A misbehavior detection system to detect novel position falsification attacks in the Internet of Vehicles

Abstract

In the Internet of Vehicle (IoV) networks, vehicles exchange periodic Basic Safety Messages (BSMs) containing information regarding speed and position. Safety-critical applications like blind-spot warning and lane change warning systems use the BSMs to ensure the safety of road users. To create chaos in the network, an insider attacker may inject false information into the BSM and broadcast it to nearby vehicles. One such attack is the position falsification attack, where the attacker inserts false information regarding the position in the BSMs. The literature has explored the use of Misbehavior Detection Systems (MDSs) to detect such attacks. But the limitaitons of the existing approaches are that they either perform exceptionally well in specific environmental settings or have compromised detection accuracy favoring a generalized model. Moreover, all the current machine learning-based detection models are signature-based, which requires prior knowledge about the attacks for effective detection. Motivated by the research gap, we propose a Novel Position Falsification Attack Detection System for the Internet of Vehicles (NPFADS for the IoV) to learn and detect novel position falsification attacks emerging in IoV networks. The performance of NPFADS is analyzed using the metrics precision, recall, F1 score, ROC, and PR curves. The Vehicular Reference Misbehavior (VeReMi) dataset is used as the benchmark for the study. The system's performance is compared to existing MDSs in the literature. The analysis shows that our proposed system outperforms the existing supervised learning models even when initialized with zero knowledge about the novel position falsification attacks.