Deep Neural Networks Driven Network Security Risk Analysis in Autonomous Driving
IEEE Transactions on Vehicular Technology · 2026-08 · 75 (8) · 15663-15673
DOI: 10.1109/tvt.2026.3672820 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Vehicular Technology, 2026-08. Verfasst von Jun Li und 2 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben.
Abstract (Originalsprache)
The application of deep neural networks (DNNs) in the area of autonomous driving is becoming increasingly widespread. However, its network security risk also highlights, which seriously threatens the security of the auto drive system. Autonomous driving network detection faces two key challenges. The high dimensional traffic data reduces specific attack detection rates and the normal abnormal sample imbalance degrades performance. For alleviating the above problems, this paper proposes auxiliary classifier generative adversarial networks improved by convolutional block attention module (ACGAN-CBAM). The main idea is to remove redundant samples from the autonomous driving network dataset and generate minority class samples, making the processed dataset more balanced. The core of Convolutional Block Attention Module (CBAM) in ACGAN lies in the collaborative work of channel and spatial attention modules to focus on key features. It combines residual blocks to construct a stable training framework, effectively addressing the issues of vanishing gradients and mode collapse that are commonly encountered in traditional generative adversarial networks. The experimental results show that when analyzing the risk characteristics of autonomous driving network, ACGAN-CBAM can more efficiently extract and utilize effective information, significantly improve the accuracy of analysis.
Zitieren
APA
Li, J., Liu, D., & Kumar, N. (2026). Deep Neural Networks Driven Network Security Risk Analysis in Autonomous Driving. IEEE Transactions on Vehicular Technology, 75(8), 15663-15673. https://doi.org/10.1109/tvt.2026.3672820
Vancouver
Li J, Liu D, Kumar N. Deep Neural Networks Driven Network Security Risk Analysis in Autonomous Driving. IEEE Transactions on Vehicular Technology. 2026;75(8):15663-15673. doi:10.1109/tvt.2026.3672820
BibTeX
@article{Li2026,
author = {Li, Jun and Liu, Dong and Kumar, Neeraj},
title = {Deep Neural Networks Driven Network Security Risk Analysis in Autonomous Driving},
journal = {IEEE Transactions on Vehicular Technology},
year = {2026},
volume = {75},
number = {8},
pages = {15663--15673},
doi = {10.1109/tvt.2026.3672820}
}
EUVIMED