基于用户满意度的边缘计算的资源分配策略  

Resource allocation strategy of edge computing based on user satisfaction

作　　者：邵旺 林晓勇[1] SHAO Wang;LIN Xiao-yong(College of Communication and Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210003,China)

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003

出　　处：《计算机工程与设计》2023年第5期1297-1304,共8页Computer Engineering and Design

基　　金：国家自然科学基金项目(61801240)。

摘　　要：现有的关于边缘计算的资源分配问题研究大多关注单一资源的分配。针对这一问题,根据用户类型的特点综合考虑计算、网络以及存储资源的分配对于用户的影响,结合边缘、云以及本地的资源情况,制定一个基于用户满意度的资源分配策略。基于流行度缓存请求内容提高用户边缘访问率,搭建用户移动、资源分配以及满意度模型,引入改进的遗传算法求解用户满意度最大化问题。场景仿真对比其它的策略,验证了其优越性。Aiming at the resource allocation of edge computing,most of the existing studies focus on the allocation of single resources.For this problem,according to the characteristics of user types,the impact of computing,network and storage resource allocation on users was comprehensively considered,and a resource allocation strategy based on user satisfaction was formulated in combination with edge,cloud and local resource conditions.The request content was cached based on popularity to improve the user edge access rate,user mobility model,resource allocation model and satisfaction model were built,and an improved genetic algorithm was introduced to solve the problem of maximizing user satisfaction.Through simulation and comparison with other strategies,its superiority is verified.

关 键 词：边缘计算 边缘缓存 用户类型 用户移动 资源分配 用户满意度 遗传算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]