雾计算环境下基于改进遗传算法的自适应任务分配算法.pdfVIP

长，信息通信

DOI:10.20153/j.issn.2096-9759.2025.03.017

雾计算环境下基于改进遗传算法的自适应任务分配算法

韩奎奎，付阿敏，洪学辰

安阳职业技术学院信息工程学院，河南安阳455000

AdaptiveTaskAllocationAlgorithmBasedonImprovedGeneticAlgorithm

InfogComputingEnvironment

HANKuikui,FUAmin,HONGXuechen

FacultyofInformationEngineering,AnyangVocationalandTechnicalCollege,AnyangHenan455000,China

摘要：终端移动用户在雾端资源节点的Abstract:Theexpenditureincurredbymobileend-usersatfogresourcenodesisprimarilydeter-

花销支出主要取决于CPU的执行时间，minedbytheCPUexecutiontime.Consequently,toreducecosts,itisnecessarytooptimizethe

因此想要降低支出成本，则需对雾端资taskallocationstrategyatfogresourcenodes,therebyenhancingprocessingefficiencyandmini-

源节点的任务分配策略进行优化，以提mizingboththeworkloadandexecutiontimeoftheCPU.Thebasicgeneticalgorithm,asaclas-

高处理效率，降低CPU的运行负担和执sicapproachintheficldoftaskallocation,exhibitsremarkableeffectivenessindistributingtask

行时间。基本遗传算法作为任务分配领requestsamongresourcenodes.However,thealgorithmfacesaprimarychallengewhenexplo-

域内的经典算法，在分配资源节点上的ringtheproblemspace:itispronetoprematureconvergence,whereitstallsprematurelyatalo-

任务请求时展现出了显著的有效性。然caloptimalsolution,failingtocomprehensivelyexploreanddiscoverasuperiorglobalallocation

而,该算法在探索问题空间时面临一个scheme.Addressingthestatedproblem,thispaperdesignsandimplementsanoptimizedadaptive

主要挑战：易于发生早熟收敛，即算法过taskallocationalgorithmbasedonthefundamentalgeneticalgorithm.Thisalgorithmemploys

早地停滞在局部最优解上，从而未能全theMetropolisacceptancecriterion,whichallowsfortheacceptanceofbadsolutionswitha

面探索并发现更优的全局分配方案。针certainprobabilityduringthesearchforallocationoutcomeswithintheproblemspace.Experi-

对所述问题，