COSOSIMO Parameter Definitions.docx
《COSOSIMO Parameter Definitions.docx》由会员分享,可在线阅读,更多相关《COSOSIMO Parameter Definitions.docx(33页珍藏版)》请在冰豆网上搜索。
COSOSIMOParameterDefinitions
COSOSIMOParameterDefinitions
JoAnnLane
UniversityofSouthernCalifornia
CenterforSystemsandSoftwareEngineering
jolane@usc.edu
Introduction
TheConstructiveSystem-of-Systems(SoS)IntegrationCostModel(COSOSIMO)isdesignedtoestimatetheeffortassociatedwiththeLeadSystemIntegrator(LSI)activitiestodefinetheSoSarchitecture,identifysourcestoeithersupplyordeveloptherequiredSoScomponentsystems,andeventuallyintegrateandtestthesehighlevelcomponentsystems.Forthepurposesofthiscostmodel,anSoSisdefinedasanevolutionarynet-centricarchitecturethatallowsgeographicallydistributedcomponentsystemstoexchangeinformationandperformtaskswithintheframeworkthattheyarenotcapableofperformingontheirownoutsideoftheframework.ThecomponentsystemsmayoperatewithintheSoSframeworkaswellasoutsideoftheframework,andmaydynamicallycomeandgoasneededoravailable.Inaddition,thecomponentsystemsaretypicallyindependentlydevelopedandmanagedbyorganizations/vendorsotherthantheSoSsponsorsortheLSI.
ResultsofrecentCOSOSIMOworkshopshaveresultedinthedefinitionofthreeCOSOSIMOsub-models:
aplanning/requirementsmanagement/architecture(PRA)sub-model,asourceselectionandsupplieroversight(SS)sub-model,andanSoSintegrationandtesting(I&T)sub-model.Theconceptualeffortprofilesforeachsub-modelareshowninFigure1.
Figure1.ConceptualOverviewofCOSOSIMOSub-Models.
ThistechnicalreportisanupdatetotheCOSOSIMOparameterdefinitionsdatedMarch2006anddescribestheparametersforeachoftheCOSOSIMOsub-models.Theparametersincludeasetofsizedriversthatareusedtocalculateanominaleffortforthesetsofactivitiesassociatedwiththesub-modelandasetofcostdriversthatareusedtoadjustthenominaleffortbasedonrelatedSoSarchitecture,process,andpersonnelcharacteristics.Eachsizedriverdescriptionincludesadefinitionoftheparameteraswellasassociatedcountingrulesandguidanceforassigningcomplexityratings.Eachcostdriverdescriptionincludesadefinitionoftheparameteraswellasguidanceforassigningtheappropriateratingfactor.
Finally,COSOSIMOworkshopfindingsindicatethatsomeoftheSoSLSIactivitiesaresimilartosystemsengineeringactivitiesaddressedbytheConstructiveSystemsEngineeringCostModel(COSYSMO)andhavesimilarsizeandcostdrivers.Therefore,someoftheCOSOSIMOparameterdefinitionsareadaptedfromtheCOSYSMOdefinitionsin[Valerdi2005]andareindicatedbyafootnote.
COSOSIMOPRAParameters
TheLSIPRAactivitiesarethoseassociatedwithSoSconceptdevelopment;requirementsidentification,analysis,andevolution;SoSarchitecturedevelopmentandevolution,aswellasthelongtermplanningforprovidingincrementalSoScapabilitiesinaccordancewiththeSoSsponsor’scostandscheduletargets.
PRASizeDrivers
NumberofSoS-RelatedRequirements:
ThisdriverrepresentsthenumberofrequirementsfortheSoSofinterestattheSoSlevel.Requirementsmaybefunctional,performance,feature,orservice-orientedinnaturedependingonthemethodologyusedforspecification.Theymayalsobedefinedbythecustomerorcontractor.SoSrequirementscantypicallybequantifiedbycountingthenumberofapplicableshalls,wills,shoulds,andmaysintheSoSormarketingspecification.Note:
someworkmayberequiredtodecomposerequirementstoaconsistentlevelsothattheymaybecountedaccuratelyfortheappropriateSoS-of-interest.Table1containsthecomplexitydefinitionsfortheSoS-relatedrequirements.
