实时信息采集任务的服务质量保证技术分析-analysis of service quality assurance technology for real-time information acquisition tasks.docxVIP

AbstractWiththerapiddevelopmentoftheInternet,informationoftheInternetincreasesrapidly.MorepeopleparticipateintheInternettoexpresstheiropinionsandcommunicatewitheachother.Inordertofullyanalyzeandpredictthetrendsofthepublicopinion,manyrelatedtechnologieshaveemerged.Bytheendof2012,thereisatotalnumberofmorethan60billionwebpagesinChina,andremained100%growthratesince2003.Itsexplosivegrowthleadsthegrowthofthepublicopiniondatabytheexponentialexpansionofthenumber.Publicopinioninformationcollectionmustbedoneatahighspeedinordertocopewiththechangesinpublicopinion.Outdatedinformationoftenmisstheopportunitiesandlosevalue.Therefore,returningthecollectionofreal-timeinformationhasimportantsignificance.Themaincontentofthispaperisthatsearchrequestsfromtheprocessorschedulingandresourcemanagementproblemsbyreptilescanbeabstractedasamodelinwhichreal-timetasksandduplicatedreal-timetasksschedulinginheterogeneousmultiprocessors.Real-timetasksschedulingmethodsaremainlybasedonthetraditionalsingle-processorschedulingmethodssuchasratemonotonicscheduling,andmultiprocessorschedulinglikelocalschedulingandglobalschedulingmethodandevolutionaryalgorithmsscheduling.Eachmethodhasitsadvantagesanddisadvantagesandadaptationconditions.Moreover,duringthereal-timetaskallocationprocess,thismethoddidnotconsidertheproblemofenergyconsumption,evenifthepartitionresultisafeasiblesolution,butincreasingenergyconsumption.Byanalyzingthelimitationsoftheabovemethods,westudythereal-timetaskallocationmethodsandproposedanimprovedalgorithmbasedonparticleswarmoptimizationinheterogeneousprocessors,increasingenergyconsumptionconstraints,satisfingtheallocationoftasksatthesametimereducingenergyconsumption.Meanwhile,astoduplicatedreal-timetasksallocationmethod,wepresentalinearprogrammingalgorithm,approximatedynamicprogrammingalgorithmandpolynomialtimecomplexityalgorithmtosolvetheproblem,whichhasreachedthepurposeofamorequicklyresultandincreasingtheposibilityoffindingafeasiblesolution.Throughtheexperiments,wenotonlysimulateonrandomly