Hybid Clustering-Transformer IDS for Rural CAVs

Document Type

Conference Paper

Publication Date

2025

DOI

10.1145/3769700.3771699

Publication Title

CoNEXT-SW '25: Proceedings of the CoNEXT'25 Student Workshop

Pages

7-8

Conference Name

CoNEXT '25: The 21st International Conference on Emerging Networking EXperiments and Technologies, December 1-4, 2025, Hong Kong, Hong Kong

Abstract

Rural CAVs face intermittent connectivity and sparse infrastructure, making cloud-based IDS unreliable. We propose an edge-first Hybrid Clustering–Transformer IDS that combines per–CAN-ID unsupervised clustering with a lightweight transformer for byte-sequence reasoning, fused via confidence weighting to handle per-ID variability. On real CAN datasets, it attains 100% recall and 98.9% accuracy, with false positives concentrated on one CAN ID—enabling targeted calibration. The design provides low-latency, connectivity-independent protection for rural deployments.

Rights

© 2025 Copyright held by the owner/authors. All rights reserved.

"ACM treats links as citations (references to objects) rather than as incorporations (embedding of objects). Permission is not needed to create links to citations in The ACM Digital Library or Online Guide to Computing Literature. ACM encourages the widespread distribution of links to the definitive Version of Records of its copyrighted works in the ACM Digital Library and does not require that authors obtain prior permission to include such links in their new works."

Original Publication Citation

Tavasoli, M., Sarrafzadeh, A., Karimoddini, A., Khaleghi, M., & Pasandi, H. B. (2025). Hybrid clustering-transformer IDS for rural CAVs. In CoNEXT-SW '25: Proceedings of the CoNEXT'25 Student workshop (pp. 7-8). Association for Computing Machinery. https://doi.org/10.1145/3769700.3771699

Link to Full Text

Share

COinS