混沌預測理論及其在VBR視頻業(yè)務中的應用.pdf

1、西南交通大學碩士學位論文混沌預測理論及其在VBR視頻業(yè)務中的應用姓名：李恒超申請學位級別：碩士專業(yè)：計算機應用技術(shù)指導教師：張家樹20040301西南交通大學碩士研究生學位論文第1l頁AbstractWiththedevelopmentofchaostheoryandapplicationtechnologystudyanalysisandpredictionofchaotictimeserieshavebecometheemphasi

2、sofchaoticsignalprecessingresearchdomain，andcansolvealotofnonlinearsignalprocessquestionsinengineeringpractice，whicharedifficulttObedonebylinearsignalprocessingmethodsMultimediatra【mchasbecomethechiefsourceofATMnetworkgr

3、aduallywiththedevelopmentofbroadbandnetworkanddigitalcompressiontechnologyandmostofthemaretransmittedthroughvariablebitrate(VBR)modeRecentresearchesindicatethatVBRvideotra伍cnotonlyhasshorttermdependencehutalsohaslong—ter

4、mdependenceConsequentlytraditionalmodelsofvariablebitratevideotra踴caredeficientindepictingvariablebitratevideotramcbutfractalmodelscandoitwellGemerallyselfsimilarbehaviorsoffractalmeantheexistenceofchaos，therebyanalyzing

5、variablebitratevideotrafficthroughchaostheorywillbelikelytorevealintrinsiclaws，whicharenotfoundedbyvtraditionaIanalysismethodResearchworksfoCUSonpredictionmethodsofchaotictimeseriesanditsapplicationinVBRvideotrafficinthi

6、sdissertation，whichmainlyincludeanalysisandlocalpredictionmethodofchaotictimeseries，nonlinearadaptivepredictionofchaotictimeseries，chaoticcharacteristicsanalysisandchaoticpredictionofVBRvideotraf五cThemainresearchfruitsar

7、easfollows：1、Predictioncapabilityandantinoiseperformanceofthreekindsoflocalzero—orderpredictionmethodsareanalysedcompletelyLocalfirst—orderpredictionofchaotictimeseriesbasedonWienersolutionisstudiedDCTdomainlocalquadrati

8、cpolynomialpredictionispresented，andsimulationresultsindicatethatthismethodnotonlycanpredictsomelowdimensionchaotictimeseriesefficientlybutalsoCanbeimplementedsimply2、Basedontheanalysisofpolynomialnonlinearadaptivepredic

9、tionmethodsexistedalreadyaDCTdomainquadraticpredictorforreal—timepredictionof10W—dimensionchaotictimeseriesisproposedThefeasibilityandpredictionperformanceofchaotictimeserieswithneuraladaptivepredictionmethodwithadaptive