云計(jì)算與物聯(lián)網(wǎng)外文翻譯文獻(xiàn)

文獻(xiàn)信息：文獻(xiàn)標(biāo)題：IntegrationofCloudComputingwithInternetofThings:ChallengesandOpenIssues（云計(jì)算與物聯(lián)網(wǎng)的集成：挑戰(zhàn)與開(kāi)放問(wèn)題）國(guó)外作者：HFAtlam等人文獻(xiàn)出處：《IEEEInternationalConferenceonInternetofThings》,2017字?jǐn)?shù)統(tǒng)計(jì)：英文4176單詞，23870字符；中文7457漢字外文文獻(xiàn)：IntegrationofCloudComputingwithInternetofThings:ChallengesandOpenIssuesAbstractTheInternetofThings(IoT)isbecomingthenextInternet-relatedrevolution.Itallowsbillionsofdevicestobeconnectedandcommunicatewitheachothertoshareinformationthatimprovesthequalityofourdailylives.Ontheotherhand,CloudComputingprovideson-demand,convenientandscalablenetworkaccesswhichmakesitpossibletosharecomputingresources;indeed,this,inturn,enablesdynamicdataintegrationfromvariousdatasources.TherearemanyissuesstandinginthewayofthesuccessfulimplementationofbothCloudandIoT.TheintegrationofCloudComputingwiththeIoTisthemosteffectivewayonwhichtoovercometheseissues.ThevastnumberofresourcesavailableontheCloudcanbeextremelybeneficialfortheIoT,whiletheCloudcangainmorepublicitytoimproveitslimitationswithrealworldobjectsinamoredynamicanddistributedmanner.ThispaperprovidesanoverviewoftheintegrationoftheCloudintotheIoTbyhighlightingtheintegrationbenefitsandimplementationchallenges.DiscussionwillalsofocusonthearchitectureoftheresultantCloud-basedIoTparadigmanditsnewapplicationsscenarios.Finally,openissuesandfutureresearchdirectionsarealsosuggested.Keywords:CloudComputing,InternetofThings,CloudbasedIoT,Integration.I.INTRODUCTIONItisimportanttoexplorethecommonfeaturesofthetechnologiesinvolvedinthefieldofcomputing.Indeed,thisiscertainlythecasewithCloudComputingandtheInternetofThings(IoT)–twoparadigmswhichsharemanycommonfeatures.Theintegrationofthesenumerousconceptsmayfacilitateandimprovethesetechnologies.Cloudcomputinghasalteredthewayinwhichtechnologiescanbeaccessed,managedanddelivered.ItiswidelyagreedthatCloudcomputingcanbeusedforutilityservicesinthefuture.AlthoughmanyconsiderCloudcomputingtobeanewtechnology,ithas,inactualfact,beeninvolvedinandencompassedvarioustechnologiessuchasgrid,utilitycomputingvirtualisation,networkingandsoftwareservices.CloudcomputingprovidesserviceswhichmakeitpossibletosharecomputingresourcesacrosstheInternet.Assuch,itisnotsurprisingthattheoriginsofCloudtechnologieslieingrid,utilitycomputingvirtualisation,networkingandsoftwareservices,aswellasdistributedcomputing,andparallelcomputing.Ontheotherhand,theIoTcanbeconsideredbothadynamicandglobalnetworkedinfrastructurethatmanagesself-configuringobjectsinahighlyintelligentway.TheIoTismovingtowardsaphasewhereallitemsarounduswillbeconnectedtotheInternetandwillhavetheabilitytointeractwithminimumhumaneffort.TheIoTnormallyincludesanumberofobjectswithlimitedstorageandcomputingcapacity.ItcouldwellbesaidthatCloudcomputingandtheIoTwillbethefutureoftheInternetandnext-generationtechnologies.However,Cloudservicesaredependentonserviceproviderswhichareextremelyinteroperable,whileIoTtechnologiesarebasedondiversityratherthaninteroperability.ThispaperprovidesanoverviewoftheintegrationofCloudComputingintotheIoT;thisinvolvesanexaminationofthebenefitsresultingfromtheintegrationprocessandtheimplementationchallengesencountered.Openissuesandresearchdirectionsarealsodiscussed.Theremainderofthepaperisorganisedasfollows:SectionIIprovidesthebasicconceptsofCloudcomputing,IoT,andCloud-basedIoT;SectionIIIdiscussesthebenefitsofintegratingtheIoTintotheCloud;Could-basedIoTArchitectureispresentedinsectionIV;SectionVillustratesdifferentCloud-basedIoTapplicationsscenarios.Followingthis,thechallengesfacingCloud-basedIoTintegrationandopenresearchdirectionsarediscussedinSectionVIandSectionVIIrespectively,beforeSectionVIIIconcludesthepaper.II.BASICCONCEPTSThissectionreviewsthebasicconceptsofCloudComputing,theIoT,andCloud-basedIoT.1.CloudComputingThereexistanumberofproposeddefinitionsforCloudcomputing,althoughthemostwidelyagreeduponseemsbethatputforthbytheNationalInstituteofStandardsandTechnology(NIST).Indeed,theNISThasdefinedCloudcomputingas"amodelforenablingubiquitous,convenient,on-demandnetworkaccesstoasharedpoolofconfigurablecomputingresources(e.g.,networks,servers,storage,applications,andservices)thatcanberapidlyprovisionedandreleasedwithminimalmanagementeffortorserviceproviderinteraction".Asstatedinthisdefinition,Cloudcomputingcomprisesfourtypesofdeploymentmodels,threedifferentservicemodels,andfiveessentialcharacteristics.CloudcomputingdeploymentmodelsaremostcommonlyclassifiedasbelongingtothepublicCloud,whereresourcesaremadeavailabletoconsumersovertheInternet.PublicCloudsaregenerallyownedbyaprofitableorganisation(e.g.AmazonEC2).Conversely,theinfrastructureofaprivateCloudiscommonlyprovidedbyasingleorganisationtoservetheparticularpurposesofitsusers.TheprivateCloudoffersasecureenvironmentandahigherlevelofcontrol(MicrosoftPrivateCloud).HybridCloudsareamixtureofprivateandpublicClouds.Thischoiceisprovidedforconsumersasitmakesitpossibletoovercomesomeofthelimitationsofeachmodel.Incontrast,acommunityCloudisaCloudinfrastructurewhichisdeliveredtoagroupofusersbyanumberoforganisationswhichsharethesameneed.Inordertoallowconsumerstochoosetheservicethatsuitsthem,servicesinCloudcomputingareprovidedatthreedifferentlevels,namely:theSoftwareasaService(SaaS)model,wheresoftwareisdeliveredthroughtheInternettousers(e.g.GoogleApps);thePlatformasaService(PaaS)model,whichoffersahigherlevelofintegratedenvironmentthatcanbuild,test,anddeployspecificsoftware(e.g.MicrosoftAzure);andfinally,withtheInfrastructureasaService(IaaS)model,infrastructuresuchasstorage,hardwareandserversaredeliveredasaservice(e.g.AmazonWebServices).2.InternetofThingsTheIoTrepresentsamodernapproachwhereboundariesbetweenrealanddigitaldomainsareprogressivelyeliminatedbyconsistentlychangingeveryphysicaldevicetoasmartalternativereadytoprovidesmartservices.AllthingsintheIoT(smartdevices,sensors,etc.)havetheirownidentity.Theyarecombinedtoformthecommunicationnetworkandwillbecomeactivelyparticipatingobjects.Theseobjectsincludenotonlydailyusableelectronicdevices,butalsothingslikefood,clothing,materials,parts,andsubassemblies;commoditiesandluxuryitems;monumentsandlandmarks;andvariousformsofcommerceandculture.Inaddition,theseobjectsareabletocreaterequestsandaltertheirstates.Thus,allIoTdevicescanbemonitored,trackedandcounted,whichsignificantlydecreaseswaste,loss,andcost.TheconceptoftheIoTwasfirstmentionedbyKevinAshtonin1999,whenhestatedthat“TheInternetofThingshasthepotentialtochangetheworld,justastheInternetdid.Maybeevenmoreso”.Later,theIoTwasformallypresentedbytheInternationalTelecommunicationUnion(ITU)in2005.AgreatmanydefinitionsoftheIoThavebeenputforthbynumerousorganisationsandresearchers.AccordingtotheITU(2012),theIoTis“aglobalinfrastructurefortheInformationSociety,enablingadvancedservicesbyinterconnecting(physicalandvirtual)thingsbasedon,existingandevolving,interoperableinformationandcommunicationtechnologies”.TheIoTintroducesavarietyofopportunitiesandapplications.However,itfacesmanychallengeswhichcouldpotentiallyhinderitssuccessfulimplementation,suchasdatastorage,heterogeneousresource-constrained,scalability,Things,variablegeospatialdeployment,andenergyefficiency.3.Cloud-BasedInternetofThingsTheIoTandCloudcomputingarebothrapidlydevelopingservices,andhavetheirownuniquecharacteristics.Ontheonehand,theIoTapproachisbasedonsmartdeviceswhichintercommunicateinaglobalnetworkanddynamicinfrastructure.Itenablesubiquitouscomputingscenarios.TheIoTistypicallycharacterisedbywidely-distributeddeviceswithlimitedprocessingcapabilitiesandstorage.Thesedevicesencounterissuesregardingperformance,reliability,privacy,andsecurity.Ontheotherhand,Cloudcomputingcomprisesamassivenetworkwithunlimitedstoragecapabilitiesandcomputationpower.Furthermore,itprovidesaflexible,robustenvironmentwhichallowsfordynamicdataintegrationfromvariousdatasources.CloudcomputinghaspartiallyresolvedmostoftheIoTissues.Indeed,theIoTandCloudaretwocomparativelychallengingtechnologies,andarebeingcombinedinordertochangethecurrentandfutureenvironmentofinternetworkingservices.TheCloud-basedInternetofThingsisaplatformwhichallowsforthesmartusageofapplications,information,andinfrastructureinacost-effectiveway.WhiletheIoTandCloudcomputingaredifferentfromeachother,theirfeaturesarealmostcomplementary,asshowninTABLE1.Thiscomplementarityistheprimaryreasonwhymanyresearchershaveproposedtheirintegration.TABLE1.COMPARISONOFTHEIOTWITHCLOUDCOMPUTINGItemsIoTCloudComputingCharacteristicsIoTispervasive(thingsareeverywhere).Thesearerealworldobjects.Cloudisubiquitous(resourcesareavailablefromeverywhere).Thesearevirtualresources.ProcessingcapabilitiesLimitedcomputationalcapabilities.Virtuallyunlimitedcomputationalcapabilities.StoragecapabilitiesLimitedstorageornostoragecapabilities.Unlimitedstoragecapabilities.ConnectivityItusestheInternetasapointofconvergence.ItusestheInternetforservicedelivery.BigdataItisasourceofbigdata.Itisameansbywhichtomanagebigdata.III.BENEFITSOFINTEGRATINGIOTWITHCLOUDSincetheIoTsuffersfromlimitedcapabilitiesintermsofprocessingpowerandstorage,itmustalsocontendwithissuessuchasperformance,security,privacy,reliability.TheintegrationoftheIoTintotheCloudiscertainlythebestwaytoovercomemostoftheseissues.TheCloudcanevenbenefitfromtheIoTbyexpandingitslimitswithrealworldobjectsinamoredynamicanddistributedway,andprovidingnewservicesforbillionsofdevicesindifferentreallifescenarios.Inaddition,theCloudprovidessimplicityofuseandreducesthecostoftheusageofapplicationsandservicesforend-users.TheCloudalsosimplifiestheflowoftheIoTdatagatheringandprocessing,andprovidesquick,low-costinstallationandintegrationforcomplexdataprocessinganddeployment.ThebenefitsofintegratingIoTintoCloudarediscussedinthissectionasfollows.1.CommunicationApplicationanddatasharingaretwosignificantfeaturesoftheCloud-basedIoTparadigm.UbiquitousapplicationscanbetransmittedthroughtheIoT,whilstautomationcanbeutilisedtofacilitatelow-costdatadistributionandcollection.TheCloudisaneffectiveandeconomicalsolutionwhichcanbeusedtoconnect,manage,andtrackanythingbyusingbuilt-inappsandcustomisedportals.Theavailabilityoffastsystemsfacilitatesdynamicmonitoringandremoteobjectscontrol,aswellasdatareal-timeaccess.Itisworthdeclaringthat,althoughtheCloudcangreatlydevelopandfacilitatetheIoTinterconnection,itstillhasweaknessesincertainareas.Thus,practicalrestrictionscanappearwhenanenormousamountofdataneedstobetransferredfromtheInternettotheCloud.2.StorageAstheIoTcanbeusedonbillionsofdevices,itcomprisesahugenumberofinformationsources,whichgenerateanenormousamountofsemi-structuredornon-structureddata.ThisisknownasBigData,andhasthreecharacteristics:variety(e.g.datatypes),velocity(e.g.datagenerationfrequency),andvolume(e.g.datasize).TheCloudisconsideredtobeoneofthemostcost-effectiveandsuitablesolutionswhenitcomestodealingwiththeenormousamountofdatacreatedbytheIoT.Moreover,itproducesnewchancesfordataintegration,aggregation,andsharingwiththirdparties.3.ProcessingcapabilitiesIoTdevicesarecharacterisedbylimitedprocessingcapabilitieswhichpreventon-siteandcomplexdataprocessing.Instead,gathereddataistransferredtonodesthathavehighcapabilities;indeed,itisherethataggregationandprocessingareaccomplished.However,achievingscalabilityremainsachallengewithoutanappropriateunderlyinginfrastructure.Offeringasolution,theCloudprovidesunlimitedvirtualprocessingcapabilitiesandanon-demandusagemodel.Predictivealgorithmsanddata-drivendecisionsmakingcanbeintegratedintotheIoTinordertoincreaserevenueandreducerisksatalowercost.4.ScopeWithbillionsofuserscommunicatingwithoneanothertogetherandavarietyofinformationbeingcollected,theworldisquicklymovingtowardstheInternetofEverything(IoE)realm-anetworkofnetworkswithbillionsofthingsthatgeneratenewchancesandrisks.TheCloud-basedIoTapproachprovidesnewapplicationsandservicesbasedontheexpansionoftheCloudthroughtheIoTobjects,whichinturnallowstheCloudtoworkwithanumberofnewrealworldscenarios,andleadstotheemergenceofnewservices.5.NewabilitiesTheIoTischaracterisedbytheheterogeneityofitsdevices,protocols,andtechnologies.Hence,reliability,scalability,interoperability,security,availabilityandefficiencycanbeveryhardtoachieve.IntegratingIoTintotheCloudresolvesmostoftheseissues.Itprovidesotherfeaturessuchasease-of-useandease-of-access,withlowdeploymentcosts.6.NewModelsCloud-basedIoTintegrationempowersnewscenariosforsmartobjects,applications,andservices.Someofthenewmodelsarelistedasfollows:?SaaS(SensingasaService),whichallowsaccesstosensordata;?EaaS(EthernetasaService),themainroleofwhichistoprovideubiquitousconnectivitytocontrolremotedevices;?SAaaS(SensingandActuationasaService),whichprovidescontrollogicsautomatically;?IPMaaS(IdentityandPolicyManagementasaService),whichprovidesaccesstopolicyandidentitymanagement;?DBaaS(DatabaseasaService),whichprovidesubiquitousdatabasemanagement;?SEaaS(SensorEventasaService),whichdispatchesmessagingservicesthataregeneratedbysensorevents;?SenaaS(SensorasaService),whichprovidesmanagementforremotesensors;?DaaS(DataasaService),whichprovidesubiquitousaccesstoanytypeofdata.IV.CLOUD-BASEDIOTARCHITECTUREAccordingtoanumberofpreviousstudies,thewell-knownIoTarchitectureistypicallydividedintothreedifferentlayers:application,perceptionandnetworklayer.MostassumethatthenetworklayeristheCloudlayer,whichrealisestheCloud-basedIoTarchitecture,asdepictedinFig.1.Fig.1.Cloud-basedIoTarchitectureTheperceptionlayerisusedtoidentifyobjectsandgatherdata,whichiscollectedfromthesurroundingenvironment.Incontrast,themainobjectiveofthenetworklayeristotransferthecollecteddatatotheInternet/Cloud.Finally,theapplicationlayerprovidestheinterfacetodifferentservices.V.CLOUD-BASEDIOTAPPLICATIONSTheCloud-basedIoTapproachhasintroducedanumberofapplicationsandsmartservices,whichhaveaffectedendusers’dailylives.TABLE2presentsabriefdiscussionofcertainapplicationswhichhavebeenimprovedbytheCloud-basedIoTparadigm.TABLE2.CLOUD-BASEDIOTAPPLICATIONSApplicationFieldDescriptionHealthcareCloud-basedIoThasbroughtmanybenefitsandopportunitiestothefieldofhealthcare.Itcanclearlydevelopandimprovehealthcareservicesandkeepthefieldinnovative(e.g.intelligentdrug/medicinecontrol,hospitalmanagement).SmartCitiesAmiddlewareforfuturesmartcitiescanbeprovidedthroughtheIoT,attainingdatafromsensinginfrastructure,IoTtechnologiesandplacinginformationinaconsistentmanner.Thiswillleadtothegenerationofservicesthatcancommunicatewiththesurroundingenvironments(e.g.Smartstreetlights,Bigbelly,ShotSpotte).SmartHomesAlargenumberofCloud-basedIoTapplicationshaveenabledtheautomationofhomeactivities,wheretheadoptionofvariousembeddeddevicesandCloudcomputinghasempoweredtheautomationofin-houseactivities(e.g.homesecuritycontrol,smartmetering,energysaving).VideosurveillanceByembracingCloud-basedIoT,intelligentvideosurveillancewillmakeitpossibletomanage,storeandprocessvideocontentfromvideosensorseasilyandefficiently;thiswillalsomakeitpossibletoextractinformationfromscenesautomatically.Ithasbecomeoneofthesupremetoolsformanysecurity-relatedapplications(e.g.WirelessCCTVCameras,Movementdetectionsystem).AutomotiveandSmartMobilityTheintegrationofCloudcomputingintoTheGlobalPositioningSystem(GPS)andothertransportationtechnologiesrepresentsapromisingopportunitytosolvemanyoftheexistingchallenges(e.g.trafficstateprediction¬ification,remotevehicles).SmartenergyandsmartgridCloudcomputingandtheIoTcanworktogethereffectivelytoprovideconsumerswithsmartmanagementofenergyconsumption(e.g.smartmeters,smartappliances,renewableenergyresources).SmartlogisticsItallowsfor,andeases,theautomatedmanagementofgoodsflowbetweenproducersandconsumers,whilesimultaneouslyenablingthetrackingofgoodsintransit(e.g.logisticsindustry,trackingshipments).EnvironmentalmonitoringBycombiningtheCloudwiththeIoT,ahigh-speedinformationsystemcanbeprovidedwhichwilllinktheentitythatmonitorswide-areaenvironmentsandsensorsthathavebeenproperlydeployedinthearea(e.g.pollutionsourcemonitoring,waterqualitymonitoring,airqualitymonitoring).VI.CHALLENGESFACINGCLOUD-BASEDIOTINTEGRATIONTherearemanychallengeswhichcouldpotentiallypreventthesuccessfulintegrationoftheCloud-basedIoTparadigm.Thesechallengesinclude:1.SecurityandprivacyCloud-basedIoTmakesitpossibletotransportdatafromtherealworldtotheCloud.Indeed,oneparticularlyimportantissueswhichhasnotyetbeenresolvedishowtoprovideappropriateauthorisationrulesandpolicieswhileensuringthatonlyauthorisedusershaveaccesstothesensitivedata;thisiscrucialwhenitcomestopreservingusers’privacy,andparticularlywhendataintegritymustbeguaranteed.Inaddition,whencriticalIoTapplicationsmoveintotheCloud,issuesarisebecauseofthelackoftrustintheserviceprovider,informationregardingservicelevelagreements(SLAs),andthephysicallocationofdata.Sensitiveinformationleakagecanalsooccurduetothemulti-tenancy.Moreover,publickeycryptographycannotbeappliedtoalllayersbecauseoftheprocessingpowerconstraintsimposedbyIoTobjects.Newchallengesalsorequirespecificattention;forexample,thedistributedsystemisexposedtonumberofpossibleattacks,suchasSQLinjection,sessionriding,cross-sitescripting,andside-channel.Moreover,importantvulnerabilities,includingsessionhijackingandvirtualmachineescapearealsoproblematic.2.HeterogeneityOneparticularlyimportantchallengefacedbytheCloud-basedIoTapproachisrelatedtotheextensiveheterogeneityofdevices,platforms,operatingsystems,andservicesthatexistandmightbeusedfornewordevelopedapplications.Cloudplatformssufferfromheterogeneityissues;forinstance,Cloudservicesgenerallycomewithproprietaryinterfaces,thusallowingforresourceintegrationbasedonspecificproviders.Inaddition,theheterogeneitychallengecanbeexacerbatedwhenend-usersadoptmulti-Cloudapproaches,andthusserviceswilldependonmultipleproviderstoimproveapplicationperformanceandresilience.3.BigdataWithmanypredictingthatBigDatawillreach50billionIoTdevicesby2020,itisimportanttopaymoreattentiontothetransportation,access,storageandprocessingoftheenormousamountofdatawhichwillbeproduced.Indeed,givenrecenttechnologicaldevelopments,itisclearthattheIoTwillbeoneofthecoresourcesofbigdata,andthattheCloudcanfacilitatethestorageofthisdataforalongperiodoftime,inadditiontosubjectingittocomplexanalysis.Handlingthehugeamountofdataproducedisasignificantissue,astheapplication’swholeperformanceisheavilyreliantonthepropertiesofthisdatamanagementservice.FindingaperfectdatamanagementsolutionwhichwillallowtheCloudtomanagemassiveamountsofdataisstillabigissue.Furthermore,dataintegrityisavitalelement,notonlybecauseofitseffectontheservice’squality,butalsobecauseofsecurityandprivacyissues,themajorityofwhichrelatetooutsourceddata.4.PerformanceTransferringthehugeamountofdatacreatedfromIoTdevicestotheCloudrequireshighbandwidth.Asaresult,thekeyissueisobtainingadequatenetworkperformanceinordertotransferdatatoCloudenvironments;indeed,thisisbecausebroadbandgrowthisnotkeepingpacewithstorageandcomputationevolution.Inanumberofscenarios,servicesanddataprovisionshouldbeachievedwithhighreactivity.Thisisbecausetimelinessmightbeaffectedbyunpredictablemattersandreal-timeapplicationsareverysensitivetoperformanceefficiency.5.LegalaspectsLegalaspectshavebeenverysignificantinrecentresearchconcerningcertainapplications.Forinstance,serviceprovidersmustadapttovariousinternationalregulations.Ontheotherhand,usersshouldgivedonationsinordertocontributetodatacollection.6.MonitoringMonitoringisaprimaryactioninCloudComputingwhenitcomestoperformance,managingresources,capacityplanning,security,SLAs,andfortroubleshooting.Asaresult,theCloud-basedIoTapproachinheritsthesamemonitoringdemandsfromtheCloud,althoughtherearestillsomerelatedchallengesthatareimpactedbyvelocity,volume,andvarietycharacteristicsoftheIoT.7.LargescaleTheCloud-basedIoTparadigmmakesitpossibletodesignnewapplicationsthataimtointegrateandanalysedatacomingfromtherealworldintoIoTobjects.Thisrequiresinteractingwithbillionsofdeviceswhicharedistributedthroughoutmanyareas.Thelargescaleoftheresultingsystemsraisesmanynewissuesthataredifficulttoovercome.Forinstance,achievingcomputationalcapabilityandstoragecapacityrequirementsisbecomingdifficult.Moreover,themonitoringprocesshasmadethedistributionoftheIoTdevicesmoredifficult,asIoTdeviceshavetofaceconnectivityissuesandlatencydynamics.VII.OPENISSUESANDRESEARCHDIRECTIONSThissectionwilladdresssomeoftheopenissuesandfutureresearchdirectionsrelatedtoCloud-basedIoT,andwhichstillrequiremoreresearchefforts.Theseissuesinclude:1.StandardisationManystudieshavehighlightedtheissuesoflackofstandards,whichisconsideredcriticalinrelationtotheCloud-basedIoTparadigm.AlthoughanumberofproposedstandardisationshavebeenputforthbythescientificsocietyforthedeploymentofIoTandCloudapproaches,itisobviousthatarchitectures,standardprotocols,andAPIsarerequiredtoallowforinterconnectionbetweenheterogeneoussmartthingsandthegenerationofnewservices,whichmakeuptheCloud-basedIoTparadigm.2.FogComputingFogcomputingisamodelwhichextendsCloudcomputingservicestotheedgeofthenetwork.SimilartotheCloud,Fogsupplycommunicatesapplicationservicestousers.FogcanessentiallybeconsideredanextensionofCloudComputingwhichactsasanintermediatebetweentheedgeofthenetworkandtheCloud;indeed,itworkswithlatency-sensitiveapplicationsthatrequireothernodestosatisfytheirdelayrequirements.Althoughstorage,computing,andnetworkingarethemainresourcesofbothFogandtheCloud,theFoghascertainfeatures,suchaslocationawarenessandedgelocation,thatprovidegeographicaldistribution,andlowlatency;moreover,therearealargenodes;thisisincontrastwiththeCloud,whichissupportedforreal-timeinteractionandmobility.3.CloudCapabilitiesAsinanynetworkedenvironment,securityisconsideredtobeoneofthemainissuesoftheCloud-basedIoTparadigm.TherearemorechancesofattacksonboththeIoTandtheCloudside.IntheIoTcontext,dataintegrity,confidentialityandauthenticitycanbeguaranteedbyencryption.However,insiderattackscannotberesolvedanditisalsohardtousetheIoTondeviceswithlimitedcapabilities.4.SLAenforcementCloud-basedIoTusersneedcreateddatatobeconveyedandprocessedbasedonapplication-dependentlimitations,whichcanbetoughinsomecases.EnsuringaspecificQualityofService(QoS)levelregardingCloudresourcesbydependingonasingleproviderraisesmanyissues.Thus,multipleCloudprovidersmayberequiredtoavoidSLAviolations.However,dynamicallychoosingthemostappropriatemixtureofCloudprovidersstillrepresentsanopenissueduetotime,costs,andheterogeneityofQoSmanagementsupport.5.BigdataIntheprevioussection,wediscussedBigDataasacriticalchallengethatistightlycoupledwiththeCloud-basedIoTparadigm.Althoughanumberofcontributionshavebeenproposed,BigDataisstillconsideredacriticalopenissue,andoneinneedofmoreresearch.TheCloud-basedIoTapproachinvolvesthemanagementandprocessingofhugeamountsofdatastemmingfromvariouslocationsandfromheterogeneoussources;indeed,intheCloud-basedIoT,manyapplicationsneedcomplicatedtaskstobeperformedinreal-time.6.EnergyefficiencyRecentCloud-basedIoTapplicationsincludefrequentdatathatistransmittedfromIoTobjectstotheCloud,whichquicklyconsumesthenodeenergy.Thus,generatingefficientenergywhenitcomestodataprocessingandtransmissionremainsasignificantopenissue.Severaldirectionshavebeensuggestedtoovercomethisissue,suchascompressiontechnologies,efficientdatatransmission;anddatacachingtechniquesforreusingcollecteddatawithtime-insensitiveapplication.7.SecurityandprivacyAlthoughsecurityandprivacyarebothcriticalresearchissueswhichhavereceivedagreatdealofattention,theyarestillopenissueswhichrequiremoreefforts.Indeed,adaptingtodifferentthreatsfro