單片機類畢業(yè)論文英文翻譯

英文資料翻譯AMODELINGBASEDMETHODOLOGYFOREVALUATINGTHEPERFORMANCEOFAREALTIMEEMBEDDEDCONTROLSYSTEMKLEMENPERKO,REMYKOCIK,REDHAHAMOUCHE,ANDREJTROSTABSTRACTTHISPAPERPRESENTSAMODELLINGBASEDMETHODOLOGYFOREMBEDDEDCONTROLSYSTEMECSDESIGNHERE,INSTEADOFDEVELOPINGANEWMETHODOLOGYFORECSDESIGN,WEPROPOSETOUPGRADEANEXISTINGONEBYBRIDGINGITWITHAMETHODOLOGYUSEDINOTHERAREASOFEMBEDDEDSYSTEMSDESIGNWECREATEDATRANSFORMATIONBRIDGEBETWEENTHECONTROLSCHEDULINGANDTHEHARDWARE/SOFTWAREHW/SWCODESIGNTOOLSBYDEFININGTHISBRIDGE,WEALLOWFORANAUTOMATICMODELTRANSFORMATIONASARESULT,WEOBTAINMOREACCURATETIMINGBEHAVIOURSIMULATIONS,CONSIDERINGNOTONLYTHEREALTIMESOFTWARE,BUTALSOTHEHARDWAREARCHITECTURESIMPACTONTHECONTROLPERFORMANCEWESHOWANEXAMPLEWITHDIFFERENTMODELEVALUATIONRESULTSCOMPAREDTOREALIMPLEMENTATIONMEASUREMENTS,WHICHCLEARLYDEMONSTRATESTHEBENEFITSOFOURAPPROACH2011ELSEVIERBVALLRIGHTSRESERVEDKEYWORDSMODELING,MODELTRANSFORMATIONS,EMBEDDEDCONTROLSYSTEMSDESIGN,REALTIMESYSTEMS1INTRODUCTIONEMBEDDEDCONTROLSYSTEMSECSSAREUBIQUITOUSNOWADAYSTHEYAREUSEDINABROADSPECTRUMOFAPPLICATIONS,FROMSIMPLETEMPERATURECONTROLINHOUSEHOLDAPPLIANCESTOCOMPLEXANDSAFETYCRITICALAUTOMOTIVEBRAKESYSTEMSORAIRCRAFTFLIGHTCONTROLSYSTEMSDIFFERENTAPPLICATIONSHAVEDIFFERENTDEMANDSWITHREGARDSTOTHEREALTIMEEXECUTION,CONTROLPERFORMANCE,ENERGYCONSUMPTION,PRICE,ETC,OFTHEECSBEINGUSEDMODERNTECHNOLOGIESFORHARDWAREHWANDSOFTWARESWDESIGNPROVIDEAVARIETYOFPOSSIBILITIESFORDESIGNINGECSSEG,DISTRIBUTEDANDNETWORKEDHW,MULTIPROCESSORSYSTEMS,AVARIETYOFSWCONTROLALGORITHMSANDREALTIMEOPERATINGSYSTEMSRTOSS,ETC1ITISCOMMONLYACKNOWLEDGEDTHATTHEDESIGNINGANDVERIFYINGOFRELIABLEANDEFFICIENTECSSFORAPARTICULARAPPLICATIONARECHALLENGINGTASKS11TRADITIONALCONTROLSYSTEMDESIGNTHEAIMOFDESIGNINGANECSISTOBUILDACOMPUTINGSYSTEMTHATISABLETOCONTROLTHEBEHAVIOROFAPHYSICALSYSTEM,EG,APLANTSUCHAPLANTISMADEUPOFINTERCONNECTEDMECHANICAL,ELECTRICALAND/ORCHEMICALELEMENTSATYPICALECSCONSISTSOFELECTRONICSENSORSFORDATAACQUISITIONFROMTHEPLANT,ACOMPUTINGSYSTEMFORPROCESSINGTHECONTROLALGORITHM,ANDELECTRONICACTUATORSTODRIVETHEPLANTTHEECSDESIGNPROCESSINVOLVESDIFFERENTACTORSANDAREASOFEXPERTISECONTROLTHEORY,SIGNALPROCESSING,REALTIMESWANDHWENGINEERSEACHOFTHESEENGINEERSISFAMILIARWITHTHEIROWNMODELINGLANGUAGES,MODELS,DESIGNTOOLS,ETCTHISHETEROGENEITYINTRODUCESCUTSINTHEDESIGNPROCESSMODELTRANSFORMATIONSARENEEDEDBETWEENEACHDESIGNSTEPHOWEVER,THEYAREOFTENCARRIEDOUTMANUALLYAND,ASARESULT,AREPRONETOMISTAKESANDSUBJECTTOINTERPRETATION,WHICHOFCOURSEDEPENDSONTHESKILLOFTHEDESIGNERTHETRADITIONALFORMOFECSDESIGNISPERFORMEDINTWOSEPARATEDDOMAINSTHECONTROLSWDOMAINANDTHEHWDOMAINUSINGSPECIFICDESIGNTOOLSANDTHEIRRESPECTIVESYSTEMMODELSINTHEFIRSTDOMAIN,CONTROLENGINEERSDEFINETHECONTROLLAWSANDTHESWENGINEERSWRITETHECODETHATEXECUTESTHEOPERATIONSREQUIREDBYTHECONTROLLAWSASOCALLEDCONTROLSCHEDULINGCODESIGNISPERFORMEDDECISIONSMADEINTHEREALTIMERTSOFTWAREDESIGNAFFECTTHECONTROLDESIGN,ANDVICEVERSAFORINSTANCE,DIFFERENTSWSCHEDULINGPOLICIESHAVEDIFFERENTIMPACTSONTHELATENCYDISTRIBUTIONSINTHECONTROLLOOPSAND,CONSEQUENTLY,ONTHEIRPERFORMANCEALSO,THECONTROLLOOPPERFORMANCEDIRECTLYAFFECTSBYCONSTRAININGTHESWEXECUTIONPARAMETERSIE,SAMPLINGPERIODS,TASKEXECUTIONJITTER,ETCINTHESECONDDOMAINTHEHWENGINEERSDESIGNANHWPLATFORMTHATWILLEXECUTETHECONTROLSWTHECONNECTIONSOFALLTHESENSORSANDACTUATORSTOTHEPLATFORMAREMADEVIATHEAVAILABLECOMMUNICATIONCHANNELSHOWEVER,BECAUSETHEHWPLATFORMISDESIGNEDSEPARATELY,CONTROLENGINEERSCANNOTESTIMATEITSIMPACTONTHECONTROLLOOPPERFORMANCEFORINSTANCE,THEDATAFROMSENSORSANDTOACTUATORSCANPASSTHROUGHONEORMORECOMMUNICATIONCHANNELSAHWENGINEERCAN,INGENERAL,CHOOSEFROMAMONGAVARIETYOFCOMMUNICATIONPROTOCOLS,ANDEACHTYPEINTRODUCESDIFFERENTLATENCIESANDJITTER,WHICHTHEREFOREAFFECTSTHESWEXECUTIONTHECONTROLENGINEERCANNOT,HOWEVER,EVALUATETHEEFFECTOFTHESELATENCIESBEFORETHESYSTEMISACTUALLYIMPLEMENTEDHENCE,THEDESIREDPERFORMANCEOFTHESYSTEMMAYNOTBEACHIEVED,ANDITISNECESSARYTOCHANGEANDTUNETHECONTROLLAWSCALIBRATIONPHASEINORDERTOCOMPENSATEFORTHEIMPACTOFTHESECOMMUNICATIONANDEXECUTIONDELAYSTHEFACTTHATTHECALIBRATIONHASTOBEPERFORMEDONANACTUALPLANTCANBEVERYEXPENSIVEANDTIMECONSUMING,ESPECIALLYWHENTHEDESIREDPERFORMANCECANNOTBEACHIEVEDUSINGTHECURRENTHWPLATFORMANDAREDESIGNISREQUIREDANOTHERSHORTCOMINGOFTRADITIONALECSDESIGNISTHEINABILITYOFCONTROLANDSWENGINEERSTOEXPLOITSOMEOFTHEADVANTAGESOFFEREDBYMODERNHWTECHNOLOGIESFORINSTANCE,CONTROLLOOPSRUNNINGINPARALLEL,INSTEADOFTHETRADITIONALSEQUENTIALEXECUTION,COULDGIVEBETTERPERFORMANCEPARALLELEXECUTIONCANBEACHIEVEDWITHTHEUSEOFMULTIPROCESSORORDISTRIBUTEDPLATFORMSMODERNECSDESIGNTECHNIQUESRELYHEAVILYONSYSTEMMODELING,WHICHPROVIDESAMEANSTOEXAMINEHOWVARIOUSCOMPONENTSWORKTOGETHERANDTOESTIMATETHEIMPACTOFTHEECSSIMPLEMENTATIONONCONTROLPERFORMANCEBEFOREITISACTUALLYIMPLEMENTEDTHISMAKESITPOSSIBLETOCORRECTTHEINITIALCONTROLLAWSINORDERTOCOMPENSATEFORTHEIMPLEMENTATIONIMPACTSEARLYINTHEDESIGNCYCLEANOTHERIMPORTANTASPECTOFMODELINGISTHEABILITYTOEXPLOREDIFFERENTPOSSIBLESYSTEMIMPLEMENTATIONSDESIGNSPACEEXPLORATIONAPPROPRIATEMODELINGCANSIGNIFICANTLYSHORTENTHEDESIGNCYCLEOFANECS2TOOVERCOMETHEPROBLEMSINTRODUCEDBYTHEHETEROGENEITYOFDESIGNMODELSANDTOOLS,DIFFERENTMETHODOLOGIESANDTOOLSWEREDEVELOPED3THESEMETHODOLOGIESUSUALLYPROVIDEAMEANSTOCREATEAUNIFORMECSMODEL,SIMULATEANDEVALUATEITSBEHAVIOR,FORMALLYTRANSFORMITTOWARDSANIMPLEMENTATION,ETC12PROPOSEDCONTROLSYSTEMDESIGNTOIMPROVEANDACCELERATETHETRADITIONALECSDESIGNWEPROPOSETHEMERGINGOFTHESESEPARATEDDOMAINSONTHEBASISOFTHISMERGING,ALLTHEACTORSINTHEDESIGNPROCESSCOULDBETTERCOLLABORATEANDEXCHANGETHEIRDATADURINGTHEDESIGNPROCESS,THEYCOULDDOAMORETHOROUGHDESIGNSPACEEXPLORATIONANDTHEDESIGNCYCLECOULDBEMADESIGNIFICANTLYSHORTERINSTEADOFDEVELOPINGANEWMETHODOLOGYFORECSDESIGN,WEPROPOSETOUPGRADETHETRADITIONALSWBASEDCONTROLSYSTEMDESIGNAPPROACHWITHEFFICIENTMODELINGANDDESIGNOFTHEHWPLATFORMSRECENTLY,SEVERALMETHODOLOGIESHAVEBEENDEVELOPEDTHATCONCERNHW/SWCODESIGNTHESEMETHODOLOGIESENABLETHEEFFICIENTDESIGNOFSWANDHWONEMBEDDEDSYSTEMSINTERMSOFSWEXECUTIONSPEED,HWRESOURCESUSAGE,SYSTEMFLEXIBILITY,FUTUREUPGRADEABILITY,FINALDESIGNCOSTS,ETCWEPROPOSECREATINGAFORMALBRIDGEBETWEENTHEEXISTINGTOOLSFORCONTROLSCHEDULINGCODESIGNANDHW/SWCODESIGNTHISBRIDGEMAKESPOSSIBLEMODELTRANSFORMATIONSANDTHEEXCHANGEOFSIMULATIONRESULTSBETWEENTOOLSFORCONTROLSCHEDULINGCODESIGNANDHW/SWCODESIGNTHEBRIDGEISBASEDONAFORMALTRANSFORMATIONOFMODELSBETWEENDIFFERENTDESIGNTOOLSOURFOUNDATIONFORTHECONTROLSCHEDULINGCODESIGNMETHODOLOGYISWORKPRESENTEDIN4ANDITSASSOCIATEDTOOL,MODEST,WHICHISPRESENTEDIN5FORTHEPURPOSEOFHW/SWCODESIGNWEHAVESELECTEDTHEMETHODOLOGYPRESENTEDIN6WITHITSASSOCIATEDABSTRACTSYSTEMMODELINGTOOL,ASYMOD,WHICHISPRESENTEDIN7WITHTHEBRIDGEWEAREABLETOOBTAINMOREACCURATECONTROLPERFORMANCEEVALUATIONSCONSIDERINGARCHITECTURALDETAILSANDEVENTHEPOSSIBILITYTOSTUDYMIXEDHW/SWIMPLEMENTATIONSOFTHECONTROLSYSTEMEVALUATINGTHEIMPACTOFIMPLEMENTATIONINTHEEARLYDESIGNSTAGESREDUCESTHENUMBEROFDESIGNLIFECYCLEITERATIONSANDSHORTENSTHETIMENEEDEDFORAFINALCALIBRATIONOFTHECONTROLLAWSINTHENEXTSECTIONWEPRESENTTHERELATEDMETHODOLOGIES,FOLLOWEDBYSHORTDESCRIPTIONSOFTHEMODESTANDASYMODTOOLSANDTHEIRMETAMODELSINSECTION3WEDESCRIBETHEFORMALRULESFORMODELTRANSFORMATIONANDTHEIMPLEMENTATIONOFTHEBRIDGEINSECTION4,TWOEXAMPLESOFANEMBEDDEDCONTROLLERAREPRESENTEDBYCOMPARINGSIMULATIONRESULTSTOMEASUREMENTSONAREALIMPLEMENTEDSYSTEM,WESHOWTHEBENEFITSOFOURAPPROACHFINALLY,THEPAPERISCONCLUDEDINSECTION513RELATEDMETHODOLOGIESANDTOOLSTHEINCREASINGNEEDTOOPTIMIZEECSSINTERMSOFTHEIRCONTROLPERFORMANCE,RTCONSTRAINTSANDCOSTEFFICIENCYHASLEDTOLIMITEDCOMPUTATIONALRESOURCESCOMBINEDWITHTHEIREFFICIENTEXPLOITATIONANDHAS,ASACONSEQUENCE,ENCOURAGEDTHEEMERGENCEOFNEWRESEARCHAREASDOMAINSPECIFICTOOLSFORCONTROLSCHEDULINGCODESIGNHAVEBEENDEVELOPEDRECENTLYTHESETOOLSSUPPORTIMPLEMENTATIONMODELINGANDANALYSISINTERMSOFCONTROLPERFORMANCESEVERALOFTHETOOLSAREBASEDONMATLAB,WHICHISTRADITIONALLYUSEDBYCONTROLENGINEERSFORTHEDESIGNOFCONTROLLAWSTHEAIDA8TOOLSETISAMODELBASEDENVIRONMENTFORTHEDESIGNANDANALYSISOFCONTROLSYSTEMS,USEDEITHERINSTANDALONEFORMORWITHMATLABTHETOOLSETSUPPORTSTHEMODELINGOFCONTROLFUNCTIONEXECUTIONONDISTRIBUTEDHWCOMPONENTSCONTAININGMULTIPROCESSORSANDCOMMUNICATIONSLINKSTHEEFFECTSOFTHECONTROLALGORITHMSIMPLEMENTATIONONCONTROLPERFORMANCECANBEANALYZEDJITTERBUG9ISAMATLABBASEDANALYSISTOOLFORCOMPUTINGAQUADRATICPERFORMANCECRITERIONINLINEARCONTROLSYSTEMSUNDERVARIOUSTIMINGCONDITIONSUSINGJITTERBUG,THESENSITIVITYOFCONTROLSYSTEMSTODELAYS,JITTERANDOTHERINTERFERENCESCANBESTUDIEDTHEEFFECTSOFDIFFERENTSWIMPLEMENTATIONSONCONTROLPERFORMANCECANBEANALYZEDTRUETIME10ISASIMULATORINMATLAB/SIMULINKDESIGNEDFORTHECOSIMULATIONOFTHEDISTRIBUTEDCONTROLLERSTASKEXECUTIONONSEVERALRTKERNELS,NETWORKTRANSMISSIONS,ANDCONTINUOUSPLANTDYNAMICSITPROVIDESACONTROLPERFORMANCEANALYSISOFDISTRIBUTEDRTCOMPUTERBASEDCONTROLSYSTEMS,CONSIDERINGTHEEFFECTSOFPROCESSORSANDNETWORKSCHEDULING,TASKATTRIBUTES,THEIRDATADEPENDENCY,ETCTRUETIMEANDJITTERBUGCANBEUSEDTOGETHERTOEVALUATETHEPERFORMANCEOFVARIOUSCONTROLLOOPIMPLEMENTATIONS11RECENTLYANESMOL12TOOLCHAINHASBEENDEVELOPEDITINCORPORATESAPROTOTYPESCHEDULINGTOOLWHICHCALCULATESSCHEDULESFORTIMETRIGGEREDNETWORKSINDISTRIBUTEDEMBEDDEDSYSTEMSTHEESMOLCANBEUSEDTOGETHERWITHTHETRUETIMEINORDERTOASSESPLATFORMEFFECTSTOCOMPUTEDSCHEDULEANDTOCONTROLPERFORMANCETHERESEARCHACTIVITIESFOCUSEDONSOFTWAREDESIGNFORDISTRIBUTEDREALTIMEEMBEDDEDSYSTEMSLEADTODEVELOPMENTSEVERALTOOLSANDLANGUAGESTIMINGDEFINITIONLANGUAGETDLISAHIGHLEVELDESCRIPTIONLANGUAGEFORSPECIFYINGTHEEXPLICITTIMINGREQUIREMENTSOFATIMETRIGGEREDAPPLICATION,WHICHMAYBECONSTRUCTEDOUTOFSEVERALCOMPONENTSITPROMOTESTHEIDEATHATTHEFUNCTIONALANDTEMPORALBEHAVIOROFDEVELOPEDSOFTWARESHOULDBEPLATFORMINDEPENDENTTHISREDUCESCOSTSOFSYSTEMINTEGRATION,VALIDATIONANDMAINTENANCEANAUTOMATICBUSSCHEDULEGENERATIONFORMESSAGESOVERNETWORKTOPOLOGYISPRESENTEDIN13ANAPPROACHTOOPTIMIZESOFTWARECOMPONENTALLOCATIONSYSTEMSONDISTRIBUTEDREALTIMEEMBEDDEDSYSTEMSISEXPLAINEDIN14AUTHORSPROVIDEBINPACKINGALGORITHMFORDEPLOYMENTANDCONFIGURATIONOFCOMPONENTSINORDERTOMEETTHEIRREQUIREDQUALITYOFSERVICEQOSPROPERTIES,SUCHASPREDICTABLELATENCY/JITTER,THROUGHPUTGUARANTEES,SCALABILITYETCCHEDDARTOOLISDESIGNEDFORSOFTWARETASKSCHEDULINGSIMULATIONANDFEASIBILITYANALYSISOFSYSTEMSDESCRIBEDWITHARCHITECTUREANALYSISTHISRAISESTHEEMBEDDEDSYSTEMDESIGNPROCESSFROMACLASSICALCODEORIENTEDDEVELOPMENTTOAMODELDRIVENDEVELOPMENTWITHAUTOMATICCODEGENERATIONTHETOOLRELIESONAPROPOSEDMETAMODELABLETOCAPTURETHETHREEFACETDESCRIPTIONSBYDEFININGAUNIFIEDTERMINOLOGYANDSEMANTICSTOSHARETHEINFORMATIONWITHOUTHAVINGTODUPLICATEITTHUS,ITENSURESCONSISTENCYBETWEENTHEMODELSUSEDINEACHDESIGNSTEPBYCARRYINGOUTMODELTRANSFORMATIONSANDRECONCILIATIONBETWEENTHEFACETSEACHCHANGEINTHEFACETDESCRIPTIONISAUTOMATICALLYRELAYEDTOANOTHERONETHETOOLPROVIDESFACILITIESFORSCHEDULINGANALYSISANDSIMULATIONASWELLASCODEGENERATIONFORMULTIPLETARGETSITINCLUDESANRTBEHAVIORANALYSISMODULETOCHECKTHESCHEDULINGFEASIBILITYRATEMONOTONICRM,DEADLINEMONOTONICDM,EARLIESTDEADLINEFIRSTEDF,ETCITISALSOABLETOSIMULATEANDDISPLAY,ASACHRONOGRAMCHART,THEEXECUTIONOFTASKSTHECODEGENERATIONANDPROFILINGMODULE,WHICHGENERATESTHESOURCECODESFORTHEMULTIPLERTKERNELSRTAI32,DSPBIOS33ANDREALTIMEJAVAANDCAPTURESAREALTEMPORALEXECUTIONTRACE,HASALSOBEENIMPLEMENTEDTHEMODESTTOOLALSOPROVIDESATIMINGMEASUREMODULETHATPERFORMSLATENCY,DELAYANDJITTERMEASUREMENTSONTASKEXECUTIONTHISTEMPORALINFORMATIONCANBEUSEDINHYBRIDSIMULATIONSINORDERTOESTIMATETHELOSSOFCONTROLPERFORMANCE211METAMODELOFTHEMODESTIMPLEMENTATIONFACETTHEBRIDGEBETWEENTHEMODESTANDASYMODTOOLSISBASEDONRULESDEFININGTHETRANSFORMATIONSBETWEENASUBSETOFTHEMODESTMETAMODELIMPLEMENTATIONFACETPARTANDTHEASYMODMETAMODELSINCETHEGOALOFTHEPAPERISTODESCRIBETHEBRIDGE,WEWILLNOTPRESENTDETAILSOFTHEWHOLEMODESTMETAMODELBUTONLYOFTHESUBSETRELEVANTTOTHEIMPLEMENTATIONFACETFORTHESAKEOFSIMPLICITY,WESHALLTERMITTHEMODESTMETAMODELASNIPPETOFTHEMETAMODELISPRESENTEDINFIG2TODEFINETHEMETAMODEL,WEUSEAGRAPHNOTATIONADATAFLOWGRAPHDFGISADIRECTEDGRAPHTHATREPRESENTSTHEDATADEPENDENCIESBETWEENTHENUMBERSOFEDGESREPRESENTTHEDATAPATHSTHEMULTIPERIODDATAFLOWGRAPHMPDFGISADFGEXTENDEDWITHINFORMATIONABOUTTHERTEXECUTIONCONSTRAINTSOFTHEDATAOPERATIONSANMPDFGISATYPEOFDIRECTEDACYCLICGRAPHDAGINTHEMODESTTOOL,AMPDFGMIC,C,TM,RISUSEDTOREPRESENTTHEIMPLEMENTATIONFACETOFEMBEDDEDSYSTEMS,WHEREICICOMP1,ICOMPNISASETOFIMPLEMENTATIONCOMPONENTSICOMPS,CC1,CMISASETOFCOMMUNICATIONSBETWEENTHEIMPLEMENTATIONCOMPONENTS,TMTM1,TMOISASETOFTASKSANDRDEFINESTHETASKSCHEDULERICOMPSREPRESENTTHENODESINTHEMPDFGANDCOMMUNICATIONREPRESENTSTHEEDGESBETWEENTHENODESICOMPISAMATHEMATICALFUNCTIONTHATCOMPUTESTHEOUTPUTSFROMTHEINPUTSICOMPCANBEPRESENTEDASATUPLEICOMPNOK,OL,IO,IP,WHEREOK,OLISALISTOFALLITSOUTPUTSANDIO,IPISALISTOFALLITSINPUTSTHESETSIMANDOMARETHESETSOFALLTHEINPUTSANDALLTHEOUTPUTSOFALLTHEICOMPSINTHEMODELMTHESHAREDVARIABLESAREREPRESENTEDASCOMMUNICATIONCOMMUNICATIONISADATATRANSFERBETWEENTHEOUTPUTANDTHEINPUTOFTWODISTINCTICOMPSACOMMUNICATIONISNOTEDASCNOK,IL,TYPE,WHEREOKOMISTHEPRODUCEROFTHEDATA,ILIMISTHECONSUMEROFTHEDATAANDTYPEDEFINESTHEDATATYPEEACHCOMMUNICATIONCANTRANSFERONLYONEDATATYPE,WHICHCANBECHOSENFROMASETOFPREDEFINEDTYPESCHAR,INTEGER,FLOAT,DOUBLETHETASKSADDRTEXECUTIONCONSTRAINTSFORPERIODICITY,DEADLINESANDPRIORITIESTOASETOFICOMPSTHEEXECUTIONOFTASKSINMODESTISTIMETRIGGERED,WITHEACHTASKDEFININGAPARTITIONOFICOMPSICTASKCANBEEXPRESSEDASTHEQUINTUPLETMNPMN,T0MN,DMN,LMN,SEQMN,WHEREPMNISTHETASKPERIOD,T0MNISTHESTARTTIMEOFFSET,DMNISTHETASKDEADLINE,LMNISTHETASKPRIORITYANDSEQMNICOMPQ,ICOMPRISANORDEREDSEQUENCETHATDEFINESTHEEXECUTIONSEQUENCEOFICOMPSTHATSHOULDBEEXECUTEDWHENTHETASKISCALLEDATARUNTIMETHETASKSCHEDULERISDEFINEDWITHTHEPAIRRSCHM,PREM,WHERESCHMDEFINESTHESCHEDULINGALGORITHMTYPEFROMAPREDEFINEDSETOFALGORITHMSRM,DM,EDFANDTHEBOOLEANPREMDEFINESWHETHERTHETASKEXECUTIONSCANBEPREEMPTED22ASYMODTOOLTHEABSTRACTSYSTEMLEVELMODELLINGASYMODTOOLISAREPRESENTATIVEOFTHEHW/SWCODESIGNTOOLSITISUSEDINTHEEVALUATIONOFEMBEDDEDSYSTEMSDURINGTHEEARLYSTAGESOFTHEDESIGNPROCESSINTERMSOFTHECONSUMPTIONOFMATERIALRESOURCESMEMORY,BUS,PROCESSORS,FPGA,ETCANDTHEDURATIONOFTHERTTASKEXECUTIONSITPROVIDESAUNIFORMMODELLINGFRAMEWORKFORTHEHWANDSWCOMPONENTSDESCRIBEDATAHIGHLEVELOFABSTRACTIONTHEMODELLINGFOLLOWSRULESDEFINEDINTHEASYMODMETAMODELTHEASYMODISCONSTRUCTEDOFTHREEMAJORPARTSTHEFIRSTPARTISAGRAPHICALMODELLINGINTERFACE7BASEDONAGENERICMODELLINGENVIRONMENTGME34GMEISACONFIGURABLETOOLKITFORCREATINGDOMAINSPECIFICMODELLINGANDPROGRAMSYNTHESISENVIRONMENTSINORDERTOCONFIGURETHEGMETOSERVEASPECIFICDOMAIN,ASUITABLEMETAMODELFORTHEABSTRACTHW/SWCODESIGNISDEFINEDTHISSPECIFICMETAMODELISCONFORMANTWITHTHEGMEMETAMODELLINGLANGUAGE,CALLEDMETAGME,ANDISBASEDONUNIFIEDMODELLINGLANGUAGEUMLCLASSDIAGRAMSANDOBJECTCONSTRAINTLANGUAGEOCLCONSTRAINTSADEFINEDMETAMODELCONTAINSALLTHESYNTACTIC,SEMANTIC,ANDPRESENTATIONINFORMATIONREGARDINGTHEASYMODDOMAINIE,WHICHHWANDSWCONCEPTSCANBEUSEDTOCONSTRUCTMODELS,THERELATIONSHIPSTHATMAYEXISTAMONGTHOSECONCEPTS,HOWTHECONCEPTSMAYBEORGANIZEDANDVIEWEDBYTHEMODELLER,ANDTHERULESGOVERNINGTHECONSTRUCTIONOFMODELS35THESECONDPARTOFASYMODISAMODELINTERPRETERTHATPARSESTHEINFORMATIONFROMTHEGRAPHICALMODELINTOATEXTUALDESCRIPTIONTHEOBTAINEDDESCRIPTIONCOMPLIESWITHTHESIMULATIONKERNELTHISKERNELISTHETHIRDPARTOFTHEASYMODTOOLITISBASEDONAMETHODOLOGYPRESENTEDIN6ANDPROVIDESTHEMEANSFORTHETEMPORALSIMULATIONOFDISTRIBUTEDSYSTEMSONTHESYSTEMLEVELTHESIMULATIONOFTHEMODELPROVIDESTEMPORALINFORMATIONABOUTTHESWTASKEXECUTIONS,THEHWRESOURCEUSAGE,THECONTENTIONSONTHECOMMUNICATIONDEVICESANDTHEIRIMPACTONTHETASKEXECUTIONSTHERESULTSARESUMMARIZEDINAHWRESOURCEUTILIZATIONTABLEANDSTATISTICALINFORMATIONABOUTTHEEXECUTIONTIMEDELAYS,JITTERFOREACHTASKBASEDONTHISINFORMATION,THEDESIGNERSCANSEEHOWTHEHWANDSWCOMPONENTSINTERACTANDIDENTIFYEVENTUALBOTTLENECKSINTHESYSTEMTHEASYMODTOOLCANEVALUATESEVERALDIFFERENTHWARCHITECTURESANDSWIMPLEMENTATIONSAGAINSTTHEDESIGNCONSTRAINTSITENABLESQUICKANDEFFICIENTDESIGNSPACEEXPLORATIONINTHEEARLYSTAGESOFTHEDESIGNPROCESSTHETOOLCANBEUSEDASINSTANDALONEFORMORINCORPORATEDINTOTHESYSTEMDESIGNTOOLCHAINBYMEANSOFIMPORTINGANABSTRACTSYSTEMMODELANDEXPORTINGTHEMODELEVALUATIONRESULTSINTHEXMLEXTENSIBLEMARKUPLANGUAGEFORMAT221ASYMODMETAMODELEMBEDDEDSYSTEMSINASYMODAREPRESENTEDINTHREEASPECTSSWFUNCTIONALITY,HWARCHITECTURE,ANDMAPPINGEACHOFTHESEASPECTSISDEFINEDBYTHEASYMODMETAMODELANDEACHASPECTCANBEREPRESENTEDBYASUITABLEGRAPHASIMPLIFIEDMETAMODELISPRESENTEDINFIG3THEELEMENTSVISIBLEINEACHASPECTAREDENOTEDWITHDASHEDLINESANEVENTFLOWGRAPHEFGISADIRECTEDGRAPHTHATREPRESENTSANEVENTDEPENDENCYBETWEENTHEELEMENTS,WHERETHEGRAPHNODESREPRESENTEVENTDEPENDENTELEMENTSANDTHEGRAPHEDGESREPRESENTEVENTPATHSTHENODESCANPRODUCEANDCONSUMEEVENTSANARCHITECTUREGRAPHAGISAGRAPHTHATREPRESENTSABASICARCHITECTURALTOPOLOGYANDWHERETHEGRAPHNODESREPRESENTARCHITECTURALCOMPONENTSANDTHEEDGESREPRESENTCONNECTIONSBETWEENTHEARCHITECTURALCOMPONENTSAMAPPINGGRAPHMGISADIRECTEDGRAPHTHATDEFINESWHICHFUNCTIONALITYWILLBEEXECUTEDONAPARTICULARARCHITECTURALCOMPONENTMODELAINASYMODCANBEPRESENTEDASASETAFG,AG,MG,WHEREFGISANEFGTHATREPRESENTSTHESWFUNCTIONALITYASPECTOFTHEASYMODMODEL,AGISANAGTHATREPRESENTSTHEHWARCHITECTUREASPECTOFTHEASYMODMODELANDMGISANMGTHATREPRESENTSTHEMAPPINGASPECTOFTHEASYMODMODEL222EVENTFLOWGRAPHINASYMODINASYMODTHESWMODELISPRESENTEDWITHINTHEFUNCTIONALITYASPECTWITHANEFGGRAPHTHEEFGFGCANBEEXPRESSEDASATRIPLEFGS,TA,XS,WHERESSA1,SANISASETOFPERIODICEVENTGENERATORSPEG,TATA1,TANISASETOFTASKSANDXSXS1,XSNISASETOFCONNECTIONSTHATDEFINETHEEVENTFLOWTHETASKSANDPEGSREPRESENTTHENODESOFTHEEFGANDTHECONNECTIONSREPRESENTITSEDGESPEGSPERIODICALLYGENERATEEVENTSANDTHEYCANBEEXPRESSEDWITHTHEPAIRSANSTAN,SDAN,WHERESTANDEFINESTHEPERIODOFEVENTSANDSDANISTHEINITIALDELAYOFTHEFIRSTEVENTPEGSCANALSOBEUSEDFORGENERATINGONETIMEEVENTSINTHISCASE,AVERYLONGPERIODSTANSHOULDBESETBEYONDTHETOTALSIMULATIONTIMEANDTHEDELAYSDANISUSEDTODEFINEWHENANEVENTSHOULDBEGENERATEDTHEASYMODTASKSAREEVENTTRIGGEREDTASKSANASYMODTASKISCOMPOSEDOFADESCRIPTIONOFTHEHWRESOURCEUSAGEFORAPARTICULARPIECEOFTHECODEANDRTEXECUTIONCONSTRAINTSFORTHEDEADLINEANDPRIORITYTHETASKCANBEEXPRESSEDWITHATRIPLETANDAN,LAN,BAN,WHEREDANISTHETASKDEADLINE,LANISTHETASKPRIORITYANDTHEBOOLEANBANDEFINESWHETHERTHESCHEDULERSHOULDBECALLEDATTHEENDOFTHETASKEXECUTIONTHISISDISCUSSEDINMOREDETAILINSUBSECTION311WITHAGUIDELINEVTHETASKEXECUTIONCANBEGINIFTHETASKISTRIGGEREDWITHANEVENTANDIFTHESCHEDULER,BASEDONTHESCHEDULINGPOLICYANDTASKRTEXECUT