Signal Processing Image Communication A software framework for model predictive control with GenOpt.docx
- 文档编号:30530872
- 上传时间:2023-08-16
- 格式:DOCX
- 页数:20
- 大小:39.72KB
Signal Processing Image Communication A software framework for model predictive control with GenOpt.docx
《Signal Processing Image Communication A software framework for model predictive control with GenOpt.docx》由会员分享,可在线阅读,更多相关《Signal Processing Image Communication A software framework for model predictive control with GenOpt.docx(20页珍藏版)》请在冰豆网上搜索。
SignalProcessingImageCommunicationAsoftwareframeworkformodelpredictivecontrolwithGenOpt
ThevulnerabilitiesintheCommunication(TCP/IP)protocolstackandtheavailabilityofmoresophisticatedattacktoolsbreedinmoreandmorenetworkhackerstoattackthenetworkintentionallyorunintentionally,leadingtoDistributedDenialofService(DDoS)attack.TheDDoSattackscouldbedetectedusingtheexistingmachinelearningtechniquessuchasneuralclassifiers.Theseclassifierslackgeneralizationcapabilitieswhichresultinlessperformanceleadingtohighfalsepositives.ThispaperevaluatestheperformanceofacomprehensivesetofmachinelearningalgorithmsforselectingthebaseclassifierusingthepubliclyavailableKDDCupdataset.Basedontheoutcomeoftheexperiments,ResilientBackPropagation(RBP)waschosenasbaseclassifierforourresearch.TheimprovementinperformanceoftheRBPclassifieristhefocusofthispaper.Ourproposedclassificationalgorithm,RBPBoost,isachievedbycombiningensembleofclassifieroutputsandNeymanPearsoncostminimizationstrategy,forfinalclassificationdecision.PubliclyavailabledatasetssuchasKDDCup,DARPA1999,DARPA2000,andCONFICKERwereusedforthesimulationexperiments.RBPBoostwastrainedandtestedwithDARPA,CONFICKER,andourownlabdatasets.DetectionaccuracyandCostpersamplewerethetwometricsevaluatedtoanalyzetheperformanceoftheRBPBoostclassificationalgorithm.Fromthesimulationresults,itisevidentthatRBPBoostalgorithmachieveshighdetectionaccuracy(99.4%)withfewerfalsealarmsandoutperformstheexistingensemblealgorithms.RBPBoostalgorithmoutperformstheexistingalgorithmswithmaximumgainof6.6%andminimumgainof0.8%.
ArticleOutline
1.Introduction
2.Relatedwork
2.1.DDoSattack
2.2.Realtimefeatureextraction
2.3.Machinelearningmethods
2.4.Softcomputingmethods
2.5.Existingtracebackmechanisms
2.6.Ensembleofclassifiers–motivation
3.Proposedsystem
3.1.Datacollection
3.2.Preprocessing
3.3.Proposedclassificationalgorithm(RBPBoost)
3.3.1.Training
3.3.2.Testing
3.3.3.NeymanPearsonapproach–costminimization
3.3.3.1.Calculationoftotalnumberofsamplesforeachclass
3.3.3.2.Calculationoftotalnumberofmisclassifiedsamplesforeachclass
3.3.3.3.Findingoptimumthreshold
3.4.Responsesystem
3.5.ComparisonofRBPBoostwiththeexistingalgorithms
4.Simulationresults
4.1.Experiment1
4.2.Experiment2
4.2.1.Experiment2.1
4.2.2.Experiment2.2
4.2.3.Experiment2.3
4.3.Detectionofnewattacks
5.Conclusion
Acknowledgements
References
Purchase
$41.95
146
AutomaticcomponentprotocoladaptationwiththeCoConut/Jtoolsuite OriginalResearchArticle
FutureGenerationComputerSystems,Volume19,Issue5,July2003,Pages627-639
RalfH.Reussner
Closepreview | Relatedarticles | Relatedreferenceworkarticles
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Abstract
WhileindustrialmiddlewareplatformssuchasCORBA,EJB,or.NETfacilitatethedevelopmentofdistributedapplicationsbyprovidingcertaininfra-structuralservicesrequiredinmanydistributedsystems(suchasnameservices,remotemethodcalls,parametermarshalling,etc.),theseindustrialplatformsfailtosupportthedevelopmentofdistributedsystemswithindependentcomponents.Inparticular,theircomponentmodelsdonotprovidesufficientinformationforcomponentinteroperabilitychecksorautomatedcomponentadaptation.Especiallywhenconsideringthepersonalandinstitutionalseparationbetweencomponentdeveloperandcomponentvendorasoneoftheprerequisitesofanindependentcomponentmarket,findingautomaticallyasmanycomponentinteroperabilityerrorsaspossibleiscrucial.Hence,itisofpracticalconcernthatcomponentinterfacesnotonlymodelthecorrectwayofcallingthesinglemethodsbutalsothevalidsequencesofmethodcalls.Likewise,practiceclearlyshowsthatcomponentreuseusuallyrequirescomponentadaptation.Thisdirectlyshowsthatonlydetectingincompatibilitiesisoflimiteduse,butadvocatesfortechniquesofautomatedcomponentadaptation.
Inthispaperwedescribealgorithmsandtoolsforspecifyingandanalysingcomponentinterfacesinordertocheckinteroperabilityandtogenerateadaptedcomponentinterfacesautomatically.Therefore,weintroducetheconceptofparameterisedcontractsandanewcomponentinterfacemodel.
