Journal of Jilin University(Engineering and Technology Edition) ›› 2025, Vol. 55 ›› Issue (10): 3337-3345.doi: 10.13229/j.cnki.jdxbgxb.20231369

Previous Articles    

Resource-efficient clustering collaborative federated learning client selection method

Qiang LI(),Ling-yu ZHANG,Xiang-yu MENG()   

  1. College of Computer Science and Technology，Jilin University，Changchun 130012，China
  • Received:2023-12-08 Online:2025-10-01 Published:2026-02-03
  • Contact: Xiang-yu MENG E-mail:li_qiang@jlu.edu.cn;xiangyumeng@jlu.edu.cn

RICH HTML

 0   

PDF (PC)

23

Abstract:

To address the key issue in federated learning the resource heterogeneity and data heterogeneity of each client， this paper proposes a resource-efficient client selection method for clustered collaborative federated learning. Firstly， where each client is grouped according to its computational power， the average accuracy of each group of clients is used as an indirect metric to select clients in the same group in each round of training. Secondly， the clients are clustered according to the model similarity of each client within each group， and the clients in different clusters within each group are selected. Evaluate the performance of the method proposed in this paper on real datasets， the experimental results show that this method can reduce the global training time， obtain faster and smoother convergence， and achieve a good balance between training efficiency and global model accuracy.

Key words: computer system architecture, federated learning, client selection, clustering, resource heterogeneity, data heterogeneity

CLC Number: 

  • TP301

Fig.1

Federated learning"

Fig.2

Structure of proposed method"

Table 1

Dataset statistics table"

数据集MNISTCIFAR10
训练集大小60 00050 000
测试集大小10 00010 000
特征数量28×283×28×28
分类数量1010

Table 2

Training model accuracy"

数据集场景FedAvg方法本文方法
MNISTIID96.9597.12
Non-IID91.4494.41
CIFAR10IID69.3970.12
Non-IID58.961.41

Fig.3

Effect of model accuracy in different scenarios"

Table 3

Total training time"

数据集场景FedAvg方法本文方法
MNISTIID16701 302
Non-IID1 5441 210
CIFAR10IID1 7321 509
Non-IID1 8941 631

Table 4

Training time and number of rounds required to achieve corresponding accuracy in MNIST dataset"

场景方 法准确率总体/%
85%90%93%95%
轮数时间/s轮数时间/s轮数时间/s轮数时间/s
IIDFedAvg317.701268.1931176.9066376.6096.95
本文313.921043.0130125.7362270.0397.12
Non-IIDFedAvg83494.942121 186.58----91.44
本文35222.7879532.86195921.58--94.41

Table 5

Training time and number of rounds required to achieve corresponding accuracy in CIFAR10 dataset"

场景方 法准确率总体/%
50%55%60%65%
轮数时间/s轮数时间/s轮数时间/s轮数时间/s
IIDFedAvg27119.7643186.1979334.64152735.1669.39
本文2684.2440129.3273283.69147615.8570.12
Non-IIDFedAvg129830.141921 260.76----58.9
本文115604.83164910.112421 347.81--61.41

Fig.4

Effect of model accuracy under different data heterogeneity"

