In the context of 6th generation (6G) networks, vehicular edge computing (VEC) is emerging as a promising solution to let battery-powered ground vehicles with limited computing and storage resources offload processing tasks to more powerful devices. Given the dynamic vehicular environment, VEC systems need to be as flexible, intelligent, and adaptive as possible. To this aim, in this paper we study the opportunity to realize VEC via non-terrestrial networks (NTNs), where ground vehicles offload resource-hungry tasks to Unmanned Aerial Vehicles (UAVs), High Altitude Platforms (HAPs), or a combination of the two. We define an optimization problem in which tasks are modeled as a Poisson arrival process, and apply queuing theory to find the optimal offloading factor in the system. Numerical results show that aerial-assisted VEC is feasible even in dense networks, provided that high-capacity HAP/UAV platforms are available.
翻译:在第6代(6G)网络中,车辆边缘计算(VEC)正在成为一个大有希望的解决方案,让计算机和储存资源有限的电池驱动地面车辆将加工任务卸到更强大的装置上。鉴于动态车辆环境,VEC系统需要尽可能灵活、智能和适应性。为此,我们在本文件中研究通过非地面网络(NTN)实现VEC的机会,地面车辆向无人驾驶航空车辆(UAVs)、高高度平台(HAPs)或两者的组合卸载资源。我们定义了一个优化问题,在其中,任务以 Poisson 抵达过程为模型,并应用排队理论来寻找系统中的最佳卸载因素。数字结果显示,即使在密集的网络中,空中辅助VEC也是可行的,条件是具备高容量HAP/UAV平台。