EUVIMEDEUVIMED entdecken
Crossref

6G Enabled Generative AI Services for Secure High-Speed Railway Networks

Ruibin Guo, Weiting Zhang, Dong Yang, Hongchao Wang, Hongke Zhang

IEEE Network · 2025-09 · 39 (5) · 105-113

DOI: 10.1109/mnet.2025.3580618 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Network, 2025-09. Verfasst von Ruibin Guo und 4 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 15 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Running trains at higher speeds is a crucial trend in railway development. For future ultra-high-speed scenarios, the sixth-generation mobile communication technology (6G) can provide reliable transmission and enable generative artificial intelligence (GAI) services for secure high-speed railway networks. In this paper, we propose a 6G-enabled GAI service architecture for railway networks to enhance train operation safety. This architecture leverages integrated air-ground-train transmission, uncrewed aerial vehicle (UAV), and distributed learning mechanisms to facilitate training and inference of GAI services. By analyzing the massive real-time data of critical equipment and areas with GAI techniques, early warnings and customized guidance for equipment breakdowns and natural disasters can be provided. Specifically, we propose integrating time-sensitive networking into 6G and jointly allocating time-spectrum resources in the air-ground-train integrated transmission to guarantee deterministic end-to-end latency in heterogeneous networks. Secondly, we develop a self-organization mechanism for a UAV cluster to perform various functions. The cluster comprises different types of UAVs and multiple intelligent algorithms are utilized during the self-organization process. Furthermore, a federated learning-based distributed training and inference framework for GAI services is studied, and several parallel computing techniques are proposed to reduce communication overhead. Simulation results demonstrate that learning-based resource allocation promotes the timeliness of model inference. Open research issues on further improving the performance of GAI services are also studied.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Guo, R., Zhang, W., Yang, D., Wang, H., & Zhang, H. (2025). 6G Enabled Generative AI Services for Secure High-Speed Railway Networks. IEEE Network, 39(5), 105-113. https://doi.org/10.1109/mnet.2025.3580618

Vancouver

Guo R, Zhang W, Yang D, Wang H, Zhang H. 6G Enabled Generative AI Services for Secure High-Speed Railway Networks. IEEE Network. 2025;39(5):105-113. doi:10.1109/mnet.2025.3580618

BibTeX
@article{Guo2025,
  author = {Guo, Ruibin and Zhang, Weiting and Yang, Dong and Wang, Hongchao and Zhang, Hongke},
  title = {6G Enabled Generative AI Services for Secure High-Speed Railway Networks},
  journal = {IEEE Network},
  year = {2025},
  volume = {39},
  number = {5},
  pages = {105--113},
  doi = {10.1109/mnet.2025.3580618}
}
Kostenlos starten →