One-Class Classification-Based Position Falsification Detection System in C-ITS Communication Network
摘要
Cooperative-Intelligent Transport Systems (C-ITS) is a system for ensuring road safety and user safety, such as traffic congestion and traffic accidents, by sharing traffic risk information in advance through real-time two-way data transmission between road traffic components. The main services provided by C-ITS, such as traffic jam and traffic accident prevention, are performed based on information derived by collecting and analyzing vehicle location information at the C-ITS Center. For this reason, if an attacker falsifies position data on the communication channel from the vehicle to the C-ITS Center with malicious purposes, most of the C-ITS services cannot operate normally, causing financial losses and casualties. This research proposes an effective position falsification attack detection system in C-ITS based on the environmental analysis of C-ITS. For the purpose of developing cyberattack detection systems, it is very difficult to collect training datasets by conducting real-world cyberattacks in most IT infrastructure environments, including C-ITS environments. For this reason, the C-ITS target position falsification attack detection system proposed in this research utilizes the method of training only normal data and detecting position falsification based on that information. For the position falsification attacks detection system proposed in this research, high performance of precision 99.97%, recall 100%, and F1 Score 99.98% was obtained using VeReMi and BurST-ADMA datasets, which are datasets containing normal data and position falsification data collected from ITS and VANET environment simulators.