Vehicle-to-Everything Cloud Collision Prediction Architecture for Random Forest and Software-Defined Networking

Abstract

The rising number of road accidents worldwide underscores the urgent need to implement Vehicle- to-Everything (V2X) communications, which can help minimize fatalities and injuries. This paper proposes a novel cloud-based architecture with machine learning capabilities that enables smooth, intelligent management of V2X services. The system features a backup controller that can take over in case of failure, ensuring uninterrupted operation. Additionally, the ML models provide invaluable optimization, from predicting traffic patterns to maximizing resource utilization and connection quality. With their independent decision-making, these Artificial Intelligent/ Machine Learning (AI/ML) functions act as a catalyst, enhancing overall efficiency. This research sheds new light on employing AI/ML in integrating non-terrestrial networks, addressing the complex challenges of massive data analytics under strict requirements. By outlining future research directions, it makes a significant contribution to the knowledge base around intelligent V2X systems. The proposed architecture demonstrates how AI/ML can be harnessed to create robust, seamless networks that help protect human lives on the road. Consequently, this paper introduces a novel architecture leveraging Software-Defined Networking (SDN) over the cloud, which decouples the control plane from the data plane, providing enhanced network programmability and scalability while minimizing costs and network congestion. Additionally, the paper presents a Vulnerable Road User (VRU) accident detection system within this architecture. The proposed method be able to analyze the incoming data in real time, identify potential collision risks with more than 90% accuracy, and provide warnings or take preventive actions to miti-gate the risk of accidents. This can greatly improve road safety by enabling vehicles to make informed decisions and take appropriate measures to avoid collisions