EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

FedSW-TSAD: SWGAN-Based Federated Time Series Anomaly Detection

Xiuxian Zhang, Hongwei Zhao iD, Weishan Zhang iD, Shaohua Cao, Haoyun Sun, Baoyu Zhang iD

Sensors · 27.06.2025 · 25 (13) · 4014

DOI: 10.3390/s25134014 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 27.06.2025. Verfasst von Xiuxian Zhang und 5 weiteren Autor:innen. Beteiligte Einrichtung: Qingdao Institute of Software, College of Computer Science and Technology, China University of Petroleum (East China), Qingdao 266580, China. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Shandong Provincial Natural Science Foundation Major Basic Research Program, Shandong Data Open Innovative Application Laboratory.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

As distributed sensing technologies evolve, the collection of time series data is becoming increasingly decentralized, which introduces serious challenges for both model training and data privacy protection. In response to this trend, federated time series anomaly detection enables collaborative analysis across distributed sensing nodes without exposing raw data. However, federated anomaly detection experiences issues with unstable training and poor generalization due to client heterogeneity and the limited expressiveness of single-path detection methods. To address these challenges, this study proposes FedSW-TSAD, a federated time series anomaly detection method based on the Sobolev–Wasserstein GAN (SWGAN). It leverages the Sobolev–Wasserstein constraint to stabilize adversarial training and combines discriminative signals from both reconstruction and prediction modules, thereby improving robustness against diverse anomalies. In addition, FedSW-TSAD adopts a differential privacy mechanism with L2-norm-constrained noise injection, ensuring privacy in model updates under the federated setting. The experimental results determined using four real-world sensor datasets demonstrate that FedSW-TSAD outperforms existing methods by an average of 14.37% in the F1-score while also enhancing gradient privacy under the differential privacy mechanism. This highlights the practical value of FedSW-TSAD for privacy-preserving anomaly detection in sensor-based monitoring systems such as industrial IoT, remote diagnostics, and predictive maintenance.

Schlagwörter

Stichwörter der Autor:innen

anomaly detectionfederated learningprivacy protection

Autor:innen und Institutionen

  • Qingdao Institute of Software, College of Computer Science and Technology, China University of Petroleum (East China), Qingdao 266580, China.
  • Shandong Data Open Innovative Application Laboratory, Qingdao 266580, China.
  • Shandong Key Laboratory of Intelligent Oil & Gas Industrial Software, Qingdao 266580, China.

ORCID-Profile: Hongwei Zhao, Weishan Zhang, Baoyu Zhang

Förderung

  • National Natural Science Foundation of China
  • Shandong Provincial Natural Science Foundation Major Basic Research Program
  • Shandong Data Open Innovative Application Laboratory

Zitieren

APA

Zhang, X., Zhao, H., Zhang, W., Cao, S., Sun, H., & Zhang, B. (2025). FedSW-TSAD: SWGAN-Based Federated Time Series Anomaly Detection. Sensors, 25(13), 4014. https://doi.org/10.3390/s25134014

Vancouver

Zhang X, Zhao H, Zhang W, Cao S, Sun H, Zhang B. FedSW-TSAD: SWGAN-Based Federated Time Series Anomaly Detection. Sensors. 2025;25(13):4014. doi:10.3390/s25134014

BibTeX
@article{Zhang2025,
  author = {Zhang, Xiuxian and Zhao, Hongwei and Zhang, Weishan and Cao, Shaohua and Sun, Haoyun and Zhang, Baoyu},
  title = {FedSW-TSAD: SWGAN-Based Federated Time Series Anomaly Detection},
  journal = {Sensors},
  year = {2025},
  volume = {25},
  number = {13},
  pages = {4014},
  doi = {10.3390/s25134014}
}
Kostenlos starten →