Performance analysis of the air-ground integrated vehicular networks: A hierarchical model approach

Abstract

The rapid development of 5G/6G networks and Internet of things (IoT) has spawned numerous vehicular applications and services. The unmanned aerial vehicles (UAVs) can be conceived as edge computing servers to provide flexible, reliable and resilient services for vehicle users with limited processing capabilities. The goal of this paper is to quantitatively evaluate the service capability of the Air-Ground Integrated Vehicular Networks (AGIVN) to provide services for vehicle tasks. We first develop a hierarchical model for capturing the behaviors of the task service in AGIVN system, where a latency sensitive vehicle task can be temporarily stored due to UAV moving, and dynamically migrated between UAV clusters due to vehicle moving. Then, we derive the calculation formulas for performance metrics, including throughput, mean response delay, and task rejection probability. Compare numerical results with simulation results to verify the approximate accuracy of metric formulas and proposed model. By numerical analysis results, we evaluate the impact of various parameters on the AGIVN service performance in order to assist service providers better manage AGIVN system.