Deep Learning-Based Intrusion Detection for Automotive Ethernet: Evaluating & Optimizing Fast Inference Techniques for Deployment on Low-Cost Platform
This work addresses the challenge of real-time security for connected vehicles by enabling efficient deployment on affordable platforms, though it is incremental as it applies existing optimization methods to a specific domain.
The paper tackled the problem of deploying deep learning-based intrusion detection systems for automotive Ethernet on low-cost hardware by evaluating fast inference techniques like distillation and pruning, achieving detection times of up to 727 μs with an AUCROC of 0.9890 on a Raspberry Pi 4.
Modern vehicles are increasingly connected, and in this context, automotive Ethernet is one of the technologies that promise to provide the necessary infrastructure for intra-vehicle communication. However, these systems are subject to attacks that can compromise safety, including flow injection attacks. Deep Learning-based Intrusion Detection Systems (IDS) are often designed to combat this problem, but they require expensive hardware to run in real time. In this work, we propose to evaluate and apply fast neural network inference techniques like Distilling and Prunning for deploying IDS models on low-cost platforms in real time. The results show that these techniques can achieve intrusion detection times of up to 727 μs using a Raspberry Pi 4, with AUCROC values of 0.9890.