基于云边协同的车联网资源分配与任务卸载

打开文本图片集

中图分类号：U463.67 文献标志码：A DOI：10.20104/j.cnki.1674-6546.20250131

Resource Allocationand Task Offloading in Vehicular Networks Based on Cloud-Edge Collaboration

Ge Pingzheng1，2 （1.Jiangxi Vocationaland Technical CollgeofCommunications，Nanchang 33013;2.East ChinaJiaotong University， Nanchang 330013）

【Abstract]Toaddresstheissues of effcientresource allocationandreasonable task offloading in the Internetof Vehicles environment，researchbasedoncloud-edgecollaboration iscariedout.A jointofloadingschemeoflocal-edgeserver-cloud serverisproposed.Byitegatingmulti-levelresoures，thetaskprocessngflowisoptiized，andteesourceutilzatiate isimproved.Meanwhile，theTwinDelayedDeepDeterministicPolicyGradient（TD3PG）algorithmisintroducedtodecidethe taskoffloadingstrategy.Thesimulationresultsshowthatthisjointofloadingschemeeffectivelyimproves thesystemresource utilizationrate.TheproposedTD3PGalgorithm performs excellntlyinoptimizing task ofloadingdecisions.Compared with the traditional DeepDeterministicPolicyGradient （DDPG），ParticleSwarmOptimization（PSO），andGeneticAlgorithm（G），the Quality of Experience （QoE） of users is increased by 7.41% 20.83% and 45.00% ，respectively.

Keywords：Cloud-edge collaboration，Resource allocation，Task offloading，Twin Delayed DeepDeterministic PolicyGradient（TD3PG）

1前言

车载移动设备已成为现代交通体系中不可或缺的组成部分-2。（剩余6661字）