EUVIMEDEUVIMED entdecken
Crossref

Optimal Client Sampling in Federated Learning With Client-Level Heterogeneous Differential Privacy

Jiahao Xu, Rui Hu, Olivera Kotevska

IEEE Internet of Things Journal · 01.05.2026 · 13 (9) · 18979-18990

DOI: 10.1109/jiot.2026.3661913 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Internet of Things Journal, erschienen am 01.05.2026. Verfasst von Jiahao Xu und 2 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 65 Literaturangaben. Gefördert von: U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research, UT-Battelle, LLC, with U.S. Department of Energy, National Science Foundation under the Harnessing the Data Revolution for Nevada Fire Science (HDRFS) Seed.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Federated learning (FL) with client-level differential privacy (DP) provides a promising framework for collaboratively training models while rigorously protecting clients’ privacy. However, classic approaches like DP-FedAvg struggle when clients have heterogeneous privacy requirements, as they must uniformly enforce the strictest privacy level across all clients, leading to excessive DP noise and significant degradation in model utility. Existing methods to improve the model utility in such heterogeneous privacy settings often assume a trusted server and are largely heuristic, resulting in suboptimal performance and lacking strong theoretical foundations. In this work, we address these challenges under a practical attack model where both clients and the server are honest-but-curious. We propose group-based differentially private federated learning (GDPFed), which partitions clients into groups based on their privacy budgets and achieves client-level DP within each group to reduce the privacy budget waste and hence improve the model utility. Based on the privacy and convergence analysis of GDPFed, we find that the magnitude of DP noise depends on both model dimensionality and the per-group client sampling ratios. To further improve the performance of GDPFed, we introduce GDPFed+, which integrates model sparsification to eliminate unnecessary noise and optimizes per-group client sampling ratios to minimize convergence error. Extensive empirical evaluations on multiple benchmark datasets demonstrate the effectiveness of GDPFed+, showing substantial performance gains compared with state-of-the-art methods.

Förderung

  • U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research
  • UT-Battelle, LLC, with U.S. Department of Energy
  • National Science Foundation under the Harnessing the Data Revolution for Nevada Fire Science (HDRFS) Seed

Zitieren

APA

Xu, J., Hu, R., & Kotevska, O. (2026). Optimal Client Sampling in Federated Learning With Client-Level Heterogeneous Differential Privacy. IEEE Internet of Things Journal, 13(9), 18979-18990. https://doi.org/10.1109/jiot.2026.3661913

Vancouver

Xu J, Hu R, Kotevska O. Optimal Client Sampling in Federated Learning With Client-Level Heterogeneous Differential Privacy. IEEE Internet of Things Journal. 2026;13(9):18979-18990. doi:10.1109/jiot.2026.3661913

BibTeX
@article{Xu2026,
  author = {Xu, Jiahao and Hu, Rui and Kotevska, Olivera},
  title = {Optimal Client Sampling in Federated Learning With Client-Level Heterogeneous Differential Privacy},
  journal = {IEEE Internet of Things Journal},
  year = {2026},
  volume = {13},
  number = {9},
  pages = {18979--18990},
  doi = {10.1109/jiot.2026.3661913}
}
Kostenlos starten →