[1] Tankard C. What the GDPR means for businesses[J].Network Security, 2016,2016(6): 5-8.
[2] McMahan H B, Mcore E, Ramage D, et al. Communication-efficient learning of deep networks from decentralized data[C]∥Proceeding of the 20th Tnternational Conference on Artificial Intelligence and Statistics,Ft. Lauderdale,USA, 2017.
[3] Abadi M, Chu A, Goodfellow L, et al. Deep learning with differential privacy[C]∥Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security,Vienna, Austria, 2016:308-318.
[4] Bonawitz K, Ivanow V, Kreuter B, et al. Practical secure aggregation for privacy-preserving machine learning[C]∥Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security,Dallas, USA, 2017: 1175-1191.
[5] McMahan H B, Ramage D, Kunal T, et al. Learning differentially private recurrent language models[J/OL].[2022-11-22]. ,2017.
[6] Chai Z, Ali A, Zawad S, et al. Tifl: A tier-based federated learning system[C]∥Proceedings of the 29th International Symposium on High-Performance Parallel and Distributed Computing,Stockholm, Sweden, 2020: 125-136.
[7] Nishio T, Yonetani R. Client selection for federated learning with heterogeneous resources in mobile edge[C]∥IEEE International Conference on Communications(ICC), Piscataway, USA, 2019:1-7.
[8] Karimireddy S P, Kale S, Mohri M, et al. SCAFFOLD: Stochastic controlled averaging for federated learning[C]∥International Conference on Machine Learning,Online, 2020:5132-5143.
[9] Fraboni Y, Vidal R, Kamenl L, et al. Clustered sampling: Low-variance and improved representativity for clients selection in federated learning[C]∥Preceeding of the 38th International Conference on Machine Learning, Online, 2021: 3407-3416.
[10] Li T, Sahu A K, Zaheer M, et al. Federated optimization in heterogeneous networks[J]. Proceedings of Machine Learning and Systems, 2020, 2: 429-450.
[11] Shu J G, Zhong W Z, Zhou Y, et al. FLAS: Computation and communication efficient federated learning via adaptive sampling[J]. IEEE Transactions on Network Science and Engineering,2021,9(4): 2003-2014.
[12] Bonawitz K, Eichner H, Grieskamp W, et al. Towards federated learning at scale: System design[J]. Proceedings of Machine Learning and Systems, 2019, 1: 374-388.
[13] Ward J H. Hierarchical grouping to optimize an objective function[J]. Journal of the American Statistical Association,1963, 58(301): 236-244.
[14] Sattler F, Müller K R, Samek W. Clustered federated learning: Model-agnostic distributed multitask optimization under privacy constraints[J]. IEEE Transactions on Neural Networks and Learning Systems, 2020, 32(8): 3710-3722.
[15] LeCun Y, Bottou L, Bengio Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
[16] Krizhevsk, Hinton G. Learning multiple layers of features from tiny images[R]. Toronto: University of Toronto, 2009.
[17] Hsu T M H, Qi H, Brown M. Measuring the effects of non-identical data distribution for federated visual classification[J/OL].[2022-11-22]. ,2019.
[1] Qiu-zhan ZHOU,Xin-meng LI,Hao-qing-zi SHEN,Hui-nan WU,Yuan-yuan LI,Jing RONG,Chun-hua HU,Ping-ping LIU. Non-intrusive load decomposition of unbalanced data based on attention mechanism [J]. Journal of Jilin University(Engineering and Technology Edition), 2026, 56(1): 239-246.
[2] Shu-xu ZHAO,Zhi-chao SUN,Xiao-long WANG. Dynamic authentication protocol for mobile edge computing scenarios [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(3): 1050-1060.
[3] Qi-liang ZHU,Xue-ting YU. k⁃prototype clustering algorithm and privacy protection for sensitive data republishing under relative entropy [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(3): 1009-1014.
[4] Shu-hong MA,Jun-jie ZHANG,Xi-fang CHEN,Guo-mei LIAO. Identifying urban functional structures using time-series taxi data [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(2): 603-613.
[5] Xin-gang GUO,Song WANG,Chao CHENG,Zhen FAN. Combined game theory and driving style hybrid traffic flow lane change decision model [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(12): 3875-3884.
[6] Song LI,Xiao-nan LIU,Juan LIU. Peak clustering algorithm for uncertain data density based on JS divergence [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(7): 2038-2048.
[7] Xi-jun ZHANG,Guang-jie YU,Yong CUI,Ji-yang SHANG. Short-term traffic flow prediction based on clustering algorithm and graph neural network [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(6): 1593-1600.
[8] Xi-guang ZHANG,Long-fei ZHANG,Yu-xi MA,Yin-ting FAN. Design of fuzzy clustering algorithm for massive cloud data based on density peak [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(5): 1401-1406.
[9] Di LIU,Yao SUN,Yun-feng HU,Hong CHEN. Commercial vehicle formation strategy based on density clustering [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(5): 1459-1468.
[10] Fu-heng QU,Yue-tao PAN,Yong YANG,Ya-ting HU,Jian-fei SONG,Cheng-yu WEI. An efficient global K-means clustering algorithm based on weighted space partitioning [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(5): 1393-1400.
[11] Gui-zhen CHEN,Hui-ting CHENG,Cai-hua ZHU,Yu-ran LI,Yan LI. A risk evaluation method for urban intersections considering drivers' physiological information [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(5): 1277-1284.
[12] Li LYU,Mei-zi ZHU,Ping KANG,Long-zhe HAN. Density peaks clustering with second-order K-nearest neighbors and multi-cluster merging [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(5): 1417-1425.
[13] Xiao-yu YI,Mian-zhu YI. Network book resource recommendation based on deep fusion of interest information [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(12): 3614-3619.
[14] De-lin LI,Jun-xian CHEN,Yong-gang WANG,Lu WANG,Zhao-qing SHEN. Identification of driving behavior on steep sharp curves based on latent class model [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(12): 3526-3533.
[15] Liang FAN,Ying-ming XU,Yang TAN. Interface slip calculation of prefabricated steel⁃concrete composite beams with clustering studs [J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(9): 2533-2541.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
Copyright © Journal of Jilin University(Engineering and Technology Edition), All Rights Reserved.
Tel: +86-431-85095297　Fax: +86-431-85094074　E-mail: xbgxb@jlu.edu.cn
Powered by Beijing Magtech Co. Ltd