Optimizing Radio Access in 5G Vehicle Networks Using Novel Machine Learning-Driven Resource Management

Abstract

Abstract The huge crossover of vehicle network needs cannot be met by present cellular technologies and vehicular networks. To achieve desired results in a vehicle context, resource management has evolved into a complicated task. The 5G wireless network claims to provide communications that are ultra-fast, delayed less, and dependable. Software-defined networking (SDN) is one of the major technologies that will enable 5G. Additionally, it guarantees increased performance all around. Managing the increased mobility of vehicles and maintaining smooth transfer between base stations are the main concerns. Additionally, providing safety-critical applications like autonomous driving requires very low latency and great dependability. Due to the limited amount of spectrum available and the dynamic nature of vehicular communication, effective resource allocation strategies are required. To maintain network efficiency, interference, and channel congestion must be reduced, and priority techniques are required to meet the needs of various services. In this study, we proposed a traffic classification using Magnified Recurrent Neural Network (MRNN) and optimized the radio access in 5g vehicle networks using Boosting Ant Colony Optimization (BACO) for the resource allocation procedure. The BACO-MRNN algorithm produced the best results based on several performance metrics, including accuracy of 98.10%, precision of 97.23%, recall of 98.15%, F1-Score of 98.45%, and RMSE of 30.10%. Additionally, the BACO-MRNN classification revealed a very advanced capacity for discrimination. The entire potential of connected vehicles and the fulfillment of smart transportation systems depend on the effective resolution of radio access difficulties in 5G vehicle networks.