Towards Ultra-Low Latency: Binarized Neural Network Architectures for In-Vehicle Network Intrusion Detection
Huiyao Dong, Igor Kotenko
TL;DR
Problem: CAN buses lack authentication and encryption, enabling various attacks. Approach: a lightweight Binarized Neural Network (BNN) intrusion detector that fuses payload, ID, and timing features using a hybrid binary quantization for non-payload attributes. Contributions: (i) a BNN framework optimized for in-vehicle IDS, (ii) a 73-dimensional binary feature representation, (iii) empirical comparisons on CAN-IDS and Car-Hacking datasets showing competitive accuracy with lower latency and memory. Impact: enables real-time IDS deployment on micro-controllers and gateway ECUs, addressing practical security needs in modern vehicles.
Abstract
The Control Area Network (CAN) protocol is essential for in-vehicle communication, facilitating high-speed data exchange among Electronic Control Units (ECUs). However, its inherent design lacks robust security features, rendering vehicles susceptible to cyberattacks. While recent research has investigated machine learning and deep learning techniques to enhance network security, their practical applicability remains uncertain. This paper presents a lightweight intrusion detection technique based on Binarized Neural Networks (BNNs), which utilizes payload data, message IDs, and CAN message frequencies for effective intrusion detection. Additionally, we develop hybrid binary encoding techniques to integrate non-binary features, such as message IDs and frequencies. The proposed method, namely the BNN framework specifically optimized for in-vehicle intrusion detection combined with hybrid binary quantization techniques for non-payload attributes, demonstrates efficacy in both anomaly detection and multi-class network traffic classification. The system is well-suited for deployment on micro-controllers and Gateway ECUs, aligning with the real-time requirements of CAN bus safety applications.