EUVIMEDEUVIMED entdecken
Crossref

Privacy Preservation in User Behavior Analysis for Mobile-Edge Computing

Song Deng, Jie Zhang, Longxiang Zhang

IEEE Transactions on Consumer Electronics · 2025-05 · 71 (2) · 3718-3730

DOI: 10.1109/tce.2024.3472608 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Consumer Electronics, 2025-05. Verfasst von Song Deng und 2 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 48 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In mobile edge computing, user behavior analysis can provide a more clear understanding of edge user behavior, which is crucial for providing more accurate recommendation services to mobile users. However, privacy leakage is inevitable in the process of edge user behavior data collection, storage and analysis. How to protect edge user privacy is very urgent for the development of mobile edge computing. Existing schemes dealing with these problems still suffer from 1) high computational overhead, 2) low model accuracy and poor convergence under privacy-preserving, and 3) poor model robustness when edge nodes fail. Therefore, we propose a novel privacy-preserving model in edge user behavioral analysis termed Safe-FL to protect the behavioral privacy of all edge participants by introducing secure data aggregation and encryption in the federated learning (FL) framework. To reduce the privacy leakage of data uploading in edge user behavior analysis, we first construct a federated learning-based edge user behavior analysis framework. Secondly, we utilize a hybrid encryption scheme to encrypt the local model hierarchically to protect the gradient privacy in edge user behavior analysis and reduce the encryption computation overhead. And finally, to improve the robustness of the model, we propose a decentralized secure gradient aggregation algorithm based on blockchain. Extensive experiments are conducted on two public datasets, whose results illustrate that our model has 1) strong privacy preservation and small computational overhead; 2) high accuracy and convergence speed; and 3) good robustness, compared to other models.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Deng, S., Zhang, J., & Zhang, L. (2025). Privacy Preservation in User Behavior Analysis for Mobile-Edge Computing. IEEE Transactions on Consumer Electronics, 71(2), 3718-3730. https://doi.org/10.1109/tce.2024.3472608

Vancouver

Deng S, Zhang J, Zhang L. Privacy Preservation in User Behavior Analysis for Mobile-Edge Computing. IEEE Transactions on Consumer Electronics. 2025;71(2):3718-3730. doi:10.1109/tce.2024.3472608

BibTeX
@article{Deng2025,
  author = {Deng, Song and Zhang, Jie and Zhang, Longxiang},
  title = {Privacy Preservation in User Behavior Analysis for Mobile-Edge Computing},
  journal = {IEEE Transactions on Consumer Electronics},
  year = {2025},
  volume = {71},
  number = {2},
  pages = {3718--3730},
  doi = {10.1109/tce.2024.3472608}
}
Kostenlos starten →