Advancements in Vehicle Intrusion Detection Systems: Enhancing Cybersecurity in Intelligent Transportation
Mahdi Sahlabadi, Md Rezanur Islam, Munkhdelgerekh Batzorig, Kangbin Yim · IntechOpen eBooks · 2025
Advancements in in-vehicle Intrusion Detection Systems (IDS) are critical to addressing cybersecurity challenges in modern intelligent transportation. This chapter explores vulnerabilities inherent in Controller Area Network (CAN) protocols, such as DoS, spoofing, and replay attacks, and their impacts on vehicular systems. By analyzing the LISA Vehicle Dataset collected through innovative techniques like gateway-based methods and custom testbeds as injection tools, the study provides insights into abnormal behaviors and attack patterns. It evaluates IDS approaches, including rule-based and machine learning-based systems, highlighting the strengths of deep learning models like RNNs and CNNs for robust anomaly detection. Feature selection and optimized chunking methodologies, such as mutual information analysis, are discussed to enhance IDS efficiency while reducing computational demands. The chapter also emphasizes real-world simulation using synchronized CAN datasets and testbeds, contributing to the development of scalable, real-time IDS frameworks for secure and reliable vehicular communication.