Vehicular ad hoc networks (VANETs) and the Internet of Things (IoT) are fundamental components of intelligent transportation systems (ITS). The evolution toward 6G‐enabled vehicular networks and the integration of heterogeneous sensors in modern vehicles have created new opportunities for improving routing efficiency, mobility management, scalability, and security in vehicular IoT (VIoT). However, highly dynamic topologies and strict quality‐of‐service requirements introduce significant technical challenges. Machine learning (ML), particularly neural network–based approaches, has emerged as a promising solution for addressing these issues. Despite the rapid growth of ML‐driven VIoT research, a structured and up‐to‐date systematic review remains limited. This paper presents a comprehensive systematic review of ML applications in VIoT, analyzing 30 selected studies and categorizing them using a newly proposed taxonomy based on training mechanisms and data utilization strategies. The reviewed works are comparatively evaluated in terms of application domains, performance metrics, challenges, and future research directions. The findings demonstrate that ML techniques enhance routing performance, enable accurate traffic congestion prediction, and support intelligent decision‐making. Nevertheless, security, scalability, and real‐world deployment remain open challenges requiring further investigation.
Song et al. (Thu,) studied this question.