基于在线双边拍卖的分层联邦学习激励机制

doi:10.11896/jsjkx.210800051

计算机科学 ›› 2022, Vol. 49 ›› Issue (3): 23-30.doi: 10.11896/jsjkx.210800051

• 新兴分布式计算技术与系统^* • 上一篇下一篇

基于在线双边拍卖的分层联邦学习激励机制

杜辉^1,2, 李卓^1,2, 陈昕²

1 网络文化与数字传播北京市重点实验室(北京信息科技大学) 北京100101
2 北京信息科技大学计算机学院北京100101

收稿日期:2021-08-05 修回日期:2021-10-19 出版日期:2022-03-15 发布日期:2022-03-15
通讯作者: 李卓(lizhuo@bistu.edu.cn)
作者简介:(duhui199801@163.com)
基金资助:
国家自然科学基金(61872044);北京市青年拔尖人才项目;网络与文化传播北京市重点实验室开放课题

Incentive Mechanism for Hierarchical Federated Learning Based on Online Double Auction

DU Hui^1,2, LI Zhuo^1,2, CHEN Xin²

1 Beijing Key Laboratory of Internet Culture and Digital Dissemination Research (Beijing Information Science & Technology University),Beijing 100101,China
2 School of Computer Science,Beijing Information Science & Technology University,Beijing 100101,China

Received:2021-08-05 Revised:2021-10-19 Online:2022-03-15 Published:2022-03-15
About author:DU Hui,born in 1998,postgraduate.His main research interests include edge computing and so on.
LI Zhuo,born in 1983,Ph.D,associate professor,is a member of China Computer Federation.His main research interests include mobile wireless network and distributed computing.
Supported by:
National Natural Science Foundation of China(61872044),Beijing Municipal Program for Top Talent and Open Program of Beijing Key Laboratory of Internet Culture and Digital Dissemination Research.

摘要/Abstract

摘要： 在分层联邦学习中,能量受限的移动设备参与模型训练会消耗自身资源。为了降低移动设备的能耗,文中在不超过分层联邦学习的最大容忍时间下,提出了移动设备能耗之和最小化问题。不同训练轮次的边缘服务器能够选择不同的移动设备,移动设备也能够为不同的边缘服务器并发训练模型,因此文中基于在线双边拍卖机制提出了ODAM-DS算法。基于最优停止理论,支持边缘服务器在合适的时刻选择移动设备,使得移动设备的平均能耗最小,然后对提出的在线双边拍卖机制进行理论分析,证明其满足激励相容性、个体理性、弱预算均衡约束等特性。模拟实验的结果证明,ODAM-DS算法产生的能耗比已有的HFEL算法平均降低了19.04%。

关键词: 分层联邦学习, 激励机制设计, 能耗最小化, 在线双边拍卖, 最优停止理论

Abstract: In hierarchical federated learning,energy constrained mobile devices will consume their own resources for participating in model training.In order to reduce the energy consumption of mobile devices,this paper proposes the problem of minimizing the sum of energy consumption of mobile devices without exceeding the maximum tolerance time of hierarchical federated learning.Different training rounds of edge server can select different mobile devices,and mobile devices can also train models under diffe-rent edge servers concurrently.Therefore,this paper proposes ODAM-DS algorithm based on an online double auction mechanism.Based on the optimal stopping theory,the edge server is supported to select the mobile device at the best time,so as to minimize the average energy consumption of the mobile device.Then,the theoretical analysis of the proposed online double auction mechanism proves that it meets the characteristics of incentive compatibility,individual rationality and weak budget equilibrium constraints.Simulation results show that the energy consumption of ODAM-DS algorithm is 19.04% lower than that of the existing HFEL algorithm.

Key words: Hierarchical federated learning, Incentive mechanism design, Minimization of energy consumption, Online double auction, Optimal stopping theory

中图分类号:

TP393

杜辉, 李卓, 陈昕. 基于在线双边拍卖的分层联邦学习激励机制[J]. 计算机科学, 2022, 49(3): 23-30. https://doi.org/10.11896/jsjkx.210800051

DU Hui, LI Zhuo, CHEN Xin. Incentive Mechanism for Hierarchical Federated Learning Based on Online Double Auction[J]. Computer Science, 2022, 49(3): 23-30. https://doi.org/10.11896/jsjkx.210800051

参考文献

[1]ZHOU Z,CHEN X,LI E,et al.Edge intelligence:paving the last mile of artificial intelligence with edge computing[J].Procee-dings of the IEEE,2019,107(8):1738-1762.
[2]MCMAHAN H B,MOORE E,RAMAGE D,et al.Communication-efficient learning of deep networks from decentralized data[J].arXiv:1602.05629v2,2016.
[3]LIM W Y B,LUONG N C,HOANG D H,et al.Federated lear-ning in mobile edge networks:A comprehensive survey[J].IEEE Communications Surveys & Tutorials,2020,22(3):2031-2063.
[4]BONAWITZ K,EICHNER H,GRIESKAM P,et al.Towardsfederated learning at scale:System design[J].arXiv:1902.01046,2019.
[5]WANG S Q,TOUR T,SALONIDIS T et al.Adaptive federated learning in resource constrained edge computing systems[J].IEEE Journal on Selected Areas in Communications (JSAC),2019,37(6):1205-1221.
[6]LIU L,ZHANG J,SONG S H,et al.Client-edge-cloud hierar-chical federated learning[J].arXiv:1095.06641v1,2019.
[7]LUO S,CHEN X,WU Q,et al.HFEL:Joint edge associationand resource allocation for cost-efficient hierarchical federated edge learning[J].arXiv:2002.11343,2020.
[8]KHAN L U,PANDEY S R,TRAN N H,et al.Federated lear-ning for edge networks:Resource optimization and incentive mechanism[J].IEEE Communications Magazine,2020,58(10):88-93.
[9]CHAI H,LENG S,CHEN Y,et al.A Hierarchical blockchain-enabled federated learning algorithm for knowledge sharing in internet of vehicles[J].IEEE Transactions on Intelligent Transportation Systems,2020,22(7):3975-3986.
[10]ZHAN Y,ZHANG J.An incentive mechanism design for efficient edge learning by deep reinforcement learning approach[C]//IEEE INFOCOM:2020 IEEE Conference on Computer Communications.Toronto,ON,Canada:IEEE,2020:1-10.
[11]TANG J.Nodes incentive strategy based on bayesian game in Ad Hoc networks[J].Computer Engineering,2019,45(8):152-158,164.
[12]ZHOU Q,LI P,NIE L.User incentive mechanism based on spatial-temporal correlation for crowd sensing[J].Computer Engineering,2021,47(3):227-236.
[13]ZENG R,ZHANG S,WANG J,et al.FMore:An IncentiveScheme of Multi-dimensional Auction for Federated Learning in MEC[C]//2020 IEEE 40th International Conference on Distri-buted Computing Systems(ICDCS 2020).Singapore,IEEE,2020:278-288.
[14]ZHANG X,YANG Z,ZHOU Z,et al.Free market of crowd-sourcing:incentive mechanism design for mobile sensing[J].IEEE Transactions on Parallel and Distributed Systems,2014,25(12):3190-3200.
[15]LI D,YANG Q,AN D,et al.On location privacy-preserving online double auction for electric vehicles in microgrids[J].IEEE Internet of Things Journal,2019,6(4):5902-5915.
[16]ZHAN Y,LI P,QU Z,et al.A learning-based incentive mechanism for federated learning[J].IEEE Internet of Things Journal,2020,7(7):6360-6368.
[17]LE T H T,TRAN N H,TUN Y K,et al.Auction based incentive design for efficient federated learning in cellular wireless networks[C]//WCNC:2020 IEEE Wireless Communications and Networking Conference.Seoul,Korea (South):IEEE,2020:1-6.
[18]PENG Y,WANG G C,HUANG S Q,et al.An energy consump-tion optimization strategy for data transmission based on optimal stopping theory in mobile network[J].Chinese Journal of Computers,2016,39(6):1162-1175.
[19]WANG F S,WANG G C.Study on energy minimization data transmission strategy in mobile cloud computing[C]//SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI:2018 IEEE SmartWorld,Ubiquitous Intelligence & Computing,Advanced &Trusted Computing,Scalable Computing & Communications,Cloud & Big Data Computing,Internet of People and Smart City Innovation.Guangzhou,China:IEEE,2018:1211-1218.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于在线双边拍卖的分层联邦学习激励机制

Incentive Mechanism for Hierarchical Federated Learning Based on Online Double Auction

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

Metrics

本文评价

推荐阅读 0

[1]	黄荣喜, 王淖, 谢天骁, 王高才. 无线网络中具有信道感知的期望能耗最小化策略研究 Study on Channel-aware Expected Energy Consumption Minimization Strategy in Wireless Networks 计算机科学, 2018, 45(10): 130-137. https://doi.org/10.11896／j.issn.1002-137X.2018.10.025
[2]	康雁. 优化能耗的可变电压禁忌任务调度算法 Variable Voltage Tabu Task Scheduling Algorithm for Optimizing Energy Consumption 计算机科学, 2010, 37(10): 287-290.