A Decentralized V2V Communication Framework with Traffic Prediction and Accident Severity Detection

Abstract

This paper presents an intelligent Vehicleto-Vehicle Communication system integrated with Machine Learning techniques to enhance road safety and traffic efficiency. The proposed system predicts accident severity, suggests optimal routes based on traffic congestion and weather conditions, and dynamically controls traffic signals. A decentralized communication framework enables real-time data exchange among vehicles, reducing latency and improving responsiveness. The system utilizes algorithms such as Decision Tree and Random Forest to achieve high prediction accuracy using real-world datasets. Simulation results demonstrate improved traffic flow and reduced congestion through adaptive signal control. Overall, the proposed approach provides a scalable and efficient solution for next-generation intelligent transportation systems. Index terms - — Vehicle-to-Vehicle Communication (V2V), Machine Learning, Traffic Prediction, Accident Severity Analysis, Intelligent Transportation Systems, Random Forest, Decision Tree, Traffic Signal Control.