EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

SD‐QFL‐Auth: A Software‐Defined Post‐Quantum Authentication Framework With Federated Learning for Secure Vehicular Networks

Qais El‐Dalahmeh, Adi El‐Dalahmeh, Jie Li

IET Networks · 2026-01 · 15 (1)

DOI: 10.1049/ntw2.70040 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IET Networks, online erschienen am 11.11.2026. Verfasst von Qais El‐Dalahmeh und 2 weiteren Autor:innen. Die Arbeit nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increasing connectivity of vehicular networks introduces significant authentication, privacy and security challenges, particularly in the presence of high mobility, malicious vehicles and emerging quantum‐computing threats. This paper proposes SD‐QFL‐Auth, a software‐defined vehicular security framework that integrates lattice‐based post‐quantum authentication, dynamic pseudonym mechanisms, trust‐aware software‐defined networking (SDN) and Federated Learning (FL)‐assisted anomaly detection. The SDN controller coordinates authentication decisions, trust evaluation, forwarding policies and malicious‐node isolation, while FL enables participating vehicles to contribute to abnormal behaviour detection without centrally collecting their raw local observations. The framework is evaluated using an integrated OMNeT++/SUMO environment under varying vehicular densities. At 300 vehicles, SD‐QFL‐Auth achieves approximately 96% trust evaluation accuracy and 97% attack detection and containment rate, while maintaining an average response time of approximately 136 ms, an average message delay of 72 ms and a packet loss ratio of 4.8%. In the extended scalability evaluation, authentication latency increases from approximately 13 ms at 50 vehicles to 56 ms at 1000 vehicles. Computational and storage analyses further show that the FL component contains 833 trainable parameters, requiring approximately 3.25 KiB for the model representation. These results demonstrate the performance characteristics of SD‐QFL‐Auth across authentication, attack detection, communication efficiency and increasing vehicular density, while providing an integrated framework for post‐quantum authentication, SDN‐based security orchestration and federated anomaly detection in vehicular networks.

Zitieren

APA

El‐Dalahmeh, Q., El‐Dalahmeh, A., & Li, J. (2026). SD‐QFL‐Auth: A Software‐Defined Post‐Quantum Authentication Framework With Federated Learning for Secure Vehicular Networks. IET Networks, 15(1). https://doi.org/10.1049/ntw2.70040

Vancouver

El‐Dalahmeh Q, El‐Dalahmeh A, Li J. SD‐QFL‐Auth: A Software‐Defined Post‐Quantum Authentication Framework With Federated Learning for Secure Vehicular Networks. IET Networks. 2026;15(1). doi:10.1049/ntw2.70040

BibTeX
@article{ElDalahmeh2026,
  author = {El‐Dalahmeh, Qais and El‐Dalahmeh, Adi and Li, Jie},
  title = {SD‐QFL‐Auth: A Software‐Defined Post‐Quantum Authentication Framework With Federated Learning for Secure Vehicular Networks},
  journal = {IET Networks},
  year = {2026},
  volume = {15},
  number = {1},
  doi = {10.1049/ntw2.70040}
}
Kostenlos starten →