Trust management in the internet of vehicles: a survey of learning-based mechanisms

Abstract

The rapid advancements in information and communication technologies have resulted in the emergence of the Internet of Vehicles (IoV) as an indispensable constituent of intelligent transportation systems, enabling vehicles to exchange real-time data for improving road safety, traffic efficacy, and users’ convenience. However, as vehicles increasingly rely on this interconnected network, robust trust management mechanisms are essential to defend against threats that could undermine network integrity and consequently compromise road safety. While conventional mechanisms provide foundational security measures, they have limitations in detecting insider threats, particularly, as IoV environments scale and diversify. Therefore, intelligent learning-based mechanisms, i.e., machine learning, deep learning, and reinforcement learning, have become crucial for addressing these limitations since they are able to continuously adapt to complex dynamic threats within IoV networks. Their ability to autonomously learn behavioral features, generalize across diverse driving scenarios, and continuously refine trust decisions allows them to address the shortcomings of conventional mechanisms. This survey, therefore, offers a comprehensive review of the said learning-based mechanisms in the context of IoV-based trust management so as to assess their respective efficaciousness in mitigating trust-related attacks. It also discusses the adaptability, scalability, and robustness of such learning-based mechanisms thus highlighting their potential to meet the evolving challenges of IoV ecosystems. It furthermore delineates open research directions for developing more adept and scalable IoV-based trust management mechanisms.