An Effective Ensemble Learning-Based Real-Time Intrusion Detection Scheme for an In-Vehicle Network

The emergence of connected and autonomous vehicles has led to complex network architectures for electronic control unit (ECU) communication. The controller area network (CAN) enables the transmission of data inside vehicle networks. However, although it has low latency and enjoys data broadcast capability, it is vulnerable to attacks on security. The lack of effectiveness of conventional security mechanisms in addressing these vulnerabilities poses a danger to vehicle safety. This study presents an intrusion detection system (IDS) that accurately detects and classifies CAN bus attacks in real-time using ensemble techniques and the Kappa Architecture. The Kappa Architecture enables real-time attack detection, while ensemble learning combines multiple machine learning classifiers to enhance the accuracy of attack detection. The scheme utilizes ensemble methods with Kappa Architecture's real-time data analysis to detect common CAN bus attacks. This study entails the development and evaluation of supervised models, which are further enhanced using ensemble techniques. The accuracy, precision, recall, and F1 score are used to measure the scheme's effectiveness. The stacking ensemble technique outperformed individual supervised models and other ensembles with accuracy, precision, recall, and F1 of 0.985, 0.987, and 0.985, respectively.