Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning

Tao Liu, Yuhang Zhang, Zhu Feng, Zhiqin Yang, Chen Xu +2 more
4/26/2024
cs.CRcs.AIcs.CV

Official code: PhD-TaoLiu/FCBA (23 stars, last commit 2023-12-25).

Abstract

Backdoors on federated learning will be diluted by subsequent benign updates. This is reflected in the significant reduction of attack success rate as iterations increase, ultimately failing. We use a new metric to quantify the degree of this weakened backdoor effect, called attack persistence. Given that research to improve this performance has not been widely noted,we propose a Full Combination Backdoor Attack (FCBA) method. It aggregates more combined trigger information for a more complete backdoor pattern in the global model. Trained backdoored global model is more resilient to benign updates, leading to a higher attack success rate on the test set. We test on three datasets and evaluate with two models across various settings. FCBA's persistence outperforms SOTA federated learning backdoor attacks. On GTSRB, postattack 120 rounds, our attack success rate rose over 50% from baseline. The core code of our method is available at https://github.com/PhD-TaoLiu/FCBA.

View on arXivView PDF

Code Implementations(1)

PhD-TaoLiu/FCBAOfficial100%

The core code for our paper "Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning".

233PythonDec 18, 20232 years ago

Similar Papers

On the Vulnerability of Backdoor Defenses for Federated Learning

90% match1 repo412023

Learning to Backdoor Federated Learning

90% match1 repo82023

Act in Collusion: A Persistent Distributed Multi-Target Backdoor in Federated Learning

90% match2024

FedBAP: Backdoor Defense via Benign Adversarial Perturbation in Federated Learning

90% match2025

Watch Out for the Lifespan: Evaluating Backdoor Attacks Against Federated Model Adaptation

89% match2025

Bad-PFL: Exploring Backdoor Attacks against Personalized Federated Learning

89% match1 repo172025

Cite this paper

@article{liu2024beyond,
  title  = {Beyond Traditional Threats: A Persistent Backdoor Attack on Federated Learning},
  author = {Tao Liu and Yuhang Zhang and Zhu Feng and Zhiqin Yang and Chen Xu and Dapeng Man and Wu Yang},
  year   = {2024},
  eprint = {2404.17617},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2404.17617v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2404.17617.svg)](https://codeofpaper.com/paper/2404.17617)

Discussion