SoS相关需求数:
这个因子表现了在该SoS层级上的相关SoS需求的数量。
需求可以是面向功能、性能、特性或服务,实际取决于系统规格的具体说明方式。
SoS需求一般可通过计算出SoS规格或营销规范说明书中类似于“应(达到)”或“要(达到)”的词句数量来确定。
注意:
有时候可能需要把需求分解成相同层级的多个小需求,以便更加精确的统计其数量。
Table1.SoS-RelatedRequirementsComplexityRatings.
Easy
Nominal
Difficult
Simpletoimplement
Familiar
Complextoimplementorengineer
Traceabletosource
Canbetracedtosourcewithsomeeffort
Hardtotracetosource
Littlerequirementsoverlap
Someoverlap
Highdegreeofrequirementsoverlap
NumberofSoSInterfaceProtocols:
Thenumberofdistinctnet-centricinterfaceprotocolstobeprovided/supportedbytheSoSframework.Note:
ThisdoesNOTincludeinterfacesinternaltotheSoScomponentsystems,butitdoesincludeinterfacesexternaltotheSoSandbetweentheSoScomponentsystems.Alsonotethatthisisnotacountoftotalinterfaces(inmanySoSs,thetotalnumberofinterfacesmaybeverydynamicascomponentsystemscomeandgointheSoSenvironment—inaddition,theremaybemultipleinstancesofagiventypeofcomponentsystem),butratheracountofdistinctprotocolsattheSoSlevel.Table2containsthecomplexitydefinitionsfortheSoSinterfaceprotocols.
SoS接口协议数:
在SoS框架中需要分别提供或给予支持的网络中心接口协议的数量。
注意:
它并不包括SoS组件系统内部的接口,但它包括了由外部提供给SoS的接口,以及SoS成员系统之间的接口。
还需注意的是,它并不是对所有接口总数的计算(在很多SoS中,其接口的总数可能是不断变化的,这是由于在SoS运行环境中,组件系统可能会有所增减——另外,一个特定类型的组件系统也可能会出现多个实例),而是在SoS层级上对独立接口数量的统计。
Table2.InterfaceProtocolComplexityRatings.
Easy
Nominal
Difficult
Simpleprotocol
Moderatelycomplexprotocol
Highlycomplexornewprotocol(s)
MayalreadybesupportedbyseveralSoScomponentsystems
MayalreadybesupportedbysomeSoScomponentsystems
CurrentlysupportedbyfewifanySoScomponentsystems
Uncoupled
Looselycoupled
Highlycoupled
Wellunderstood
Predictablebehavior
Noteasilypredicable
PRACostDrivers
RequirementsUnderstanding1:
ThiscostdriverratesthelevelofunderstandingoftheSoSrequirementsbyalloftheSoSstakeholdersincludingtheSoScustomersandsponsors,SoSPRAteammembers,componentsystemowners,users,etc.Primarysourcesofaddedsystemsengineeringeffortareunprecedentedcapabilities,unfamiliardomains,orcapabilitieswhoserequirementsareemergentwithuse.Table3definesthevariousratingvaluesfortherequirementsunderstandingcostdriver.(Note:
TheseratingdefinitionsarethesameastheonesfortheSSandI&Trequirementsunderstandingcostdrivers,butshouldbeevaluatedintermsofthelevelofunderstandingbetweenthealloftheSoSstakeholderswithemphasisontheSoSPRAteammembers.)
需求理解度:
这个因子统计出了所有SoS利益相关者们(包括SoS客户与赞助者、SoSPRA团队成员、各组件系统的所有者、用户等等)对SoS需求的理解程度。
导致额外的系统工程工作的首要因素是前所未有的新技能、不熟悉的领域,以及需要紧急投入应用的技能。
Table3.PRARequirementsUnderstandingRatings.
