Impact of Latency and Packet Loss on Virtual Platooning-Based Intersection Control in 5G-V2X MEC Environment

Abstract

Intelligent intersection management represents a promising future of transportation systems, aiming to coordinate connected and autonomous vehicles to improve safety and efficiency. However, many current coordination strategies rely on ideal communication assumptions that are rarely met in vehicle-to-everything (V2X) networks. This paper investigates the impact of non-ideal network conditions, specifically end-to-end latency and packet loss, on intersection management performance within a Multi-access Edge Computing (MEC) environment. A comparative analysis is conducted between a First-Come-First-Served baseline strategy and an advanced Virtual Platooning algorithm using an integrated simulation framework combining SUMO, OMNeT++, Simu5G, and a custom MEC application, enabling realistic simulation of traffic dynamics and 5G network behavior. The results demonstrate that while network degradation significantly increases safety risks for the baseline approach, the virtual platooning strategy provides substantial robustness by dynamically adapting vehicle trajectories and speeds. The findings show that virtual platooning significantly reduces the number of collisions and vehicle stops while maintaining higher traffic efficiency under different communication conditions. Nevertheless, the persistence of safety-critical events in extreme scenarios underscores that communication-dependent coordination cannot fully guarantee safety without additional network-aware fail-safe mechanisms. This study highlights the critical limitations of existing approaches and provides insights for designing resilient intersection management systems for real-world deployment.