基于移动边缘计算的车联网缓存策略研究  

Research on caching strategy of Internet of Vehicles based on mobile edge computing

作　　者：刘可欣 陈桂芬[1] Liu Kexin;Chen Guifen(School of Electronics&Information Engineering,Changchun University of Science&Technology,Changchun 130022,China)

机构地区：[1]长春理工大学电子信息工程学院,长春130022

出　　处：《计算机应用研究》2021年第3期851-854,870,共5页Application Research of Computers

基　　金：吉林省科技厅资助项目(20190302103GX)。

摘　　要：针对车联网中数据流量爆炸式增长而引起的业务响应时延过高的问题,提出了一种基于移动边缘计算的蚁群模拟退火算法缓存策略(ACSAM)。首先,在基于5G的车—边—云协同系统架构下,以最小化内容下载时延为目标,建立了通信计算模型;其次,采用蚁群算法构造了使内容下载时延最小的局部最优解;最后,使用模拟退火算法对局部最小下载时延进行扰动,并以一定概率接受新解,从而得到全局最小下载时延,即保证了内容被预缓存在最佳的位置。仿真结果表明,在车—边—云协同架构下,ACSAM缓存策略可显著减少传输冗余,降低下载时延。To solve the problem of high service response delay caused by explosive growth of data traffic in the Internet of Vehicles,this paper proposed an ant colony simulated annealing algorithm cache strategy based on mobile edge computing(ACSAM).Firstly,under the 5G based vehicle-edge-cloud collaborative system architecture,it established a communication computing model to minimize the content download delay.Secondly,it used ant colony optimization to construct a local optimal solution to minimize the content download delay.Finally,it used the search ability of simulated annealing algorithm to disturb the local optimal solution and accepted the new solution with a certain probability,thereby obtained the global optimal solution and improved cache hit rate.Simulation results show that under the vehicle-edge-cloud collaborative architecture,the ACSAM strategy can significantly reduce transmission redundancy and reduce download latency.

关 键 词：车联网 移动边缘计算 缓存策略 蚁群模拟退火算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]