Research Article
Byzantine-Robust Federated Learning for Unreliable Environments by Random-Matching Verification and Credibility Table
@INPROCEEDINGS{10.1007/978-3-031-94445-1_1, author={Yanjun Li and Xiaomeng Li and Hui Wan and Cong Wang}, title={Byzantine-Robust Federated Learning for Unreliable Environments by Random-Matching Verification and Credibility Table}, proceedings={Security and Privacy in Communication Networks. 20th EAI International Conference, SecureComm 2024, Dubai, United Arab Emirates, October 28--30, 2024, Proceedings, Part I}, proceedings_a={SECURECOMM}, year={2026}, month={9}, keywords={Federated Learning Byzantine Robustness Parameter Verification Credibility}, doi={10.1007/978-3-031-94445-1_1} }- Yanjun Li
Xiaomeng Li
Hui Wan
Cong Wang
Year: 2026
Byzantine-Robust Federated Learning for Unreliable Environments by Random-Matching Verification and Credibility Table
SECURECOMM
Springer
DOI: 10.1007/978-3-031-94445-1_1
Abstract
Federated learning (FL) is a technique that establishes a global model based on distributed data across multiple clients, while preserving data privacy. However, the frequent client-server communication in FL provides an opportunity for Byzantine clients, who can tamper with or falsify model parameters. Consequently, the presence of Byzantine clients impacts the convergence of the global model. To address this issue, we propose a framework that maintains Byzantine-robustness for FL in unreliable environments (BRFLUE). Specifically, BRFLUE consists of two components: a random-matching verification module and a credibility-table-based aggregation module. The former randomly matches clients and allows them to verify each other’s behavior anonymously. The latter updates clients’ credibility after receiving the client verification results and aggregates the global model based on the credibility table. The combination of these two modules can effectively identify possible Byzantine clients and reduce their adverse impact on the global model. Through implementations in different scenarios with classical datasets, we demonstrate the effectiveness of the proposed BRFLUE.