Verylow
Low
Nominal
High
VeryHigh
Poor:
emergentrequirementsorunprecedentedcapabilities
Minimal:
manyundefinedareas
Reasonable:
someundefinedareas
Strong:
fewundefinedareas
Fullunderstandingofrequirements,familiarcapabilities
LevelofServiceRequirements1:
ThiscostdriverratesthedifficultyandcriticalityofsatisfyingtheensembleoflevelofservicerequirementsorKeyPerformanceParameters(KPPs),suchassecurity,safety,transactionspeed,communicationlatency,interoperability,flexibility/adaptability,andreliability.Table4definesthevariousratingvaluesforthelevelofservicerequirementscostdriver.(Note:
TheseratingdefinitionsarethesameastheonesfortheSSandI&Tlevelofservicerequirementscostdrivers,butshouldbeevaluatedintermsoftheirimpactstothePRAactivities.)
服务需求水平:
这个因子统计出了满足全层级服务需求或者关键性能参数(KPP)的困难与临界点。
例如:
保密性、安全性、业务速度、交流速度、协同性、灵活性/适应性以及可靠性。
Table4.PRALevelofServiceRequirementsRatings.
Aspect
Verylow
Low
Nominal
High
VeryHigh
Difficulty
Simple;singledominantKPP
Low,somecouplingamongKPPs
Moderatelycomplex,coupledKPPs
Difficult,coupledKPPsorsomeconflictsbetweenKPPsmayberequiringtradeoffs
Verycomplex,tightlycoupledKPPsorsignificantconflictsbetweenKPPsrequiringtradeoffs
Criticality
Slightinconvenience
Easilyrecoverablelosses
Someloss
Highfinancialloss
Risktohumanlife
SoSStakeholderTeamCohesion1:
Representsamulti-attributeparameterwhichincludesleadership,sharedvision,diversityofstakeholders,approvalcycles,groupdynamics,IntegratedProductTeam(IPT)framework,teamdynamics,trust,andamountofchangeinresponsibilities.Itfurtherrepresentstheheterogeneityinstakeholdercommunityoftheendusers,customers,implementers,anddevelopmentteam.Table5definesthevariousratingvaluesfortheSoSstakeholderteamcohesioncostdriver.(Note:
TheseratingdefinitionsarethesameastheonesfortheSSandI&Tteamcohesioncostdrivers,butshouldbeevaluatedintermsofpotentialimpactstothePRAactivities.)
SoS利益相关者的团队凝聚力:
表现为一个多重属性的参数,包括领导能力、共同愿景、利益相关者的差异性、审批周期、群体动力、集成产品开发团队(IPT)的构架、团队动力、信赖度和岗位职责的变动量。
它进一步体现了利益相关者们比如终端用户、客户、实施者和开发团队之间的异质性。
Table5.SoSStakeholderTeamCohesionRatings.
Aspect
VeryLow
Low
Nominal
High
VeryHigh
Culture
Stakeholderswithdiverseexpertise,tasknature,language,culture,infrastructure
Highlyheterogeneousstakeholdercommunities
Heterogeneousstakeholdercommunity
Somesimilaritiesinlanguageandculture
Sharedprojectculture
Strongteamcohesionandprojectculture
Multiplesimilaritiesinlanguageandexpertise
Virtuallyhomogeneousstakeholdercommunities
Institutionalizedprojectculture
Compatibility
Highlyconflictingorganizationalobjectives
Convergingorganizationalobjectives
Compatibleorganizationalobjectives
Clearrolesandresponsibilities
Strongmutualadvantagetocollaboration
Familiarityandtrust
Lackoftrust
Willingtocollaborate,littleexperience
Somefamiliarityandtrust
Extensivesuccessfulcollaboration
Veryhighleveloffamiliarityandtrust
LSIPRATeamCapability:
RepresentstheanticipatedlevelofPRAteamcooperationandcohesion,personnelcapabilityandcontinuity,aswellasPRApersonnelexperiencewiththerelevantdomains,applications,language,andtoolsforSoSLSIpersonnelworkingonth