• Home
  • Search
  • G 2 uardFL: Safeguarding Federated Learning Against Backdoor Attacks via Attributed Client Graph Clustering
  • Cite Icon4
  • https://doi.org/10.1109/tifs.2025.3639985Copy DOI Icon

G 2 uardFL: Safeguarding Federated Learning Against Backdoor Attacks via Attributed Client Graph Clustering

  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Federated Learning (FL) offers collaborative model training across multiple decentralized devices without the need to share data directly, enhancing privacy and data security. However, FL systems are susceptible to backdoor attacks, where malicious clients inject poisoned weights during training. Existing defenses, primarily based on anomaly detection, are prone to erroneous rejections of normal weights while accepting poisoned ones, largely due to shortcomings in quantifying similarities among client models. Furthermore, other defenses demonstrate effectiveness only when dealing with a limited number of malicious clients, typically fewer than 10%. To alleviate these vulnerabilities, we present G<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>uardFL, a protective framework that translates the detection of malicious clients into an attributed graph clustering problem, thus safeguarding FL systems. Specifically, this framework employs a client graph clustering approach to identify malicious clients and integrates an adaptive mechanism to amplify the discrepancy between the aggregated model and the poisoned ones, effectively eliminating embedded backdoors. Through empirical evaluation, comparing G<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>uardFL with cutting-edge defenses, such as FLAME (USENIX Security 2022) [37] and DeepSight (NDSS 2022) [43], against various backdoor attacks, including 3DFed (SP 2023) [26], our results demonstrate its significant effectiveness in mitigating backdoor attacks while having a negligible impact on the aggregated model’s performance on benign samples (i.e., the primary task performance). For instance, in an FL system with 25% malicious clients, G<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>uardFL reduces the attack success rate to 10.61%, while maintaining a primary task performance of 80.98% on the CIFAR-10 dataset. This surpasses the performance of the best-performing baseline, which merely achieves the attack success rate of 19.54%.

Similar Papers
  • Conference Article
  • Citations11

FedLesScan: Mitigating Stragglers in Serverless Federated Learning

  • Dec 17, 2022
  • Mohamed Elzohairy +6
  • Research Article
  • Citations1

FinBack: Infiltrating Backdoors into Gradient Compressors on Federated Learning

  • Jan 01, 2025
  • IEEE Transactions on Information Forensics and Security
  • Xiangyun Tang +7
  • Conference Article
  • Citations13

FLAC: Federated Learning with Autoencoder Compression and Convergence Guarantee

  • Dec 04, 2022
  • Mahdi Beitollahi +1
  • Book Chapter
  • Citations160

CONTRA: Defending Against Poisoning Attacks in Federated Learning

  • Jan 01, 2021
  • Sana Awan +2
  • PDF
  • Research Article
  • Citations50

Blockchain-Enabled: Multi-Layered Security Federated Learning Platform for Preserving Data Privacy

  • May 19, 2022
  • Electronics
  • Zeba Mahmood +1
  • Conference Article
  • Citations47

Federated Learning Attacks and Defenses: A Survey

  • Dec 17, 2022
  • Yao Chen +4
  • PDF
  • Research Article
  • Citations1

ARMOR: Differential Model Distribution for Adversarially Robust Federated Learning

  • Feb 07, 2023
  • Electronics
  • Yanting Zhang +5
  • Conference Article
  • Citations3

Towards Robust and Bias-free Federated Learning

  • May 08, 2023
  • Ousmane Touat +1
  • Conference Article
  • Citations7

Coalition Formation Game in the Cross-Silo Federated Learning System

  • Oct 01, 2022
  • Suhan Jiang +1
  • Conference Article
  • Citations57

FedLess: Secure and Scalable Federated Learning Using Serverless Computing

  • Dec 15, 2021
  • Andreas Grafberger +4
  • Research Article
  • Citations12

Towards Practical Backdoor Attacks on Federated Learning Systems

  • Nov 01, 2024
  • IEEE Transactions on Dependable and Secure Computing
  • Chenghui Shi +8
  • Research Article
  • Citations13

BapFL: You can Backdoor Personalized Federated Learning

  • Jun 19, 2024
  • ACM Transactions on Knowledge Discovery from Data
  • Tiandi Ye +4
  • Conference Article
  • Citations10

DSFL: Dynamic Sparsification for Federated Learning

  • Dec 27, 2022
  • Mahdi Beitollahi +2
  • Conference Article
  • Citations1

Personalized Federated Learning with Optimized Contrastive Learning for Intrusion Detection System

  • Dec 21, 2024
  • Quan Hong Ngo +2
  • Conference Article
  • Citations2

Maverick: Personalized Edge-Assisted Federated Learning with Contrastive Training

  • Apr 22, 2025
  • Kaibin Wang +7
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.