{"title":"Clustered federated learning enhanced by DAG-based blockchain with adaptive tip selection algorithm","authors":"Xiaofeng Xue , Haokun Mao , Qiong Li , Xin Guan","doi":"10.1016/j.iot.2025.101573","DOIUrl":null,"url":null,"abstract":"<div><div>Federated learning (FL) enables machine learning on distributed data while preserving client privacy. However, FL faces challenges such as device heterogeneity, central server vulnerabilities, and non-independent and identically distributed data. To address these challenges, researchers proposed an asynchronous and decentralized clustered FL (CFL) using a directed acyclic graph (DAG)-based blockchain, called specializing DAG FL (SDAGFL). However, SDAGFL consumes high communication and storage resources, posing a substantial burden on devices with limited resources. To overcome these limitations, we propose a novel CFL framework called DAG-CFL. DAG-CFL consists of a server layer with multiple servers implementing DAG-based blockchain and a client layer. Within this framework, we propose an adaptive tip selection algorithm (ATSA) to select the most suitable tip nodes for model aggregation. The analysis indicates that DAG-CFL significantly reduces communication and storage resource consumption on the client side compared with SDAGFL. In addition, the convergence of DAG-CFL and the time and space complexity of ATSA are analyzed to show the effectiveness of DAG-CFL. We evaluate DAG-CFL and ATSA on cluster-wise MNIST and CIFAR-10 datasets. The results show that DAG-CFL achieves comparable performance to the best CFL baseline method while eliminating the need for a predefined number of clusters. Notably, DAG-CFL achieves an 8% increase in accuracy compared with SDAGFL. The experiment results also show the robustness of DAG-CLF in various data distribution shift scenarios and indicate that ATSA can effectively cluster clients with a modularity value of 0.66 for the MNIST dataset and 0.71 for the CIFAR-10 dataset.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101573"},"PeriodicalIF":6.0000,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Internet of Things","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2542660525000861","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
Federated learning (FL) enables machine learning on distributed data while preserving client privacy. However, FL faces challenges such as device heterogeneity, central server vulnerabilities, and non-independent and identically distributed data. To address these challenges, researchers proposed an asynchronous and decentralized clustered FL (CFL) using a directed acyclic graph (DAG)-based blockchain, called specializing DAG FL (SDAGFL). However, SDAGFL consumes high communication and storage resources, posing a substantial burden on devices with limited resources. To overcome these limitations, we propose a novel CFL framework called DAG-CFL. DAG-CFL consists of a server layer with multiple servers implementing DAG-based blockchain and a client layer. Within this framework, we propose an adaptive tip selection algorithm (ATSA) to select the most suitable tip nodes for model aggregation. The analysis indicates that DAG-CFL significantly reduces communication and storage resource consumption on the client side compared with SDAGFL. In addition, the convergence of DAG-CFL and the time and space complexity of ATSA are analyzed to show the effectiveness of DAG-CFL. We evaluate DAG-CFL and ATSA on cluster-wise MNIST and CIFAR-10 datasets. The results show that DAG-CFL achieves comparable performance to the best CFL baseline method while eliminating the need for a predefined number of clusters. Notably, DAG-CFL achieves an 8% increase in accuracy compared with SDAGFL. The experiment results also show the robustness of DAG-CLF in various data distribution shift scenarios and indicate that ATSA can effectively cluster clients with a modularity value of 0.66 for the MNIST dataset and 0.71 for the CIFAR-10 dataset.
期刊介绍:
Internet of Things; Engineering Cyber Physical Human Systems is a comprehensive journal encouraging cross collaboration between researchers, engineers and practitioners in the field of IoT & Cyber Physical Human Systems. The journal offers a unique platform to exchange scientific information on the entire breadth of technology, science, and societal applications of the IoT.
The journal will place a high priority on timely publication, and provide a home for high quality.
Furthermore, IOT is interested in publishing topical Special Issues on any aspect of IOT.