Indian SUMO traffic scenario-based misbehaviour detection dataset for connected vehicles

Abstract

The Internet of Vehicles (IoV) plays a crucial role in intelligent transportation systems (ITS) by enabling communication between interconnected vehicles and supporting infrastructure. Connected vehicles utilize basic safety messages (BSMs) to exchange kinematic data, such as vehicle acceleration, velocity, position, and direction, with neighbouring nodes in the ITS network to enhance road safety. However, these BSMs are susceptible to various security attacks, which disrupt the collaborative functionality of ITS, potentially resulting in accidents or traffic congestion. The scientific community has proposed numerous security mechanisms to protect BSMs. The majority of these assessments have been conducted utilizing either the vehicular reference misbehaviour (VeReMi) dataset or the VeReMi extension dataset. These datasets are specifically designed for the Luxembourg SUMO Traffic (LuST) scenario and are suitable for only evaluating misbehaviour detection methods within a European ITS context. However, there is a notable scarcity of publicly accessible misbehaviour datasets that faithfully depict Indian ITS scenarios. To overcome this limitation, we introduce a new scenario, i.e., the Ahmedabad SUMO Traffic (AhmST) scenario, based on the city of Ahmedabad in Gujarat, India. Moreover, we also introduce the Indian dataset for misbehaviour analysis (AhmST). The proposed dataset includes cases of false data injections affecting the vehicle position, heading, and speed information within BSMs. Finally, we compare the AhmST dataset with recent datasets, assess the proposed dataset using various machine learning techniques and present an optimized model with improved accuracy.