DeTrAP: A Novel AI/ML V2X 5G NR Adaptive Physical Layer Configuration
Thanh-Son-Lam Nguyen, Sondes Kallel, Nadjib Aitsaadi · 2023
The 5G cellular network provides vital support for enabling fast and dependable communication in dynamic environments, which is crucial for connected autonomous vehicles. To achieve this goal, telecommunication operators must prioritize speedy and efficient radio resource management in 5G New Radio (NR) systems, achieved by dynamically adapting the configuration of the physical (PHY) layer. To address this issue, we introduce a novel method called Decision Tree Adaptive Physical Layer Configuration (DeTrAP), which utilizes machine learning and observational data to real-time fine-tune the PHY layer for efficient radio resource management. Extensive simulations demonstrate that DeTrAP achieves the expected performance for safety and non-safety traffic scenarios, while significantly reducing the convergence time.