Byzantine-robust federated learning via credibility assessment on non-IID data
Authors:
Open Access
- 1 January 2021
- journal article
- research article
- Published by American Institute of Mathematical Sciences (AIMS) in Mathematical Biosciences and Engineering
- Vol. 19 (2) , 1659-1676
Abstract
Federated learning is a novel framework that enables resource-constrained edge devices to jointly learn a model, which solves the problem of data protection and data islands. However, standard federated learning is vulnerable to Byzantine attacks, which will cause the global model to be manipulated by the attacker or fail to converge. On non-iid data, the current methods are not effective in defensing against Byzantine attacks. In this paper, we propose a Byzantine-robust framework for federated learning via credibility assessment on non-iid data (BRCA). Credibility assessment is designed to detect Byzantine attacks by combing adaptive anomaly detection model and data verification. Specially, an adaptive mechanism is incorporated into the anomaly detection model for the training and prediction of the model. Simultaneously, a unified update algorithm is given to guarantee that the global model has a consistent direction. On non-iid data, our experiments demonstrate that the BRCA is more robust to Byzantine attacks compared with conventional methods.Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- Federated Learning for Cybersecurity: Concepts, Challenges, and Future DirectionsIEEE Transactions on Industrial Informatics, 2021
- Byzantine-Resilient Secure Federated LearningIEEE Journal on Selected Areas in Communications, 2020
- Computational intelligence intrusion detection techniques in mobile cloud computing environments: Review, taxonomy, and open research issuesJournal of Information Security and Applications, 2020
- Federated Learning: Challenges, Methods, and Future DirectionsIEEE Signal Processing Magazine, 2020
- KeySplitWatermark: Zero Watermarking Algorithm for Software Protection Against Cyber-AttacksIEEE Access, 2020
- Byzantine-Resilient Stochastic Gradient Descent for Distributed Learning: A Lipschitz-Inspired Coordinate-wise Median Approach2019
- Security and Privacy for the Industrial Internet of Things: An Overview of Approaches to Safeguarding EndpointsIEEE Signal Processing Magazine, 2018
- The Internet of Things: Secure Distributed InferenceIEEE Signal Processing Magazine, 2018
- Distributed Statistical Machine Learning in Adversarial SettingsProceedings of the ACM on Measurement and Analysis of Computing Systems, 2017
- Anomaly Detection Using Autoencoders with Nonlinear Dimensionality Reduction2014