Abstract
The lack of security mechanism of Controller Area Network (CAN) bus protocol introduces structural vulnerabilities, requiring intrusion detection agents for message injection attacks. This study presents an intrusion detection system based on Recurrent Neural Networks (RNNs), capturing time-series patterns. We compare the performance of various RNNs and analyze how each architecture influences its detection capability.
| Translated title of the contribution | A Study on Recurrent Neural Network Performance for In-Vehicle Intrusion Detection Agents |
|---|---|
| Original language | Korean |
| Pages (from-to) | 1842-1845 |
| Number of pages | 4 |
| Journal | Journal of Korean Institute of Communications and Information Sciences |
| Volume | 50 |
| Issue number | 12 |
| DOIs | |
| State | Published - Dec 2025 |
Bibliographical note
Publisher Copyright:© 2025, Journal of Korean Institute of Communications and Information Sciences. All rights reserved.
Keywords
- CAN
- Intrusion detection agent
- Multi-class classification
- Recurrent neural network
Fingerprint
Dive into the research topics of 'A Study on Recurrent Neural Network Performance for In-Vehicle Intrusion Detection Agents'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver