EUVIMEDEUVIMED entdecken
Crossref

Federated Learning With Enhanced Privacy via Model Splitting and Random Client Participation

Yiwei Li, Shuai Wang, Zhuojun Tian, Xiuhua Wang, Shijian Su

IEEE Transactions on Cognitive Communications and Networking · 2026 · 12 · 8225-8238

DOI: 10.1109/tccn.2026.3694029 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Cognitive Communications and Networking, 2026. Verfasst von Yiwei Li und 4 weiteren Autor:innen. Die Arbeit nennt 53 Literaturangaben. Gefördert von: Natural Science Foundation of Xiamen City, Key Science and Technology Project of Xiamen City, Fundamental Research Funds for the Central Universities.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Federated Learning (FL) often adopts differential privacy (DP) to protect client data, but the added noise required for privacy guarantees can substantially degrade model accuracy. To resolve this challenge, we propose model-splitting privacy-amplified federated learning (MS-PAFL), a novel framework that combines structural model splitting with statistical privacy amplification. In this framework, each client’s model is partitioned into a private submodel, retained locally, and a public submodel, released for global aggregation. The calibrated Gaussian noise is injected only into the public submodel, thereby confining its adverse impact while preserving the utility of the retained local model components. We further present a rigorous theoretical analysis that characterizes the joint privacy amplification achieved through random client participation and local data subsampling under this architecture. The analysis provides explicit bounds on both single-round and total privacy loss, demonstrating that MS-PAFL significantly reduces the noise necessary to satisfy a target privacy protection level. Extensive experiments validate our theoretical findings, showing that MS-PAFL consistently attains a superior privacy–utility trade-off compared with conventional DP-FL baselines and enables the training of highly accurate models under strong privacy guarantees.

Förderung

  • Natural Science Foundation of Xiamen City
  • Key Science and Technology Project of Xiamen City
  • Fundamental Research Funds for the Central Universities
  • National Natural Science Foundation of China

Zitieren

APA

Li, Y., Wang, S., Tian, Z., Wang, X., & Su, S. (2026). Federated Learning With Enhanced Privacy via Model Splitting and Random Client Participation. IEEE Transactions on Cognitive Communications and Networking, 12, 8225-8238. https://doi.org/10.1109/tccn.2026.3694029

Vancouver

Li Y, Wang S, Tian Z, Wang X, Su S. Federated Learning With Enhanced Privacy via Model Splitting and Random Client Participation. IEEE Transactions on Cognitive Communications and Networking. 2026;12:8225-8238. doi:10.1109/tccn.2026.3694029

BibTeX
@article{Li2026,
  author = {Li, Yiwei and Wang, Shuai and Tian, Zhuojun and Wang, Xiuhua and Su, Shijian},
  title = {Federated Learning With Enhanced Privacy via Model Splitting and Random Client Participation},
  journal = {IEEE Transactions on Cognitive Communications and Networking},
  year = {2026},
  volume = {12},
  pages = {8225--8238},
  doi = {10.1109/tccn.2026.3694029}
}
Kostenlos starten →