能耗约束下分层联邦学习模型质量优化的节点协作机制

doi:10.11772/j.issn.1001-9081.2024050704

《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (5): 1589-1594.DOI: 10.11772/j.issn.1001-9081.2024050704

• 先进计算 • 上一篇

能耗约束下分层联邦学习模型质量优化的节点协作机制

范亚州¹^,², 李卓¹^,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:2024-05-30 修回日期:2024-09-25 接受日期:2024-09-26 发布日期:2024-10-09 出版日期:2025-05-10
通讯作者: 李卓
作者简介:范亚州（1997—），男，安徽阜阳人，硕士研究生，主要研究方向：边缘计算
李卓（1983—），男，河南南阳人，教授，博士，CCF会员，主要研究方向：移动无线网络、分布式计算。
基金资助:
国家重点研发计划项目(2022YFF0604502);北京市自然科学基金资助项目(4232024)

Node collaboration mechanism for quality optimization of hierarchical federated learning models under energy consumption constraints

Yazhou FAN¹^,², Zhuo LI¹^,²()

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

Received:2024-05-30 Revised:2024-09-25 Accepted:2024-09-26 Online:2024-10-09 Published:2025-05-10
Contact: Zhuo LI
About author:FAN Yazhou， born in 1997， M. S. candidate. His research interests include edge computing.
LI Zhuo， born in 1983， Ph. D.， professor. His research interests include mobile wireless network， distributed computing.

摘要/Abstract

摘要：

边缘生成的大量数据可以通过联邦学习（FL）的方式训练全局模型，因此边缘计算与联邦学习的结合已成为降低网络能耗的关键技术。在分层联邦学习（HFL）中，边缘设备的局部数据量和数据质量的差异会直接影响HFL全局模型的质量。为此，提出一种传输能量约束下的节点协作算法（NCATTECC）解决传输能耗限制下全局模型质量最优化问题，证明了该问题是一个NP-hard（Non-deterministic Polynomial-hard）问题，同时证明了所提算法具有（1-1/ $e$ ）的近似比例。具体而言，通过协作，可以在不超过传输能耗限制的情况下，让更多的优质节点参与训练。在广泛使用的CIFAR-10、FashionMNIST数据集上的仿真实验结果表明，所提算法对所选节点进行训练，得到的模型准确率比FedAvg（Federated Averaging）和Fed-CBS（Federated Class-balanced Sampling）分别高出4.47%、6.64%和3.47%、4.58%。

关键词: 分层联邦学习, 端到端通信, 节点协作, 模型质量优化, 能耗限制

Abstract:

The massive data generated at the edge can be used to train global models through Federated Learning （FL）， making the combination of edge computing and federation learning become a key technology for reducing network energy consumption. In Hierarchical Federated Learning （HFL）， the difference in the amount of local data and data quality of edge devices directly affects the quality of the global model of HFL. To address these issues， a Node Cooperation Algorithm under Transmission Energy Consumption Constraint （NCATTECC） was proposed to solve the global model quality optimization problem， which was proved to be an Non-deterministic Polynomial-hard （NP-hard） problem， and it was also proved that the proposed algorithm has an approximate ratio of （1-1/ $e$ ）. Specifically， node collaboration enabled the participation of more high-quality nodes in training without exceeding energy consumption limits. Simulation experimental results on the widely used CIFAR-10 and FashionMNIST datasets prove that the proposed algorithm achieves model accuracy improvements of 4.47% and 6.64% compared to FedAvg （Federated Averaging）， and 3.47% and 4.58% compared to Fed-CBS （Federated Class-balanced Sampling）， respectively， when training with selected nodes.

Key words: Hierarchical Federated Learning (HFL), Device-to-Device (D2D) communication, node cooperation, model quality optimization, energy consumption limit

中图分类号:

TP393

范亚州, 李卓. 能耗约束下分层联邦学习模型质量优化的节点协作机制[J]. 计算机应用, 2025, 45(5): 1589-1594.

Yazhou FAN, Zhuo LI. Node collaboration mechanism for quality optimization of hierarchical federated learning models under energy consumption constraints[J]. Journal of Computer Applications, 2025, 45(5): 1589-1594.

