Edge Intelligence Empowered Vehicular Metaverse: Key Design Aspects and Future Directions

Latif U. Khan, Ahmed Elhagry, Mohsen Guizani, Abdulmotaleb El-Saddik

Abstract: Emerging intelligent transportation system applications witnessed significantly different requirements and performance metrics (e.g., latency, reliability, and quality of experience). To meet the diverse requirements, one can use a convergence of the metaverse with vehicular networks at the network edge which offers proactive analysis and efficient real-time control for the management of vehicular network resources. Therefore, in this article, we present key design aspects of an edge intelligence-enabled vehicular metaverse. We also present a high-level architecture for an edge intelligence-based vehicular metaverse that has three main aspects: a metaverse engine, offline learning, and online real-time control. Moreover, we present two case studies: joint sampling and packet error rate minimization and object detection task at the network edge. Finally, we conclude the article.