计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (2): 1-7.doi: 10.11896/jsjkx.240100023

所属专题： 联邦学习

• 学科前沿 • 上一篇    下一篇

联邦学习通信效率研究综述

郑剑文1, 刘波1, 林伟伟2,3, 谢家晨1   

  1. 1 华南师范大学计算机学院 广州 510631
    2 鹏城实验室 广东 深圳 518066
    3 华南理工大学计算机科学与工程学院 广州 510640
  • 收稿日期:2024-01-02 修回日期:2024-05-22 出版日期:2025-02-15 发布日期:2025-02-17
  • 通讯作者: 林伟伟(linww@scut.edu.cn)
  • 作者简介:(1024109066@qq.com)
  • 基金资助:
    国家自然科学基金面上项目(62072187);广州市开发区国际合作项目(2023GH02);鹏城实验室重大任务项目(PCL2023A09)

Survey of Communication Efficiency for Federated Learning

ZHENG Jianwen1, LIU Bo1, LIN Weiwei2,3, XIE Jiachen1   

  1. 1 School of Computer Science,South China Normal University,Guangzhou 510631,China
    2 Pengcheng Laboratory,Shenzhen,Guangdong 518066,China
    3 School of Computer Science and Engineering,South China University of Technology,Guangzhou 510640,China
  • Received:2024-01-02 Revised:2024-05-22 Online:2025-02-15 Published:2025-02-17
  • About author:ZHENG Jianwen,born in 2000,postgraduate.His main research interest is federated learning.
    LIN Weiwei,born in 1980,Ph.D,professor,is a distinguished member of CCF(No.37313D).His main research interestes include cloud computing,big data technology and AI application technology.
  • Supported by:
    National Natural Science Foundation of China(62072187),Guangzhou Development Zone Science and Technology Project(2023GH02) and Major Key Project of PCL,China(PCL2023A09).

PDF (PC)

283

摘要： 作为一种分布式机器学习范式,联邦学习(Federated Learning,FL)旨在在保护数据隐私的前提下,实现在多方数据上共同训练机器学习模型。在实际应用中,FL在每轮迭代中需要大量的通信来传输模型参数和梯度更新,从而提高通信效率,这是FL面临的一个重要挑战。文中主要介绍了FL中通信效率的重要性,并依据不同的侧重点将现有FL通信效率的研究分为客户端选择、模型压缩、网络拓扑重构以及多种技术结合等方法。在现有的FL通信效率研究的基础上,归纳并总结出通信效率在FL发展中面临的困难与挑战,探索FL通信效率未来的研究方向。

关键词: 联邦学习, 通信效率, 客户端选择, 模型压缩, 网络拓扑重构

Abstract: As a distributed machine learning paradigm,federated learning(FL) aims to collaboratively train machine learning models on decentralized data sources while ensuring data privacy.However,in practical applications,FL faces the challenge of communication efficiency,as significant communication is required in each iteration to transmit model parameters and gradient updates,leading to communication costs far surpass computation costs.Thus,effectively enhancing communication efficiency poses a significant challenge in FL research.This paper mainly introduces the importance of communication efficiency in FL,and divides the existing research on FL communication efficiency into client selection,model compression,network topology reconstruction,and the combination of multiple technologies according to different emphases.On the basis of the existing research on FL communication efficiency,this paper summarizes the difficulties and challenges in communication efficiency in the development of FL,and explores the future research direction of FL communication efficiency.

Key words: Federated learning, Communication efficiency, Client selection, Model compression, Network topology refactoring

中图分类号: 

