App. Res. of Comp. Accepted Paper
Vol. 43, 2026 No. 6

Differential privacy-based personalized federated learning for heterogeneous data

Hou Junkang1,2
Guo Rui1,2
Zhang Yinghui1,2
Liu Guangjun3
1. School of Cyberspace Security, Xi'an University of Posts & Telecommunications, Xi'an 710121, China
2. National Engineering Research Center for Secured Wireless, Xi'an University of Posts and Telecommunications, Xi'an 710121, China
3. School of Information Engineering, Xi'an University, Xi'an 710065, China
DOI: 10.19734/j.issn.1001-3695.2025.09.0411

Abstract

This study addresses the coexistence of performance degradation and privacy leakage risks caused by data heterogeneity in federated learning, and aims to design a personalized federated learning scheme that both protects privacy and adapts to heterogeneous data. It proposes a differential privacy-based personalized federated learning framework (DP-PFL) , which mainly includes three components: 1) An adaptive differential privacy gradient perturbation mechanism that dynamically attenuated noise intensity according to training progress and allocated noise differentially based on the importance of gradient components; 2) A meta-learning-based personalized training strategy that achieved rapid local adaptation by optimizing global initialization parameters and adaptive model mixing; 3) A multi-level privacy defense system that combined Bayesian uncertainty enhancement, temperature scaling, and output noise injection. Theoretical analysis confirms that DP-PFL satisfies the (ε, δ) -differential privacy constraint and maintains an O(1/√T) convergence rate. Experiments on MNIST and CIFAR-10 datasets show that DP-PFL achieved accuracy rates of 85.5% and 74.2% respectively in the highly heterogeneous scenario with α=0.1; its defense success rates against membership inference, attribute inference, and gradient inversion attacks reached 54.6%, 51.5%, and PSNR=6.8dB respectively, representing an average improvement of 10.2% over baselines. DP-PFL realizes effective personalized learning for heterogeneous data under strict privacy protection constraints and provides theoretical support and practical guidance for the secure deployment of federated learning.

Foundation Support

国家密码科学基金资助项目(2025NCSF02037)
国家自然科学基金资助项目(62072369)
陕西省重点研发计划基金资助项目(2020ZDLGY08-04)
陕西省创新能力支持计划基金资助项目(2020KJXX-052)
陕西省自然科学基金一般项目(2024JC-YBMS-545,2024JC-YBMS-557)
陕西省高校青年创新团队(23JP160)
西安市科技计划项目(23KGDW0018-2023)

Publish Information

DOI: 10.19734/j.issn.1001-3695.2025.09.0411
Publish at: Application Research of Computers Accepted Paper, Vol. 43, 2026 No. 6

Publish History

[2026-02-25] Accepted Paper

Cite This Article

侯俊康, 郭瑞, 张应辉, 等. 异构数据的差分隐私个性化联邦学习方案 [J]. 计算机应用研究, 2026, 43 (6). (2026-02-25). https://doi.org/10.19734/j.issn.1001-3695.2025.09.0411. (Hou Junkang, Guo Rui, Zhang Yinghui, et al. Differential privacy-based personalized federated learning for heterogeneous data [J]. Application Research of Computers, 2026, 43 (6). (2026-02-25). https://doi.org/10.19734/j.issn.1001-3695.2025.09.0411. )

About the Journal

  • Application Research of Computers Monthly Journal
  • Journal ID ISSN 1001-3695
    CN  51-1196/TP

Application Research of Computers, founded in 1984, is an academic journal of computing technology sponsored by Sichuan Institute of Computer Sciences under the Science and Technology Department of Sichuan Province.

Aiming at the urgently needed cutting-edge technology in this discipline, Application Research of Computers reflects the mainstream technology, hot technology and the latest development trend of computer application research at home and abroad in a timely manner. The main contents of the journal include high-level academic papers in this discipline, the latest scientific research results and major application results. The contents of the columns involve new theories of computer discipline, basic computer theory, algorithm theory research, algorithm design and analysis, blockchain technology, system software and software engineering technology, pattern recognition and artificial intelligence, architecture, advanced computing, parallel processing, database technology, computer network and communication technology, information security technology, computer image graphics and its latest hot application technology.

Application Research of Computers has many high-level readers and authors, and its readers are mainly senior and middle-level researchers and engineers engaged in the field of computer science, as well as teachers and students majoring in computer science and related majors in colleges and universities. Over the years, the total citation frequency and Web download rate of Application Research of Computers have been ranked among the top of similar academic journals in this discipline, and the academic papers published are highly popular among the readers for their novelty, academics, foresight, orientation and practicality.

Indexed & Evaluation

  • The Second National Periodical Award 100 Key Journals
  • Double Effect Journal of China Journal Formation
  • the Core Journal of China (Peking University 2023 Edition)
  • the Core Journal for Science
  • Chinese Science Citation Database (CSCD) Source Journals
  • RCCSE Chinese Core Academic Journals
  • Journal of China Computer Federation
  • 2020-2022 The World Journal Clout Index (WJCI) Report of Scientific and Technological Periodicals
  • Full-text Source Journal of China Science and Technology Periodicals Database
  • Source Journal of China Academic Journals Comprehensive Evaluation Database
  • Source Journals of China Academic Journals (CD-ROM Version), China Journal Network
  • 2017-2019 China Outstanding Academic Journals with International Influence (Natural Science and Engineering Technology)
  • Source Journal of Top Academic Papers (F5000) Program of China's Excellent Science and Technology Journals
  • Source Journal of China Engineering Technology Electronic Information Network and Electronic Technology Literature Database
  • Source Journal of British Science Digest (INSPEC)
  • Japan Science and Technology Agency (JST) Source Journal
  • Russian Journal of Abstracts (AJ, VINITI) Source Journals
  • Full-text Journal of EBSCO, USA
  • Cambridge Scientific Abstracts (Natural Sciences) (CSA(NS)) core journals
  • Poland Copernicus Index (IC)
  • Ulrichsweb (USA)