資源描述：
《異構(gòu)云計算平臺中節(jié)能的任務(wù)調(diào)度策略研究》由會員上傳分享，免費在線閱讀，更多相關(guān)內(nèi)容在工程資料-天天文庫。
1、AbstractWiththepopularityandrapiddevelopmentofCloudComputingtechnology,taskscomingtoCloudComputingareinfinitevariety.Inordertomeetkindsofdemandsoftasks,computingnodesintheCloudComputingplatformhavetoremainopentowaitfortasks,whichmakesenergyefficiencyo
2、fclouddatacenterhascharacteristicoflowutilization,highwaste.AsanimportantpartofCloudComputingplatform,taskschedulingistheprocessoftasksdirectlymappedtoresourcesofclouddatacenter,whichdirectlyrespondsenergycostofdatacenterresources.Therefore,energyopti
3、mizationofclouddatacentercanbeimplementedbyreasonabletaskschedulingstrategy.Clouddatacenterisusuallycomposedoflarge-scaleheterogeneouscomputingnodeswhicharelinkedbydifferenttransmissionrates.Fromthedependenceofthedivisionoftasks,thisthesistakesenergy
4、savingindependenttaskschedulingtechnologyandenergy-savingdependenttaskschedulingtechnologyinheterogeneouscloudcomputingplatformasresearchobjectives.Themajorworkinthisthesiscanbeincludedfromthreeaspects:(1)Theexistingsourcesofhigh-energyconsumptionofth
5、eclouddatacenterareanalyzed,theenergyoptimizationmethodsinexistingdatacenters,researchstatusareexpounded,theproblemofindependenttaskschedulingforenergy-savinganddependenttaskschedulingforenergy-savingofheterogeneousCloudComputingplatformareanalyzed.(2
6、)TaskscomingtotheCloudComputingarerandomandcomputingnodesofCloudComputinghavetoremainopentowaitfortasks,whichproducesmuchwastedenergy.Anenergy-savingtaskschedulingalgorithmbasedonvocationqueuingmodelfortheCloudComputingispresented.First,taskscheduling
7、modeloftheheterogeneousCloudComputingisestablishedbyusingexhaustiveservice,vacationqueuingmodelsystem.Andthen,theaverageresponsetimeoftasksandtheaveragepowerofheterogeneouscomputenodesareanalyzedbyusingthebusyperiodandbusyspinundersteadystatecondition
8、sintheCloudComputing.Afterthat,ataskschedulingalgorithmbasedonsimilartasksisproposedtoreduceenergyconsumption.Simulationresultsshowthattheproposedalgorithmcanensurethetaskperformance,andreducetheenergyconsumptionoftheCloudComputingeffectively.