ArticleOutline
1.Introduction
2.Exampleapplication
3.Componentprotocols
4.Parameterisedcontractsforprotocoladaptation
4.1.Interoperabilitychecks,substitutabilitychecksandclassicalcontractsforsoftwarecomponents
4.2.Parameterisedcontractsasgeneralisationofclassicalcontracts
5.Specifyingcomponentprotocols
5.1.Specification-orientedapproach
5.2.Sourcecode-orientedapproach
6.Relatedwork
7.Conclusion
Acknowledgements
References
Vitae
Purchase
$19.95
147
Versatilesoftwareforsemiautomaticanalysisandprocessingoflaser-inducedplasmaspectra OriginalResearchArticle
SpectrochimicaActaPartB:
AtomicSpectroscopy,Volume60,Issues7-8,31August2005,Pages1202-1210
M.P.Mateo,G.Nicolás,V.Piñón,J.C.Alvarez,A.Ramil,A.Yáñez
Closepreview | Relatedarticles | Relatedreferenceworkarticles
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Abstract
ThepresentarticledescribesthemaincharacteristicsandoperationsofSALIPS(softwarefortheanalysisoflaser-inducedplasmaspectra),acomputerprogramdesignedforuseinSpectroscopy.Duringthelastyearslaser-inducedplasmaspectroscopy(LIPS)hasgrowninpopularityanddifferentapplicationshavebeendevelopedinseveralfields.However,untilnowthereisnosoftwarereportedtoperformtherecognitionoftheelementalcompositionofagenericsamplefromitsLIPspectrum,whichmustbeachievedbyhandinatediouscomparativeprocessofexperimentalpeakswithemissionlinesfromdatabases.Forthisreason,acomputerprogramthatincludesseveraltoolstoprovideasemi-automaticidentificationofthepeaksofaLIPspectrumhasbeendeveloped.Theprogram,writteninMicrosoft®VisualBasic®code,hasauser-friendlygraphicalinterfaceandisaflexibletoolthatenablestohandle,edit,copyandprintaquickpresentationofthedataincludingautomaticallytheidentificationresultsinthegraph.SALIPSalsoprovidessomephysicalpropertiesoftheelementsandincludesalgorithmsforperformingthesimulationofspectra.Thepotentialoftheprogramisillustratedwithsomeexamples.
ArticleOutline
1.Introduction
2.Programperformanceandmodules
2.1.Propertiesandemissionlinesmodule
2.2.Simulatedspectrummodule
2.3.Spectrumanalysismodule
2.3.1.Spectrumoptionsform
2.3.2.Searchoptionsform
2.3.3.Spectrumanalysisscreen
2.3.4.Performanceofspectrumanalysismodule
3.Conclusions
References
Purchase
$31.50
148
Videocompressionwithparallelprocessing OriginalResearchArticle
ParallelComputing,Volume28,Issues7-8,August2002,Pages1039-1078
IshfaqAhmad,YongHe,MingL.Liou
Closepreview | Relatedarticles | Relatedreferenceworkarticles
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Noabstractisavailableforthisarticle.
ArticleOutline
1.Introduction
2.Digitalvideoandcompressiontechnologies
2.1.Fundamentals
2.2.Basiccompressiontechniques
2.3.Videocompressionstandardization
2.4.Implementationstrategies
2.4.1.Hardware-basedapproach
2.4.2.Software-basedapproach
3.Parallelprocessingtechnologies
3.1.Parallelarchitectures
3.1.1.Symmetricmultiprocessors
3.1.2.Massivelyparallelprocessors
3.1.3.Distributedshared-memorymachines
3.1.4.Clustercomputing
3.2.Processorarchitectures
3.3.Softwareforparallelprocessing
4.Parallelprocessingapproachesforvideocompression
4.1.ParallelvideocodingalgorithmsonVLSI
4.1.1.Vectorquantization
4.1.2.Paralleldiscretecosinetransform
4.1.3.Parallelwavelet
4.1.4.Parallelvariablelengthcoding
4.1.5.Motionestimation
4.1.6.Ratecontrol
4.2.Completeencodersimplementedonparallelhardware
4.3.Software-basedencoders
4.3.1.Spatialparallelism
4.3.2.Temporalparallelism
4.3.3.Shared-memoryimplementations
4.3.4.Pipelining
4.3.5.Object-basedparallelization
4.3.6.Exploitingmultithreading
4.3.7.Miscellaneousparallelapproaches
4.4.Exploitingparallelismwithinasingleprocessor
5.Conclusions
References
Purchase
$41.95
Abstract
Drivenbytherapidlyincreasingdemandforaudio-visualapplications,digitalvideocompressiontechnologyhasbecomeamaturefield,offeringseveralavailableproductsbasedonbothhardwareandsoftwareimplementations.Takingadvantageofspatial,temporal,andstatisticalredundanciesinvideodata,avideocompressionsystemaimstomaximizethecompressionratiowhilemaintainingahighpicturequality.Despitethetremendousprogressinthisarea,videocompressionremainsachallengingresearchproblemduetoitscomputationalrequirementsandalsobecauseoftheneedforhigherpicturequalityatlowerdatarates.Designingefficientcodingalgorithmscontinuestobeaprolificareaofresearch.Forcircumventthecomputationalrequirement,researchershasresortedtoparallelprocessingwithavarietyofapproachesusingdedicatedparallelVLSIarchitecturesaswellassoftwareongeneral-purposeavailablemultiprocessorsystems.Despitetheavailabilityoffastsingleprocessors,parallelprocessinghelpstoexploreadvancedalgorithmsandtobuildmoresophisticatedsystems
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- Signal Processing Image Communication software framework for model predictive control with GenOpt
链接地址:https://www.bdocx.com/doc/30530872.html