ORIGINAL RESEARCH article
Front. Phys.
Sec. Social Physics
Volume 13 - 2025 | doi: 10.3389/fphy.2025.1387285
This article is part of the Research TopicSecurity, Governance, and Challenges of the New Generation of Cyber-Physical-Social SystemsView all 16 articles
Research on a Secure Federated Learning Methods Based on DWT and SVD Digital Watermarking
Provisionally accepted
- 1Jilin University, Changchun, China
- 2Hubei University of Arts and Science, Xiangyang, Hubei, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Due to the ease with which data can be replicated and cannot be tracked once it is spread, it becomes very difficult to verify the ownership of the data. Federated learning in privacy computing allows data to participate in model training without leaving the local area, resulting in better models and protecting data and privacy security. However, federated learning is also vulnerable to attacks in the application process. This article uses digital watermarks in federated learning, uses singular value decomposition (SVD) to generate robust digital watermarks for traceability, and uses discrete wavelet transform (DWT) to generate fragile digital watermarks for attack detection (1-6). The experimental results demonstrate that this greatly improves the security performance of federated learning.
Keywords: federated learning1, digital watermarking2, DWT3, SVD4, IID5
Received: 17 Feb 2024; Accepted: 21 May 2025.
Copyright: © 2025 Luan, Wen and Hang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Quan Wen, Jilin University, Changchun, China
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.