  • TP181
[1]ZHANG C,XIE Y,BAI H,et al.A survey on federated learning[J].Knowledge-Based Systems,2021,216:106775.
[2]MAMMEN P M.Federated learning:Opportunities and challenges [EB/OL].(2021-01-14) [2024-03-12].https://arxiv.org/abs/2101.05428.
[3]SHAHID O,POURIYEH S,PARIZI R M,et al.Communication efficiency in federated learning:Achievements and challenges [EB/OL].(2021-07-23)[2024-03-12].https://arxiv.org/abs/2107.10996.
[4]ASAD M,MOUSTAFA A,ITO T,et al.Evaluating the communication efficiency in federated learning algorithms[C]//2021 IEEE 24th International Conference on Computer Supported Cooperative Work in Design(CSCWD).IEEE,2021:552-557.
[5]WEN J,ZHANG Z,LAN Y,et al.A survey on federated lear-ning:challenges and applications[J].International Journal of Machine Learning and Cybernetics,2023,14(2):513-535.
[6]KAIROUZ P,MCMAHAN H B,AVENT B,et al.Advances and open problems in federated learning[J].Foundations and Trends in Machine Learning,2021,14(1/2):214-217.
[7]LIU B,LV N,GUO Y,et al.Recent Advances on Federated Learning:A Systematic Survey[EB/OL].(2023-01-03)[2024-03-12].https://arxiv.org/abs/2301.01299.
[8]POURIYEH S,SHAHID O,PARIZI R M,et al.Securesmart communication efficiency in federated learning:Achievements and challenges[J].Applied Sciences,2022,12(18):8980.
[9]SHAHEEN M,FAROOQ M S,UMER T,et al.Applications of federated learning;Taxonomy,challenges,and research trends[J].Electronics,2022,11(4):670.
[10]WNAG L P,WEI W,LI B.CMFL:Mitigating communicationoverhead for federated learning[C]//2019 IEEE 39th International Conference on Distributed Computing Systems(ICDCS).IEEE,2019:954-964.
[11]LI C,ZENG X,ZHANG M,et al.PyramidFL:A fine-grainedclient selection framework for efficient federated learning[C]//Proceedings of the 28th Annual International Conference on Mobile Computing and Networking.2022:158-171.
[12]WANG H,KAPLAN Z,NIU D,et al.Optimizing federatedlearning on non-iid data with reinforcement learning[C]//IEEE INFOCOM 2020－IEEE Conference on Computer Communications.IEEE,2020:1698-1707.
[13]NISHIO T,YONETANI R.Client selection for federated lear-ning with heterogeneous resources in mobile edge[C]//2019 IEEE International Conference on Communications(ICC).IEEE,2019:1-7.
[14]DU C,XIAO J,GUO W.Bandwidth constrained client selection and scheduling for federated learning over SD-WAN[J].IET Communications,2022,16(2):187-194.
[15]XU J,DU W,JIN Y,et al.Ternary compression for communication-efficient federated learning[J].IEEE Transactions on Neural Networks and Learning Systems,2020,33(3):1162-1176.
[16]MALEKIJOO A,FADAEIESLAM M J,MALEKIJOU H,et al.Fedzip:A compression framework for communication-efficient federated learning[EB/OL].(2021-02-02)[2024-03-12].https://arxiv.org/abs/2102.01593.
[17]AJI A F,HEAFIELD K.Sparse Communication for Distributed Gradient Descent[C]//Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing.2017:440-445.
[18]HUFFMAN D A.A method for the construction of minimum-redundancy codes[J].Proceedings of the IRE,1952,40(9):1098-1101.
[19]ZHU X,WANG J,CHEN W,et al.Model compression and privacy preserving framework for federated learning[J].Future Generation Computer Systems,2023,140:376-389.
[20]ZHAO J,ZHU H,WANG F,et al.CORK:A privacy-preserving and lossless federated learning scheme for deep neural network[J].Information Sciences,2022,603:190-209.
[21]HE Y,WANG H P,ZENK M,et al.CosSGD:Communication-efficient federated learning with a simple cosine-based quantization [EB/OL].(2020-12-15)[2024-03-12].https://arxiv.org/abs/2012.08241.
[22]HU C,JIANG J,WANG Z.Decentralized federated learning:A segmented gossip approach[EB/OL].(2019-08-21)[2024-03-12].https://arxiv.org/abs/1908.07782.
[23]WANG Z,HU Y,YAN S,et al.Efficient ring-topology decentralized federated learning with deep generative models for medical data in ehealthcare systems[J].Electronics,2022,11(10):1548.