图/表 6

图1 HFL的框架

Fig. 1 Framework of HFL

表1 CNNFashioin的结构

Tab. 1 Structure of CNNFashion

网络层	核大小	步长	输出通道数
Conv1	5×5	1	10
Conv2	5×5	1	20
Drop	—	—	—
FC1	—	—	50
FC2	—	—	10

图2 总传输能限制为20 J时在CIFAR-10上选取节点的全局模型准确率随训练轮次的变化

Fig. 2 Variation of global model accuracy with training rounds for nodes selected by algorithms on CIFAR-10 at a total transmission energy limit of 20 J

图3 总传输能耗限制为10 J时在FashionMNIST上选取节点的全局模型准确率随训练轮次的变化

Fig. 3 Variation of global model accuracy with training rounds for nodes selected by algorithms on FashionMNIST at a total transmission energy limit of 10 J

表2 不同传输能耗下算法所选节点数对比

Tab. 2 Comparison of number of nodes selected by algorithms under different transmission energy consumption

能耗/J	NCATTECC	FedAvg
10	42	34
20	53	42

表3 传输能耗为10 J时算法所选节点质量对比

Tab. 3 Quality comparison of nodes selected by algorithms with transmission energy consumption of 10 J

算法	Level1	Level2	Level3	Level4	Level5	Level6
NCATTECC	7	7	6	6	7	9
FedAvg	8	5	4	6	5	6

