Abstract
With the rapid advancement of digitalization and automation, modern vehicles, especially in the light commercial segment, have evolved into complex, interconnected platforms resembling mobile computing systems. This transformation has increased the dependency on in-vehicle communication networks and, as a result, exposed them to a wider range of cybersecurity threats. A fundamental aspect of the proposed method is the use of a lightweight CNN model specific for deployment in embedded automotive environments with limited computational resources and optimized for efficiency. Operating on low-power hardware platforms such as edge ECUs, the tiny device developed in this study works effectively unlike conventional deep learning architectures seeking high processing power and memory. Despite its minimal computational footprint, the model is capable of accurately distinguishing between legitimate and spoofed communication traffic, as well as detecting a variety of attack forms that target different CAN protocol components. The performance metrics of the model further highlight its effectiveness, achieving a ROC AUC Score of 0.9887, an Accuracy of 0.9887, a Precision of 0.9825, a Recall of 0.9952, and an F1-Score of 0.9888. Particularly for real-time on-vehicle intrusion detection systems, this harmony between performance and efficiency makes the strategy especially important. Just as importantly is the introduction of a specifically produced hybrid dataset, which is fundamental for system evaluation and training. The dataset aggregates synthetic generated attack scenarios with real-world spoofing, injection, and denial-of- service (DoS) conditions using actual CAN traffic acquired from a J1939-compliant light commercial vehicle. Standard 11-bit identities combined with industrial communication protocols help the dataset to reflect real-world vehicle dynamics across several ECUs under various scenarios. The model can learn fine-grained patterns often missed by conventional rule-based or manually engineered approaches by means of the image-like transformation of CAN messages—preserving bit-level and temporal information. In intelligent transportation systems, the lightweight CNN architecture and the strong dataset combine to create a scalable and deployable IDS framework that can improve in-vehicle cybersecurity.
Keywords
CAN Bus Intrusion Detection System Cybersecurity Vehicular Communication Deep Learning CNN Artificial Intelligence Anomaly Detection Light Commercial Vehicles
Thanks
We would want to sincerely thank you to Karsan for their essential contribution in allowing the data collecting process. Particularly, we would want to thank Mr. M. Alper BALIM, E&E Vehicle Software Engineering Administrator, and Mr. Nurettin ÖZEKMEKCİ, Manager of E&E Design Engineering, whose priceless direction, insights, and ongoing support have greatly helped this research to be successful. We also want to thank the whole Karsan R&D team for their kind cooperation, test vehicle preparation, and guarantee of access to the required tools and systems for experimental use. Their dedication and transparency made it feasible to create an industrial-grade, realistic dataset and to do all stages of this work under reasonable, dependable surroundings.
References
- [1] Alfardus, A., & Rawat, D. B. (2021, 1–4 Aralık). Intrusion detection system for CAN bus in-vehicle network based on machine learning algorithms.In Proceedings of the 2021 IEEE 12th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON) (pp. 944–949).IEEE. https://doi.org/10.1109/UEMCON53757.2021.9666745
- [2]Avatefipour, O., Al-Sumaiti, A. S., El-Sherbeeny, A. M., Awwad, E. M., Elmeligy, M. A., Mohamed, M. A., & Malik, H. (2019). An intelligent secured framework for cyber-attack detection in electric vehicles’ CAN bus using machine learning. IEEE Access,7,127,580-127 https://doi.org/10.1109/ACCESS.2019.2937576
- [3] Jichici, C., Groza, B., Ragobete, R., Murvay, P.-S., & Andreica, T. (2022). Effective intrusion detection and prevention for the commercial vehicle SAE J1939 CAN bus. IEEE Transactions on Intelligent Transportation Systems, 23(10), 17 425–17 439. https://doi.org/10.1109/TITS.2022.3151712
- [4] Marchetti, M., & Stabili, D. (2017, 11–14 Haziran). Anomaly detection of CAN bus messages through analysis of ID sequences [Conference paper]. In Proceedings of the 2017 IEEE Intelligent Vehicles Symposium(IV)(pp.1577–1583).IEEE. https://doi.org/10.1109/IVS.2017.7995934
