2022]2-4);Science and Technology Program of Guizhou Province([2020]5017);Science and Technology Program of Guizhou Province([2022]065);Guizhou University Talent Introduction Research Fund([2015]-53)

摘要/Abstract

摘要：

针对联邦学习的训练过程迭代次数多、训练时间长、效率低等问题，提出一种基于激励机制的联邦学习优化算法。首先，设计与时间和模型损失相关的信誉值，基于该信誉值，设计激励机制激励拥有高质量数据的客户端加入训练。其次，基于拍卖理论设计拍卖机制，客户端通过向雾节点拍卖本地训练任务，委托高性能雾节点训练本地数据从而提升本地训练效率，解决客户端间的性能不均衡问题。最后，设计全局梯度聚合策略，增加高精度局部梯度在全局梯度中的权重，剔除恶意客户端，从而减少模型训练次数。

关键词: 联邦学习, 激励机制, 信誉值, 拍卖策略, 聚合策略

Abstract:

Federated learning optimization algorithm based on incentive mechanism was proposed to address the issues of multiple iterations, long training time and low efficiency in the training process of federated learning.Firstly, the reputation value related to time and model loss was designed.Based on the reputation value, an incentive mechanism was designed to encourage clients with high-quality data to join the training.Secondly, the auction mechanism was designed based on the auction theory.By auctioning local training tasks to the fog node, the client entrusted the high-performance fog node to train local data, so as to improve the efficiency of local training and solve the problem of performance imbalance between clients.Finally, the global gradient aggregation strategy was designed to increase the weight of high-precision local gradient in the global gradient and eliminate malicious clients, so as to reduce the number of model training.

Key words: federated learning, incentive mechanism, reputation value, auction strategy, aggregation strategy

中图分类号:

TN92

田有亮, 吴柿红, 李沓, 王林冬, 周骅. 基于激励机制的联邦学习优化算法[J]. 通信学报, 2023, 44(5): 169-180.

Youliang TIAN, Shihong WU, Ta LI, Lindong WANG, Hua ZHOU. Federated learning optimization algorithm based on incentive mechanism[J]. Journal on Communications, 2023, 44(5): 169-180.

