Threat Analysis for Automotive CAN Networks: A GAN Model-Based Intrusion Detection Technique

Article Computer Science, Hardware & Architecture

Intrusion Detection for In-vehicle Network by Using Single GAN in Connected Vehicles

Yuanda Yang et al.

Summary: Controller area network (CAN) bus-based connected and even self-driving vehicles face severe cybersecurity challenges due to external connections and vulnerabilities, leading to privacy and security threats. Generative adversarial nets (GAN)-based intrusion detection systems (IDSs) can overcome limitations of insufficient attack data types in deep learning-based IDSs. This study proposes an improved GAN-based intrusion detection method for in-vehicle networks, which enhances evaluation metrics and reduces detection time by utilizing a new loss function and sparse enhancement training method.

JOURNAL OF CIRCUITS SYSTEMS AND COMPUTERS (2021)

Add to Collection

Article Engineering, Civil

Security Enhancement for Real-Time Parallel In-Vehicle Applications by CAN FD Message Authentication

Guoqi Xie et al.

Summary: The study proposes a security enhancement for CAN FD in-vehicle communication through a two-stage method, aiming to improve the security of in-vehicle applications.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2021)

Add to Collection

Article Engineering, Civil

Preserving Privacy in the Internet of Connected Vehicles

Soheila Ghane et al.

Summary: This paper discusses privacy issues in distributed edge computing, addressing data communication between multiple vehicles in the IoT layer and potentially untrusted edge controllers. The proposed mechanism ensures a strong privacy level by adding correlated noise in the vehicle's side, scaling with data correlation. Extensive experiments demonstrate that this approach outperforms recent methods in terms of utility.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2021)

Add to Collection

Article Computer Science, Hardware & Architecture