Enhancing Vehicular Network Security, Privacy, and Trust Through Reinforcement Learning: A Comprehensive Survey

The evolution of vehicular networks from vehicular ad-hoc networks (VANETs) to Internet of Vehicles (IoVs) has played a pivotal role in the intelligent transportation system (ITS). However, these networks are increasingly vulnerable to security, privacy, and trust (SPT) threats due to various emerging attacks. In response, Reinforcement Learning (RL) has emerged as a promising technique for strengthening vehicular network security. This paper provides a comprehensive exploration of the SPT challenges within vehicular networks and presents RL as a promising solution for enhancing SPT provisioning. First, we provide a tutorial on vehicular networks and integrated concepts, and the overview of RL concepts and different types of RL. Then, we conduct a detailed analysis of existing RL-based solutions, categorizing them within two novel taxonomies: one based on the specific SPT focused area and the other on the specific RL methods employed. We conclude by discussing key lessons learnt, current open challenges, and potential future directions in this rapidly evolving field.