图/表 14

图1

表1

图2

图3

表2

表3

表4

图4

图5

图6

图7

图8

表5

表6

参考文献 34

[1]	GE Z Q , SONG Z H , DING S X ,et al. Data mining and analytics in the process industry:the role of machine learning[J]. IEEE Access, 2017,5: 20590-20616.
[2]	MOHASSEL P , ZHANG Y P . Secure ML:a system for scalable privacy-preserving machine learning[C]// Proceedings of 2017 IEEE Symposium on Security and Privacy (SP). Piscataway:IEEE Press, 2017: 19-38.
[3]	ZHANG Y , XU C X , LI H W ,et al. HealthDep:an efficient and secure deduplication scheme for cloud-assisted eHealth systems[J]. IEEE Transactions on Industrial Informatics, 2018,14(9): 4101-4112.
[4]	AHISKA K , OZGOREN M K , LEBLEBICIOGLU M K . Autopilot design for vehicle cornering through icy roads[J]. IEEE Transactions on Vehicular Technology, 2018,67(3): 1867-1880.
[5]	YANG Q , LIU Y , CHEN T J ,et al. Federated machine learning:concept and applications[J]. ACM Transactions on Intelligent Systems and Technology, 2019,10(2): 1-19.
[6]	PANDEY S R , TRAN N H , BENNIS M ,et al. A crowdsourcing framework for on-device federated learning[J]. IEEE Transactions on Wireless Communications, 2020,19(5): 3241-3256.
[7]	ZHAN Y F , LI P , QU Z H ,et al. A learning-based incentive mechanism for federated learning[J]. IEEE Internet of Things Journal, 2020,7(7): 6360-6368.
[8]	LIU J , WANG J H , PONG C ,et al. FedPA:an adaptively partial model aggregation strategy in Federated Learning[J]. Computer Networks, 2021,199:108468.
[9]	MCMAHAN H B , MOORE E , RAMAGE D ,et al. Communication-efficient learning of deep networks from decentralized data[J]. arXiv Preprint,arXiv:1602.05629, 2016.
[10]	BONOMI F , MILITO R , ZHU J ,et al. Fog computing and its role in the Internet of things[C]// Proceedings of the First Edition of the MCC Workshop on Mobile Cloud Computing. New York:ACM Press, 2012: 13-16.
[11]	LI Y J , TAO X F , ZHANG X F ,et al. Privacy-preserved federated learning for autonomous driving[J]. IEEE Transactions on Intelligent Transportation Systems, 2022,23(7): 8423-8434.
[12]	KIM H , PARK J , BENNIS M ,et al. Blockchained on-device federated learning[J]. IEEE Communications Letters, 2020,24(6): 1279-1283.
[13]	陈明鑫, 张钧波, 李天瑞 . 联邦学习攻防研究综述[J]. 计算机科学, 2022,49(7): 310-323.
	CHEN M X , ZHANG J B , LI T R . Survey on attacks and defenses in federated learning[J]. Computer Science, 2022,49(7): 310-323.
[14]	LI Y P , COURCOUBETIS C , DUAN L J . Recommending paths:follow or not follow?[C]// Proceedings of IEEE Conference on Computer Communications. Piscataway:IEEE Press, 2019: 928-936.
[15]	KANG J W , XIONG Z H , 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(6): 10700-10714.
[16]	ZENG R F , ZHANG S X , WANG J Q ,et al. FMore:an incentive scheme of multi-dimensional auction for federated learning in MEC[C]// Proceedings of 2020 IEEE 40th International Conference on Distributed Computing Systems (ICDCS). Piscataway:IEEE Press, 2021: 278-288.
[17]	DENG Y H , LYU F , REN J ,et al. FAIR:quality-aware federated learning with precise user incentive and model aggregation[C]// Proceedings of IEEE Conference on Computer Communications. Piscataway:IEEE Press, 2021: 1-10.
[18]	LIAN X R , HUANG Y J , LI Y C ,et al. Asynchronous parallel stochastic gradient for nonconvex optimization[C]// Proceedings of the 28th International Conference on Neural Information Processing Systems. New York:ACM Press, 2015: 2737-2745.
[19]	DAI W , KUMAR A , WEI J L ,et al. High-performance distributed ML at scale through parameter server consistency models[C]// Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto:AAAI Press, 2015: 79-87.
[20]	NIU F , RECHT B , RE C ,et al. HOGWILD!:a lock-free approach to parallelizing stochastic gradient descent[J]. Advances in Neural Information Processing Systems, 2011,24: 693-701.
[21]	NISHIO T , YONETANI R . Client selection for federated learning with heterogeneous resources in mobile edge[C]// Proceedings of 2019 IEEE International Conference on Communications (ICC). Piscataway:IEEE Press, 2019: 1-7.
[22]	余佳仁, 田有亮, 林晖 . 基于信誉管理模型的矿工类型鉴别机制设计[J]. 网络与信息安全学报, 2022,8(1): 128-138.
	YU J R , TIAN Y L , LIN H . Design of miner type identification mechanism based on reputation management model[J]. Chinese Journal of Network and Information Security, 2022,8(1): 128-138.
[23]	王勇, 李国良, 李开宇 . 联邦学习贡献评估综述[J]. 软件学报, 2023,34(3): 1168-1192.
	WANG Y , LI G L , LI K Y . Survey on contribution evaluation for federated learning[J]. Journal of Software, 2023,34(3): 1168-1192.
[24]	XU G W , LI H W , LIU S ,et al. VerifyNet:secure and verifiable federated learning[J]. IEEE Transactions on Information Forensics and Security, 2020,15: 911-926.
[25]	WANG S Q , TUOR T , SALONIDIS T ,et al. When edge meets learning:adaptive control for resource-constrained distributed machine learning[C]// Proceedings of IEEE 2018 IEEE Conference on Computer Communications. Piscataway:IEEE Press, 2018: 63-71.
[26]	ZHENG S Y , CAO Y , YOSHIKAWA M . Incentive mechanism for privacy-preserving federated learning[J]. arXiv Preprint,arXiv:2106.04384, 2021.
[27]	MIAO Y B , LIU Z T , LI H W ,et al. Privacy-preserving Byzantine-robust federated learning via blockchain systems[J]. IEEE Transactions on Information Forensics and Security, 2022,17: 2848-2861.
[28]	XIONG L , LIU L . PeerTrust:supporting reputation-based trust for peer-to-peer electronic communities[J]. IEEE Transactions on Knowledge and Data Engineering, 2004,16(7): 843-857.
[29]	HUANG C Y , WANG Z Y , CHEN H X ,et al. RepChain:a reputation-based secure,fast,and high incentive blockchain system via sharding[J]. IEEE Internet of Things Journal, 2021,8(6): 4291-4304.
[30]	TRAN N H , BAO W , ZOMAYA A ,et al. Federated learning over wireless networks:optimization model design and analysis[C]// Proceedings of 2019 IEEE Conference on Computer Communications. Piscataway:IEEE Press, 2019: 1387-1395.
[31]	ABDALI T A N , HASSAN R , AMAN A H M ,et al. Fog computing advancement:concept,architecture,applications,advantages,and open issues[J]. IEEE Access, 2021,9: 75961-75980.
[32]	ZHOU C Y , FU A M , YU S ,et al. Privacy-preserving federated learning in fog computing[J]. IEEE Internet of Things Journal, 2020,7(11): 10782-10793.
[33]	WAN C B , JIN F S , QIAO Z ,et al. Unsupervised active learning with loss prediction[J]. Neural Computing and Applications, 2023,35(5): 3587-3595.
[34]	FU A M , ZHANG X L , XIONG N X ,et al. VFL:a verifiable federated learning with privacy-preserving for big data in industrial IoT[J]. IEEE Transactions on Industrial Informatics, 2022,18(5): 3316-3326.

