Abstract

Philipp Meyer, Teresa Lübeck, Timo Häckel, Franz Korf, Thomas C. Schmidt,
Anomaly Detection in Real-Time Networks Using Asynchronous Traffic Shaping,
In: Proc. of the 16th IEEE Vehicular Networking Conference (VNC), IEEE, October 2025.
[pdf][BibTeX][Abstract]

Abstract: Connected vehicles are susceptible to safety-critical failures caused by malfunctioning components or intrusions, which calls for robust protection mechanisms of their backbone networks. The detection of misbehavior in the network is one key aspect of vehicular security. The IEEE Time-Sensitive Networking standard introduced Asynchronous Traffic Shaping (ATS) to guarantee deadlines for asynchronous communication in real-time networks. Extending previous work on monitoring synchronous real-time traffic, we propose an approach to detect network anomalies in real-time networks by surveying ATS. We show that the inherent configuration of ATS can serve as the definition of the normal behavior of individual streams. Deviations from this behavior can be detected by observing ATS statistics. Our findings reveal that ATS effectively identifies multiple classes of abnormal network behavior without producing false positives. Compared to previous work that relies on ingress metering algorithms, ATS proves to be stricter on asynchronous data streams, which results in higher detection rates while avoiding a separate ingress metering configuration.

Themes: Time-Sensitive Networking , Network Security

 

This page generated by bibTOhtml on Fri Jun 6 12:05:04 AM UTC 2025