AI-Assisted Dynamic RSU Deployment Using Real Time Traffic Density Estimation in VANETs: A Comprehensive Survey

Abstract

Vehicular Ad Hoc Networks (VANETs) have become a cornerstone of Intelligent Transportation Systems (ITS), enabling real-time communication among vehicles, roadside infrastructure, and transportation management entities. Roadside Units (RSUs) play a critical role in supporting Vehicle-to-Infrastructure (V2I) communications, traffic monitoring, safety applications, and information dissemination. However, conventional static RSU deployment strategies often fail to accommodate rapidly changing traffic conditions, resulting in reduced coverage efficiency, increased communication delays, and inefficient utilization of infrastructure resources. To address these limitations, recent research has focused on integrating Artificial Intelligence (AI) techniques with real-time traffic density estimation to support dynamic and adaptive RSU deployment. This survey presents a comprehensive review of AI assisted dynamic RSU deployment approaches in VANET environments. The study examines traffic density estimation methods, RSU deployment strategies, optimization algorithms, machine learning models, deep learning techniques, and emerging intelligent transportation frameworks. Furthermore, the survey categorizes existing literature according to deployment methodology, traffic prediction approach, and intelligence level while providing a comparative analysis of current solutions. Major research challenges including scalability, computational complexity, security, privacy, and infrastructure costs are discussed. Finally, future research directions involving reinforcement learning, federated learning, edge intelligence, digital twins, unmanned aerial vehicle-assisted RSUs, and 6G-enabled vehicular networks are highlighted. The survey aims to serve as a comprehensive reference for researchers and practitioners working toward the development of adaptive, intelligent, and efficient vehicular communication infrastructures.