参数	描述
data_size	本地训练数据的大小
P_i	雾节点i竞拍的价格
T_i	雾节点i一次迭代的时间
c_n	执行一个数据样本的CPU周期数
f_i	设备i的频率
	客户端i用于计算的能量消耗
	客户端i在第n轮训练后的模型损失
	第n轮训练后的平均模型损失
d	任务发布者与客户端签署合约需要交纳的押金
r	客户端与雾节点签署合约需要交纳的押金
task_x	任务x
curret_task	当前任务
	客户端i的第 j轮训练梯度
w^j	第 j轮训练的全局梯度
T_ex	任务发布者期望的总训练时间
T_r	实际训练的总时间
t ^ji	客户端i在第 j轮训练所用的时间

方案	MNIST	CIFAR-10
本文方案	96.3%	96.1%
FAIR方案	51.8%	51.8%

方案	MNIST/s	CIFAR-10/s
本文方案	421.4	391.58
FAIR方案	611.4	568.85

方案	MNIST/s	CIFAR-10/s
本文方案	296.03+	1 579.82+
FedAvg方案	550.37	2 682.26
FAIR方案	467.92	2 490.69

方案	客户端激励	流浪者效应	聚合策略
FedPA方案	×	√	×
FedAvg方案	×	×	×
FAIR方案	√	×	√
本文方案	√	√	√

基于激励机制的联邦学习优化算法

Federated learning optimization algorithm based on incentive mechanism

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 34

相关文章 15

Metrics

推荐阅读 0

方案	MNIST（96%）/轮	CIFAR-10（60%）/轮
本文方案	16	86
FedAvg方案	20	98
FAIR方案	17	91

[1]	马鑫迪, 李清华, 姜奇, 马卓, 高胜, 田有亮, 马建峰. 面向Non-IID数据的拜占庭鲁棒联邦学习[J]. 通信学报, 2023, 44(6): 138-153.
[2]	金彪, 李逸康, 姚志强, 陈瑜霖, 熊金波. GenFedRL：面向深度强化学习智能体的通用联邦强化学习框架[J]. 通信学报, 2023, 44(6): 183-197.
[3]	李开菊, 许强, 王豪. 冗余数据去除的联邦学习高效通信方法[J]. 通信学报, 2023, 44(5): 79-93.
[4]	余晟兴, 陈泽凯, 陈钟, 刘西蒙. DAGUARD：联邦学习下的分布式后门攻击防御方案[J]. 通信学报, 2023, 44(5): 110-122.
[5]	姜慧, 何天流, 刘敏, 孙胜, 王煜炜. 面向异构流式数据的高性能联邦持续学习算法[J]. 通信学报, 2023, 44(5): 123-136.
[6]	张佳乐, 朱诚诚, 孙小兵, 陈兵. 基于GAN的联邦学习成员推理攻击与防御方法[J]. 通信学报, 2023, 44(5): 193-205.
[7]	余晟兴, 陈钟. 基于同态加密的高效安全联邦学习聚合框架[J]. 通信学报, 2023, 44(1): 14-28.
[8]	汤凌韬, 王迪, 刘盛云. 面向非独立同分布数据的联邦学习数据增强方案[J]. 通信学报, 2023, 44(1): 164-176.
[9]	范绍帅, 吴剑波, 田辉. 面向能量受限工业物联网设备的联邦学习资源管理[J]. 通信学报, 2022, 43(8): 65-77.
[10]	莫梓嘉, 高志鹏, 杨杨, 林怡静, 孙山, 赵晨. 面向车联网数据隐私保护的高效分布式模型共享策略[J]. 通信学报, 2022, 43(4): 83-94.
[11]	康海燕, 冀源蕊. 基于本地化差分隐私的联邦学习方法研究[J]. 通信学报, 2022, 43(10): 94-105.
[12]	陶梅霞, 王栋, 孙瑞, 张乃夫. 联邦学习中基于时分多址接入的用户调度策略[J]. 通信学报, 2021, 42(6): 1-29.
[13]	贺文晨, 郭少勇, 邱雪松, 陈连栋, 张素香. 基于DRL的联邦学习节点选择方法[J]. 通信学报, 2021, 42(6): 62-71.
[14]	李尤慧子, 殷昱煜, 高洪皓, 金一, 王新珩. 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.
[15]	黄永明, 郑冲, 张征明, 尤肖虎. 大规模无线通信网络移动边缘计算和缓存研究[J]. 通信学报, 2021, 42(4): 44-61.