Intrusion Detection System for CAN Networks Using PCA with Random Forest Approach

The rapid integration of advanced electronic control units (oecus) in modern cars has made the Controller Area Network (CAN) bus a crucial communications backbone. bone. However, CAN protocols are susceptible to a number of cyberattacks, including spoofing, fuzzing, and denial-of-service (DoS) attacks, due to their lack of built-in security measures. This paper proposes an efficient Intrusion Detection System (IDS) for CAN. networks constructed by combining Principal Component Analysis (PCA) and Random Forest (RF). In order to efficiently extract the most important features from high-dimensional CAN traffic data, PCA is used for dimensionality reduction and noise elimination. These condensed features are then used to train the Random Forest classifier, which correctly detects and categorises unusual network behaviours. Experiments on benchmark CAN intrusion datasets show that the proposed PCA-RF model achieves low false positive rates and excellent detection accuracy while maintaining computational efficiency. suitable for real-time in-car deployment. The findings demonstrate that combining PCA and Random Forest improves detection robustness and performance, providing a dependable and expandable IDS framework to protect automotive CAN networks from new online threats.