KD-GAT: Combining Knowledge Distillation and Graph Attention Transformer for a Controller Area Network Intrusion Detection System

TL;DR

This paper presents KD-GAT, a novel intrusion detection system for CAN networks that combines graph attention networks with knowledge distillation to improve accuracy and reduce model size.

Contribution

It introduces a two-phase training method for a compact GAT-based student model using knowledge distillation from a multi-layer GAT teacher, applied to CAN intrusion detection.

Findings

Student model achieves over 99% accuracy on two datasets.

The approach effectively captures temporal and relational patterns in CAN traffic.

Class imbalance affects performance on the third dataset, indicating future work is needed.

Abstract

The Controller Area Network (CAN) protocol is widely adopted for in-vehicle communication but lacks inherent security mechanisms, making it vulnerable to cyberattacks. This paper introduces KD-GAT, an intrusion detection framework that combines Graph Attention Networks (GATs) with knowledge distillation (KD) to enhance detection accuracy while reducing computational complexity. In our approach, CAN traffic is represented as graphs using a sliding window to capture temporal and relational patterns. A multi-layer GAT with jumping knowledge aggregation acting as the teacher model, while a compact student GAT--only 6.32% the size of the teacher--is trained via a two-phase process involving supervised pretraining and knowledge distillation with both soft and hard label supervision. Experiments on three benchmark datasets--Car-Hacking, Car-Survival, and can-train-and-test demonstrate that both teacher and student models achieve strong results, with the student model attaining 99.97% and 99.31% accuracy on Car-Hacking and Car-Survival, respectively. However, significant class imbalance in can-train-and-test has led to reduced performance for both models on this dataset. Addressing this imbalance remains an important direction for future work.