Intrusion, anomaly, and attack detection in smart vehicles

Abstract

With the advancement of technology and Internet penetration in all aspects of life in today's modern world, smart vehicles, particularly connected vehicles, are rising. As a result, as smart vehicles have evolved and grown in popularity, so have the cyber threats that threaten them. Vehicles may communicate with other vehicles, road infrastructure, and other smart devices; as a result, the security of related equipment must be ensured. Researchers presented many articles investigating smart vehicle intrusion, anomaly, and attack detection methods. In most of these, researchers only examined the proposed detection methods on in-vehicle networks (IVNs) and external networks for smart ground vehicles. However, reviewing and analyzing the detection methods provided in all communications and connections related to smart ground/air vehicles is necessary. This paper examines the detection methods for in-vehicle networks, inter-vehicle networks, ground vehicle power stations, and the Internet of Drones (IoD). Besides, it analyzes recently published studies in the four categories: intrusion detection systems (IDSs), anomaly detection, attack detection, and hybrid detection methods in smart vehicles. The methods examined in this survey were published in 2018–2022 in reputable journals such as Elsevier, IEEE, Springer, ScienceDirect, etc. We examine datasets, simulations/implementations, and key evaluation criteria for each detection method studied. Also, we compare these methods using appropriate tables and diagrams and analyze the problems in detecting intrusions, anomalies, and attacks in smart vehicles. Finally, we address the research questions and discuss future research challenges.