摘要
在混合部署场景下,为减少时延,提升多个终端用户的综合体验,提出一种无人机作为动态边缘服务器部署到高空平台辅助卸载,并结合视距通信路径损耗、通信、计算多阶段模型的时延优化任务卸载方法。将时延优化建模为一个非凸问题,用整数非线性规划模型来表述,求解中进一步将其分割为多个时隙内任务卸载时延的最小化问题,结合系统模型、任务卸载流程、优化目标,将卸载策略优化分为两阶段,找到原问题的较优解。仿真结果表明,该方法与其它同类研究方案相比,能够充分利用无人机动态边缘服务器资源减少时延,具有良好的可扩展性。
To enhance the comprehensive experience of multiple end-users in hybrid deployment scenarios,a method of optimizing the latency of computation offloading was presented.It involved the application of UAVs as dynamic edge servers to assist offloading on the high-altitude platform station and the multi-stage model of line-of-sight communication path loss,communication,and computation was combined.To be specific,the delay optimization was modeled as a non-convex problem,which was formulated with an integer nonlinear programming model,and solved by treating it as the problem of minimizing the offloading delay over multiple time slots.Taking the system model,task offloading process,and optimization objective into consideration,the optimization strategy was divided into two phases,to find the better solution of the original problem.Simulation results show that compared with other solutions,this method is able to fully utilize the UAV dynamic edge server resources to reduce the latency and it can be easily extended.
作者
霍英卓
李爱萍
段利国
赵菊敏
HUO Ying-zhuo;LI Ai-ping;DUAN Li-guo;ZHAO Ju-min(College of Computer Science and Technology(College of Data Science),Taiyuan University of Technology,Jinzhong 030600,China)
出处
《计算机工程与设计》
北大核心
2025年第3期726-733,共8页
Computer Engineering and Design
基金
山西省自然科学基金面上基金项目(202303021211052、202203021221234)
国家自然科学基金面上基金项目(61972273)
山西省关键核心技术和共性技术研发攻关专项子课题基金项目(20201102021)。
关键词
多接入边缘计算
无人机
混合部署
计算卸载
视距通信
卸载时延
时延优化
multi-access edge computing
unmanned aerial vehicles
hybrid deployment
computation offloading
line-of-sight communication
unloading delay
time delay optimization
作者简介
霍英卓(1999),男,山西吕梁人,硕士研究生,CCF学生会员,研究方向为边缘计算;通讯作者:李爱萍(1974),女,山西吕梁人,博士,教授,CCF高级会员,研究方向为边缘计算、智能信息处理;段利国(1970),男,山西忻州人,博士,副教授,CCF高级会员,研究方向为自然语言处理、智能信息处理;赵菊敏(1976),女,山西大同人,博士,教授,研究方向为物联网关键技术、信号与信息处理。E-mail:1915498598@qq.com。
