EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

G-PFL-ID: Graph-Driven Personalized Federated Learning for Unsupervised Intrusion Detection in Non-IID IoT Systems

Daniel Ayo Oladele, Ayokunle Ige, Olatunbosun Agbo-Ajala, Olufisayo Ekundayo, Sree Ganesh Thottempudi, Malusi Sibiya, Ernest Mnkandla

IoT · 29.01.2026 · 7 (1) · 13

DOI: 10.3390/iot7010013 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IoT, online erschienen am 29.01.2026. Verfasst von Daniel Ayo Oladele und 6 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 100 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Intrusion detection in IoT networks is challenged by data heterogeneity, label scarcity, and privacy constraints. Traditional federated learning (FL) methods often assume IID data or require supervised labels, limiting their practicality. We propose G-PFL-ID, a graph-driven personalized federated learning framework for unsupervised intrusion detection in non-IID IoT systems. Our method trains a global graph encoder (GCN or GAE) with a DeepSVDD objective under a federated regularizer (FedReg) that combines proximal and variance penalties, then personalizes local models via a lightweight fine-tuning head. We evaluate G-PFL-ID on the IoT-23 (Mirai-based captures) and N-BaIoT (device-level dataset) under realistic heterogeneity (Dirichlet-based partitioning with concentration parameters α∈{0.1,0.5,∞} and client counts K∈{10,15,20} for IoT-23, and natural device-based partitioning for N-BaIoT). G-PFL-ID outperforms global FL baselines and recent graph-based federated anomaly detectors, achieving up to 99.46% AUROC on IoT-23 and 97.74% AUROC on N-BaIoT. Ablation studies confirm that the proximal and variance penalties reduce inter-round drift and representation collapse, and that lightweight personalization recovers local sensitivity—especially for clients with limited data. Our work bridges graph-based anomaly detection with personalized FL for scalable, privacy-preserving IoT security.

Zitieren

APA

Oladele, D. A., Ige, A., Agbo-Ajala, O., Ekundayo, O., Thottempudi, S. G., Sibiya, M., & Mnkandla, E. (2026). G-PFL-ID: Graph-Driven Personalized Federated Learning for Unsupervised Intrusion Detection in Non-IID IoT Systems. IoT, 7(1), 13. https://doi.org/10.3390/iot7010013

Vancouver

Oladele DA, Ige A, Agbo-Ajala O, Ekundayo O, Thottempudi SG, Sibiya M, et al. G-PFL-ID: Graph-Driven Personalized Federated Learning for Unsupervised Intrusion Detection in Non-IID IoT Systems. IoT. 2026;7(1):13. doi:10.3390/iot7010013

BibTeX
@article{Oladele2026,
  author = {Oladele, Daniel Ayo and Ige, Ayokunle and Agbo-Ajala, Olatunbosun and Ekundayo, Olufisayo and Thottempudi, Sree Ganesh and Sibiya, Malusi and Mnkandla, Ernest},
  title = {G-PFL-ID: Graph-Driven Personalized Federated Learning for Unsupervised Intrusion Detection in Non-IID IoT Systems},
  journal = {IoT},
  year = {2026},
  volume = {7},
  number = {1},
  pages = {13},
  doi = {10.3390/iot7010013}
}
Kostenlos starten →