参考文献 28

1	LIM W Y B， LUONG N C， HOANG D T， et al. Federated learning in mobile edge networks： a comprehensive survey［J］. IEEE Communications Surveys and Tutorials， 2020， 22（3）： 2031-2063.
2	WANG Z， XU H， LIU J， et al. Resource-efficient federated learning with hierarchical aggregation in edge computing［C］// Proceeding of the 2021 IEEE Conference on Computer Communications. Piscataway： IEEE， 2021： 1-10.
3	McMAHAN H B， MOORE E， RAMAGE D， et al. Communication-efficient learning of deep networks from decentralized data［C］// Proceeding of the 20th Artificial Intelligence and Statistics. New York： PMLR， 2017： 1273-1282.
4	MUHAMMAD K， WANG Q， O'REILLY-MORGAN D， et al. FedFast： going beyond average for faster training of federated recommender systems［C］// Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York： ACM， 2020： 1234-1242.
5	ELAYAN H， ALOQAILY M， GUIZANI M. Sustainability of healthcare data analysis IoT-based systems using deep federated learning［J］. IEEE Internet of Things Journal， 2022， 9（10）： 7338-7346.
6	ABDULRAHMAN S， TOUT H， MOURAD A， et al. FedMCCS： multicriteria client selection model for optimal IoT federated learning［J］. IEEE Internet of Things Journal， 2021， 8（6）： 4723-4735.
7	WANG L， WANG W， LI B. CMFL： mitigating communication overhead for federated learning［C］// Proceeding of the IEEE 39th International Conference on Distributed Computing Systems. Piscataway： IEEE， 2019： 954-964.
8	KANG J， XIONG Z， 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.
9	KRIZHEVSKY A， SUTSKEVER I， Hinton G E. ImageNet classification with deep convolutional neural networks［J］. Communications of the ACM， 2017， 60（6）： 84-90.
10	HINTON G， DENG L， YU D， et al. Deep neural networks for acoustic modeling in speech recognition： the shared views of four research groups［J］. IEEE Signal Processing Magazine， 2012， 29（6）： 82-97.
11	CHEN M， YANG Z， SAAD W， et al. A joint learning and communications framework for federated learning over wireless networks［J］. IEEE Transactions on Wireless Communications， 2021， 20（1）： 269-283.
12	DUAN M， LIU D， CHEN X， et al. Astraea： self-balancing federated learning for improving classification accuracy of mobile deep learning applications［C］// Proceeding of the IEEE 37th International Conference on Computer Design. Piscataway： IEEE， 2019： 246-254.
13	YANG W J， CHUNG P C. Significant weighted aggregation method for federated learning in non-iid environment［C］// Proceeding of the 6th International Symposium on Computer， Consumer and Control. Piscataway： IEEE， 2023： 330-333.
14	ABDULRAHMAN S， OULD-SLIMANE H， CHOWDHURY R， et al. Adaptive upgrade of client resources for improving the quality of federated learning model［J］. IEEE Internet of Things Journal， 2023， 10（5）： 4677-4687.
15	CUI Y， CAO K， CAO G， et al. Client scheduling and resource management for efficient training in heterogeneous IoT-edge federated learning［J］. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems， 2022， 41（8）： 2407-2420.
16	WANG F， LI B， LI B. Quality-oriented federated learning on the fly［J］. IEEE Network， 2022， 36（5）： 152-159.
17	DENG Y， LYU F， REN J， et al. SHARE： shaping data distribution at edge for communication-efficient hierarchical federated learning［C］// Proceeding of the IEEE 41st International Conference on Distributed Computing Systems. Piscataway： IEEE， 2021： 24-34.
18	LI Q， LI X， ZHOU L， et al. AdaFL： adaptive client selection and dynamic contribution evaluation for efficient federated learning［C］// Proceeding of the 2024 IEEE International Conference on Acoustics， Speech and Signal Processing. Piscataway： IEEE， 2024： 6645-6649.
19	WANG L， XU S， WANG X， et al. Addressing class imbalance in federated learning［C］// Proceedings of the 35th AAAI Conference on Artificial Intelligence. Palo Alto： AAAI Press， 2021： 10165-10173.
20	CHANG X， AHMED S M， KRISHNAMURTHY S V， et al. FLASH： Federated Learning Across Simultaneous Heterogeneities［EB/OL］. ［2024-07-30］. .
21	WANG H-P， STICH S， HE Y， et al. ProgFed： effective， communication， and computation efficient federated learning by progressive training［C］// Proceedings of the 39th International Conference on Machine Learning. New York： PMLR， 2022： 23034-23054.
22	ZHOU X， ZHAO J， HAN H， et al. Joint optimization of energy consumption and completion time in federated learning［C］// Proceedings of the IEEE 42nd International Conference on Distributed Computing Systems （ICDCS）. Piscataway： IEEE， 2022： 1005-1017.
23	WANG T， LIU Y， ZHENG X， et al. Edge-based communication optimization for distributed federated learning［J］. IEEE Transactions on Network Science and Engineering， 2022， 9（4）： 2015-2024.
24	CHEN H， HUANG S， ZHANG D， et al. Federated learning over wireless IoT networks with optimized communication and resources［J］. IEEE Internet of Things Journal， 2022， 9（17）： 16592-16605.
25	LI A， ZHANG L， TAN J， et al. Sample-level data selection for federated learning［C］// Proceeding of the 2021 IEEE Conference on Computer Communications. Piscataway： IEEE， 2021： 1-10.
26	HOCHBAUM D S， PATHRIA A. Analysis of the greedy approach in covering problems［J］. Naval Research Logistics， 1998， 45（6）： 615-627.
27	ZHANG J， LI A， TANG M， et al. Fed-CBS： a heterogeneity-aware client sampling mechanism for federated learning via class-imbalance reduction［C］// Proceedings of the 40th International Conference on Machine Learning. New York： PMLR， 2023： 41354-41381.
28	JIANG Y， WANG S， VALLS V， et al. Model pruning enables efficient federated learning on edge devices［J］. IEEE Transactions on Neural Networks and Learning Systems， 2023， 34（12）： 10374-10386.
	This work is partially supported byFoundation: National Key Research and Development Program of China （2022YFF0604502）； Beijing Natural Science Foundation （4232024）.

[1]	项钰斐, 倪郑威. 基于演化博弈的分层联邦学习边缘联合动态分析[J]. 《计算机应用》唯一官方网站, 2025, 45(4): 1077-1085.
[2]	赵晓焱, 匡燕, 王梦含, 袁培燕. 基于知识图谱的端到端内容共享机制[J]. 《计算机应用》唯一官方网站, 2024, 44(4): 995-1001.
[3]	耿方兴, 李卓, 陈昕. 基于多领导者Stackelberg博弈的分层联邦学习激励机制设计[J]. 《计算机应用》唯一官方网站, 2023, 43(11): 3551-3558.
[4]	冬欣松, 郑建超, 蔡跃明, 尹廷辉, 张潇毅. 端到端通信中基于时间转换能量采集的计算迁移方案[J]. 计算机应用, 2018, 38(12): 3535-3540.

能耗约束下分层联邦学习模型质量优化的节点协作机制

Node collaboration mechanism for quality optimization of hierarchical federated learning models under energy consumption constraints

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 6

参考文献 28

相关文章 4

编辑推荐

Metrics