Optimizing end-to-end latency in C-V2X networks: A novel FD-RAN and MEC integration approach

Abstract

The increasing demand for low-latency services in cellular vehicle-to-everything (C-V2X) communications is crucial for the efficient operation of connected vehicles and autonomous driving systems. As C-V2X networks become integral to modern transportation infrastructure, minimizing end-to-end (E2E) latency remains a significant challenge in ensuring system reliability and effectiveness. To this end, we propose a novel approach that integrates a fully decoupled radio access network (FD-RAN) with multi-access edge computing (MEC) in C-V2X networks, aiming to optimize latency-sensitive applications. We apply a tractable analytical model for the E2E latency that accounts for latency contributions across the radio, backhaul, network, and processing layers. By leveraging FD-RAN's decoupled access and MEC's distributed processing, our approach effectively mitigates latency bottlenecks inherent in conventional RAN architectures. Additionally, we propose a stochastic optimization-based resource allocation method using Lyapunov techniques and Markov decision processes to dynamically manage base station selection and bandwidth allocation, thereby enhancing the system performance. Simulation results demonstrate that FD-RAN with MEC significantly reduces E2E latency compared with conventional RAN architectures, even under high traffic densities, while maintaining high data rates. These findings validate the proposed approach and offer key insights for developing low-latency infrastructures for next-generation V2X applications.