2024年│1期│2072024年第46卷第1期基于强化学习的移动边缘计算资源分配方法韩宏飞袁倩王益军徐磊徐肃涵作者简介:韩宏飞(2001-),本科,研究方向为计算机。(淮阴工学院江苏淮安223003)摘要当前的移动边缘计算资源分配结构多为单向形式,资源分配效率较低,导致资源分配比下降,文中设计了一种基于强化学习的移动边缘计算资源分配方法,并通过实验验证了其有效性。根据当前的测试需求,首先部署了资源采集节点,然后采用多阶的方式,提升整体的资源分配效率,构建多阶迁移资源分配结构,最后设计了移动边缘计算强化学习资源分配模型,采用动态化辅助协作处理的方式来实现资源分配。测试结果表明,对于选定的5个测试周期,经过3个分配组的测定及比对,最终得出的资源分配比均可以达到5.5以上,这说明在强化学习技术的辅助下,文中设计的移动边缘计算资源分配方法更加灵活、多变,针对性较强,具有实际的应用价值。关键词:强化学习;移动边缘;边缘计算;资源分配;分配方法;资源整合中图分类号TP311.5ResourceAllocationMethodofMobileEdgeComputingBasedonReinforcementLearningHANHongfei,YUANQian,WANGYijun,XULeiandXUSuhan(HuaiyinInstituteofTechnology,Huai’an,Jiangsu223003,China)AbstractThecurrentmobileedgecomputingquotastructureismostlyone-wayform,andthequotaefficiencyislow,resultinginadecreaseinthequotaratio.Amobileedgecomputingquotamethodbasedonreinforcementlearningisde-signedinthispaper,anditseffectivenessisverifiedbyexperiments.Accordingtothecurrenttestingrequirements,theresourceacquisitionnodeisfirstdeployed,andthenthemulti-ordermethodisadoptedtoimprovetheoverallquotaeffi-ciencyandconstructamulti-ordermigrationquotastructure.Finally,themobileedgecomputingreinforcementlearningquotamodelisdesigned,andthedynamicauxiliarycooperativeprocessingmethodisusedtorealizethequota.Thetestre-sultsshowthatfortheselected5testcycles,afterthemeasurementandcomparisonof3allocationgroups,thefinalquotaratiocanreachmorethan5.5,indicatingthatwiththeassistanceofreinforcementlearningtechnology,themobileedgecomputingquotamethoddesignedinthispaperismoreflexible,changeable,andhighlytargeted,andhaspracticalappli-cationvalue.KeywordsReinforcementlearning,Movingedges,Edgecomputing,Resourcealloc...
