Intelligent O-RAN-Based Proactive Handover in Vehicular Networks

The ongoing efforts of wireless service providers to build a Radio Access Network (RAN) architecture have been noted in recent years. The primary goal was to design an operator-defined RAN architecture capable of providing intelligent radio control for Fifth-Generation (5G) wireless networks as well as Beyond 5G (B5G). The Open Radio Access Network (O-RAN) Alliance was formed to transform the telecommunications ecosystem. In this paper, we propose an intelligent O-RAN framework for vehicle communication. We propose a Machine Learning (ML)-based model to predict the compatibility time during which a vehicle remains within communication range with another vehicle to establish a connection. We compare the performance of Gaussian Naive Bayes (GNB), K Nearest Neighbor (KNN), and Neural Networks (NN) in terms of training and testing accuracy. We believe compatibility time estimation helps to implement proactive forwarding for optimal network performance. Finally, we conclude our work by providing directions for future research.