面向低空智联网的无人机网络切片资源管理机制研究  

Research on UAV network slicing resource management for low-altitude intelligent network

作　　者：陈鸿杰[1] 郭昱甫 吴凡 张科 熊凯 CHEN Hongjie;GUO Yufu;WU Fan;ZHANG Ke;XIONG Kai(Southwest China Institute of Electronic Technology,Chengdu 610036,China;School of Information and Communication Engineering,University of Electronic Science and Technology of China,Chengdu 611731,China)

机构地区：[1]中国西南电子技术研究所,四川成都610036 [2]电子科技大学信息与通信工程学院,四川成都611731

出　　处：《电信科学》2025年第3期38-51,共14页Telecommunications Science

基　　金：国家自然科学基金青年项目(No.62201122)。

摘　　要：在低空智联网中,无人机作为空中通信基站、数据传输中继节点和移动网络终端的重要组成部分,凭借其卓越的机动性和适应性,广泛应用于扩展网络覆盖和支持多种业务服务。然而,由于低空智联网面临着网络拓扑动态变化、空域资源稀缺以及多样化业务需求等挑战,实现有限资源的高效编排和管理仍然是一项艰巨任务。为解决这一问题,通过对无人机网络进行端到端切片,构建满足特定需求的逻辑无人机网络架构。首先,设计了一种分群轨迹预测模型,用于确定分群接入节点的位置,为网络切片的资源预留与优化提供支持。基于此,提出了一种双时间尺度的资源管理框架:在大时间尺度上,采用非线性规划方法将切片重配置问题转化为约束优化问题,优化整体切片效益并合理预留资源;在小时间尺度上,通过针对切片内业务需求的资源调度策略,满足具体业务的传输服务质量(quality of service,QoS)需求。仿真结果表明,该方法增强了低空无人机智联网络在动态环境中的适应性与服务质量,为低空智联网复杂场景下的资源管理和业务保障提供了有效支持。In low-altitude intelligent networks,unmanned aerial vehicles(UAV)play a crucial role as aerial communi cation base stations,data relay nodes,and mobile network terminals.Leveraging their exceptional mobility and adapt ability,UAV can extend network coverage and support a wide range of service applications.However,the challenges of dynamic network topology,constrained airspace resources,and diverse service demand pose significant difficulties in achieving efficient resource orchestration and management.To address these challenges,an end-to-end slicing ap proach for UAV network was introduced,enabling the construction of logical network architectures tailored to spe cific requirements.A cluster trajectory prediction model was developed to identify the positions of clustered access nodes,providing essential support for resource reservation and optimization in network slicing.Building on this,a dual time-scale resource management framework was proposed.At a larger time scale,the slice reconfiguration prob lem was transformed into a constrained optimization task by a nonlinear programming approach,maximizing overall slice efficiency and ensuring rational resource reservation.At a finer time scale,intra-slice resource scheduling strate gies were implemented to meet the QoS requirements of specific services.Simulation results demonstrate that the pro posed method significantly improves the communication performance of low-altitude dynamic intelligent networks.It enhances the adaptability and service quality of UAV network slicing in dynamic environments,offering effective sup port for resource management and service assurance in complex scenarios.

关 键 词：无人机网络 网络切片 资源预留 双时间尺度 

分 类 号：TN915.5[电子电信—通信与信息系统]