- [5] Shahriar, M. H., Xiao, Y., Moriano, P., Lou, W., & Hou, Y. T. (2023). CANShield: Deep-learning-based intrusion detection framework for controller area networks at the signal level. IEEE Internet of Things Journal, 10(24), 22111–22127. https://doi.org/10.1109/JIOT.2023.3303271
- [6] Wu, S., Li, S., & Sun, W. (2023, 6–8 Aralık). ConvIDS: A convolutional LSTM-based intrusion detection model for in-vehicle CAN bus [Conference paper]. In Proceedings of the 2023 IEEE 6th International Conference on Automation, Electronics & Electrical Engineering (AUTEEE) (pp. 285–290). IEEE. https://doi.org/10.1109/AUTEEE60196.2023.10408040
- [7] LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278–2324.https://doi.org/10.1109/5.726791
- [8] Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet classification with deep convolutional neural networks. In F. Pereira, C. J. C. Burges, L. Bottou, & K. Q. Weinberger (Eds.), Advances in Neural Information Processing Systems 25 (pp. 1097–1105). Curran Associates. (https://papers.nips.cc/paper_files/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html)
- [9] Yamashita, R., Nishio, M., Do, R. K. G., & Togashi, K. (2018). Convolutional neural networks: An overview and application in radiology. Insights into Imaging, 9(4), 611–629. https://doi.org/10.1007/s13244-018-0639-9
- [10] Albawi, S., Mohammed, T. A., & Al-Zawi, S. (2017, 21–23 Ağustos). Understanding of a convolutional neural network [Conference paper]. In Proceedings of the 2017 International Conference on Engineering & Technology (ICET) (pp. 1–6). IEEE. https://doi.org/10.1109/ICEngTechnol.2017.8308186
- [11] Zhao, W., Yang, Y., Hu, H., Chen, Y., & Yu, F. (2025). A lightweight intrusion detection approach for CAN bus using depthwise separable convolutional Kolmogorov–Arnold network. Scientific Reports, 15, 17550. https://doi.org/10.1038/s41598-025-02474-1
- [12] Zhang, Y., Song, J., Sun, Y., Gao, Z., & Hu, Z. (2025). Federated two-stage transformer-based network for intrusion detection in non-IID data of controller area networks. Cybersecurity, 8(1), 29. https://doi.org/10.1186/s42400-024-00329-2
- [13] Wei, P., Wang, B., Dai, X., Li, L., & He, F. (2023). A novel intrusion detection model for the CAN bus packet of in-vehicle network based on attention mechanism and autoencoder. Digital Communications and Networks, 9, 14–21. https://doi.org/10.1016/j.dcan.2022.04.021
- [14] Shi, J. Z., Dong, L., Jiang, X., & Jin, X. (2024). IDS-DEC: A novel intrusion detection for CAN bus traffic based on deep embedded clustering. Vehicular Communications, 49, Article 100830. https://doi.org/10.1016/j.vehcom.2024.100830
- [15] Shi, D., Xu, M., Qin, Z., & Zhong, Y. (2025). Local-enhanced Toeplitz neural network for in-vehicle network intrusion detection. Journal of King Saud University – Computer and Information Sciences, 37, Article 8. https://doi.org/10.1007/s44443-025-00006-5
- [16] Gou, W., Zhang, H., & Zhang, R. (2023). Multi-classification and tree-based ensemble network for the intrusion detection system in the Internet of Vehicles. Sensors, 23(21), 8788. https://doi.org/10.3390/s23218788
- [17] Rai, R., Grover, J., Sharma, P., & Pareek, A. (2025). Securing the CAN bus using deep learning for intrusion detection in vehicles. Scientific Reports, 15, 13820. https://doi.org/10.1038/s41598-025-98433-x
- [18] Yang, H., & Effatparvar, M. (2025). A deep learning based intrusion detection system for CAN vehicle based on combination of triple attention mechanism and GGO algorithm. Scientific Reports, 15, 19462. https://doi.org/10.1038/s41598-025-04720-y
- [19] Devnath, M. K. (2023). GCNIDS: Graph convolutional network-based intrusion detection system for CAN bus. arXiv Preprint, arXiv:2309.10173. https://doi.org/10.48550/arXiv.2309.10173
- [20] Kabilan, N., Ravi, V., & Sowmya, V. (2024). Unsupervised intrusion detection system for in-vehicle communication networks. Journal of Safety Science and Resilience, 5, 119–129. https://doi.org/10.1016/j.jnlssr.2023.12.004
Details
| Primary Language | English |
|---|---|
| Subjects | Deep Learning |
| Journal Section | Research Articles |
| Authors | |
| Early Pub Date | September 30, 2025 |
| Publication Date | October 23, 2025 |
| Submission Date | June 11, 2025 |
| Acceptance Date | July 7, 2025 |
| Published in Issue | Year 2025 Volume: 9 Issue: 2 |
Cite
The works published in Journal of Innovative Science and Engineering (JISE) are licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.