ST-AAE: 시공간 적대적 오토인코더를 사용한 차량 내부 네트워크 침입 탐지
ST-AAE: Intrusion Detection in In-Vehicle Networks Using Spatio-Temporal Adversarial Autoencoder
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
차량 내 네트워크인 컨트롤러 영역 네트워크(CAN)는 악의적인 물리적 및 사이버 공격에 매우 취약하다. 사용자와 차량 제공자의 안전을 보장하기 위해 적절한 보안 조치가 필요하다. 본 논문에서는 차량 내 네트워크를 위한 새로운 침입 탐지 시스템(IDS)인 시공간 적대적 오토인코더(ST-AAE)를 제안한다. ST-AAE는 차량 내 IDS 분야에서 새로운 접근법으로, CAN 트래픽의 시공간 특성을 활용한다. 이 프레임워크는 비정상 트래픽을 탐지하고 공격 유형을 분류한다. 실제 운전 데이터에 대한 실험 결과, ST-AAE는 다양한 공격 유형에서 높은 정확도와 낮은 오탐률로 기존 모델보다 뛰어난 탐지 성능을 보였다. 이러한 성능은 시공간 특징을효과적으로 학습하고, 적대적 학습을 통해 일반화를 향상시키며, CAN ID와 페이로드 정보를 통합하는 능력에 기인한다. 이러한 결과는 ST-AAE가 자동차 네트워크 보안 분야에서 효과적인 솔루션이 될 수 있음을 시사한다.
Abstract
The Controller Area Network (CAN), the in-vehicle network, is highly vulnerable to malicious physical and cyber attacks. To ensure the safety of both users and vehicle providers, appropriate security measures are necessary. This paper proposes a novel intrusion detection system (IDS) called Spatio-Temporal Adversarial Autoencoder (ST-AAE) for in-vehicle networks. ST-AAE utilizes the spatio-temporal characteristics of CAN traffic, a new approach in the field of in-vehicle IDS. The framework detects abnormal traffic and classifies attack types. Experimental results on real-world driving data show that ST-AAE outperforms existing models with high accuracy and low false positive rates across various attack types. Its performance is due to effective spatio-temporal feature learning, enhanced generalization through adversarial learning, and integration of CAN ID and payload information. These findings suggest that ST-AAE is an effective solution for automotive network security.
참고문헌 (19)
- F. Ferrero, "Car-sharing services : An annotated review", Sustainable Cities and Society, 37, pp.501~518 (2018)[DOI]
- E. W. Martin, "Greenhouse Gas Emission Impacts of Carsharing in North America", IEEE Transactions on Intelligent Transportation Systems, 12(4), pp.1074~1086 (2011)[DOI]
- S. Woo, "A Practical Wireless Attack on the Connected Car and Security Protocol for In-Vehicle CAN", IEEE Transactions on Intelligent Transportation Systems, 15(1), pp.1~14 (2014)[DOI]
- F. Fakhfakh, "Cybersecurity attacks on CAN bus based vehicles : a review and open challenges", Library Hi Tech, 40(5), pp.1179~1203 (2021)[DOI]
- M. Bozdal, "Evaluation of CAN Bus Security Challenges", Sensors, 20(8), pp.2364~ (2020)[DOI]
- "Hackers remotely kill a Jeep on the highway—with me in it" (2015)[URL]
- S. Chen, "Evaluation of DoS Attacks on Vehicle CAN Bus System, Smart Innovation, Systems and Technologies", pp.308~314 (2018), Springer International Publishing
- K. Iehira, "Spoofing attack using bus-off attacks against a specific ECU of the CAN bus", Proceedings of the 2018 15th IEEE Annual Consumer Communications & Networking Conference (CCNC), pp.1~4 (2018)
- A. Taylor, "Anomaly Detection in Automobile Control Network Data with Long Short-Term Memory Networks", Proceedings of the 2016 IEEE International Conference on Data Science and Advanced Analytics (DSAA), pp.130~139 (2016)
- E. Seo, "GIDS: GAN based Intrusion Detection System for In-Vehicle Network", Proceedings of the 2018 16th Annual Conference on Privacy, Security and Trust (PST), pp.1~6 (2018)
- M. Delwar Hossain, "An Effective In-Vehicle CAN Bus Intrusion Detection System Using CNN Deep Learning Approach", Proceedings of the GLOBECOM 2020 - 2020 IEEE Global Communications Conference, pp.1~6 (2020)
- H. Kang, "Car Hacking and Defense Competition on In-Vehicle Network", Proceedings Third International Workshop on Automotive and Autonomous Vehicle Security, Internet Society (2021)
- H. K. Kim, "Car Hacking: Attack &Defense Challenge 2020 Dataset", IEEE (2021)
- A. Makhzani, "Adversarial Autoencoders", Proceedings of the ICLR Workshop (2016)
- A. Paszke, "Pytorch : An imperative style, high-performance deep learning library", Proceedings of the 33rd International Conference on Neural Information Processing Systems, (721), pp.8026~8037 (2019)
- X. Shi, "Convolutional LSTM network : A machine learning approach for precipitation nowcasting", Advances in Neural Information Processing Systems, 28, pp.802~810 (2015)[DOI]
- D. P. Kingma, "Adam : A method for stochastic optimization", Proceedings of the International Conference on Learning Representations (ICLR) (2015)
- D. Park, "A Multimodal Anomaly Detector for Robot-Assisted Feeding Using an LSTM-Based Variational Autoencoder", IEEE Robotics and Automation Letters, 3(3), pp.1544~1551 (2018)[DOI]
- D. P. Kingma, "Autoencoding variational bayes", Proceedings of the 2nd International Conference on Learning Representations (ICLR), Conference Track Proceedings (2014)