[24]ABDELLATIF A A,MHAISEN N,MOHAMED A,et al.Communication-efficient hierarchical federated learning for IoT he-terogeneous systems with imbalanced data[J].Future Generation Computer Systems,2022,128:406-419.
[25]LU X,LIAO Y,LIO P,et al.Privacy-preserving asynchronousfederated learning mechanism for edge network computing[J].IEEE Access,2020,8:48970-48981.
[26]LIU S,YU G,YIN R,et al.Joint model pruning and device selection for communication-efficient federated edge learning[J].IEEE Transactions on Communications,2021,70(1):231-244.
[27]BUYUKATES B,ULUKUS S.Timely communication in fede-rated learning[C]//IEEE Conference on Computer Communications Workshops(INFOCOM WKSHPS).IEEE,2021:1-6.
[28]CALDAS S,KONECNY J,MCMAHAN H B,et al.Expanding the reach of federated learning by reducing client resource requirements[EB/OL].(2018-12-18)[2024-03-12].https://ar-xiv.org/abs/1812.07210.
[29]CHEN Y,SUN X,JIN Y.Communication-efficient federated deep learning with layerwise asynchronous model update and temporally weighted aggregation[J].IEEE Transactions on Neural Networks and Learning Systems,2019,31(10):4229-4238.
[30]PARK S,SUH Y,LEE J.FedPSO:Federated learning using particle swarm optimization to reduce communication costs[J].Sensors,2021,21(2):600.
[31]MILLS J,HU J,MIN G.Communication-efficient federatedlearning for wireless edge intelligence in IoT[J].IEEE Internet of Things Journal,2019,7(7):5986-5994.
[32]ASAD M,MOUSTAFA A,ITO T.FedOpt:Towards communication efficiency and privacy preservation in federated learning[J].Applied Sciences,2020,10(8):2864.
[33]LI C,LI G,VARSHNEY P K.Communication-efficient federated learning based on compressed sensing[J].IEEE Internet of Things Journal,2021,8(20):15531-15541.
[34]WU C,WU F,LYU L,et al.Communication-efficient federated learning via knowledge distillation[J].Nature Communications,2022,13(1):2032.
[35]ZHU H,JIN Y.Multi-objective evolutionary federated learning[J].IEEE Transactions on Neural Networks and Learning Systems,2019,31(4):1310-1322.
[36]MARFOQ O,XU C,NEGLIA G,et al.Throughput-optimal topology design for cross-silo federated learning[J].Advances in Neural Information Processing Systems,2020,33:19478-19487.
[37]GUPTA R,ALAM T.Survey on federated-learning approaches in distributed environment[J].Wireless Personal Communications,2022,125(2):1631-1652.
[38]ZHANG J,ZHU H,WANG F,et al.Security and privacythreats to federated learning:Issues,methods,and challenges[J].Security and Communication Networks,2022,2022(1):2886795.
[39]SONG D,SHEN G,GAO D,et al.Fast heterogeneous federated learning with hybrid client selection[C]//Uncertainty in Artificial Intelligence.PMLR,2023:2006-2015.
[40]XUE Y,LAU V.Riemannian Low-Rank Model Compression for Federated Learning with Over-the-Air Aggregation[J/OL].https://onlinelibrary.wiley.com/doi/10.1155/2022/2886795.
[41]ZITZLER E,LAUMANNS M,THIELE L.SPEA2:Improving the strength Pareto evolutionary algorithm[C]//Proceedings of the EUROGEN'2001.2001.
[42]COELLO C A C,LECHUGA M S.MOPSO:A proposal formultiple objective particle swarm optimization[C]//Procee-dings of the 2002 Congress on Evolutionary Computation,CEC'02(Cat,No,02TH8600).IEEE,2002,2:1051-1056.
[43]PUTRA M A P,PUTRI A R,ZAINUDIN A,et al.Acs:Accuracy-based client selection mechanism for federated industrial iot[J].Internet of Things,2023,21:100657.
[44]RAI S,KUMARI A,PRASAD D K.Client selection in federatedlearning under imperfections in environment[J].AI,2022,3(1):124-145.
[45]CHAI Z Y,YANG C,LI Y L.Communication efficiency optimization in federated learning based on multi-objective evolutio-nary algorithm[J].Evolutionary Intelligence,2023,16(3):1033-1044.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发