The invention discloses a dynamic task unloading method and system for assisting a high-speed rail edge computing network by an unmanned aerial vehicle, and relates to the technical field of wireless communication, and the method comprises the steps: constructing an air-space integrated communication system for high-speed rail communication, and the system is composed of a high-speed rail base station, a roadside base station and an unmanned aerial vehicle base station; based on an air-space integrated communication system, constructing an optimization problem taking minimization of calculation time delay and energy consumption as a target; decomposing and converting the optimization problem into sub-problems of a plurality of time slots; on the basis of the sub-problems, a D3QN-TD3 deep reinforcement learning method is adopted, and a domain adaptive algorithm is used for processing until the optimization problem is converged to preset precision; according to the method, the unloading strategy and the transmitting power are jointly optimized, the energy storage stability of the unmanned aerial vehicle is guaranteed at the same time, the calculation time delay is reduced by 10% compared with that only the ground base station participates in calculation, and the service time delay can be effectively reduced.
本发明公开了一种无人机协助高铁边缘计算网络的动态任务卸载方法及系统,涉及无线通信技术领域,方法包括:构建一个高铁通信的空天一体化通信系统,所述系统由高铁基站、路边基站,无人机基站组成;基于空天一体化通信系统,构建以最小化计算时延和能耗为目标的优化问题;将所述优化问题分解转化为多个时隙的子问题;基于所述子问题,采用D3QN‑TD3的深度强化学习方法,以及利用域自适应算法进行处理,直到优化问题收敛到预先设定的精度;本发明通过联合优化卸载策略和发射功率,同时保证无人机储能稳定性,相较于只有地面基站参与计算的计算时延降低了10%,能够有效地降低服务时延。
Dynamic task unloading method and system of unmanned aerial vehicle assisted high-speed rail edge computing network
无人机协助高铁边缘计算网络的动态任务卸载方法及系统
2024-07-02
Patent
Electronic Resource
Chinese
European Patent Office | 2023
|European Patent Office | 2025
|