ExecutiveSummaryThedevelopmentofinformationtechnologyisaccompaniedbythecontinuousemergenceandintegrationofnewtechnologies.Asinformationandcommunicationtechnology(ICT)evolvestowards6G,itfurtherintegrateswithbigdataandartificialintelligencetechnologies,presentingthecharacteristicsofICDT(Information,Communication,BigDataTechnology)integration.TheICDTintegrated6Gsystemwillbecomeanend-to-endinformationprocessingandservicesystem,representinganmobileinformationnetworkthatcombinescommunicationnetworks,sensingnetworks,andcomputingnetworks.Newtechnologiesandfunctionalitieshavebecomecandidatesolutionsfor6Gdesign,validatedatvariouslevels,includingtheoretical,simulation,andprototypestages.Thiswhitepaperclassifiesandanalyzespotential6Gtechnologiesandsolutionsthroughthreeaspects:thenewtrendsinICDTintegration,wirelessairinterfacetechnologiesinICDTintegration,andnetworkaspectsofICDTintegration.Itproposescorrespondingtechnicalsystemsuggestions.ThenewtrendsinICDTintegrationencompasscommunicationandcomputationtechnology,semanticcommunicationtechnology,andAIlargemodeltechnology.Communicationandcomputationintegrationtechnologyisbasedoncoretechnicalcapabilitiessuchasthecoexistenceofcommunicationandcomputationresources,functionalintegration,andserviceintegration,enablingtheevolutionofwirelessaccessnetworkcomputingandtheupgradeofgeneralintelligenceservices.Semanticcommunicationtechnologyfocusesonthemeaningandsignificanceoftransmittedinformation,aimingtoovercomethelimitationsimposedbyShannontheoryoncommunicationsystems.AIlargemodels,asoneofthemostprominentscientificdirections,continuouslyenhancetheintelligentautonomouscapabilitiesofcommunicationnetworks.ICDTintegratedwirelessairinterfacetechnologyincludesintegrationofsensingandcommunications,airinterfaceAI,multipleaccessmethods,encodingandmodulation,massiveMIMO,andnear-fieldtechnology.Novelairinterfacetechnologiesareessentialforconstructinga6Gnetworkwithstrongconnectivity,computingpower,intelligence,andsecurity.Currently,thesetechnologieshavecompletedthedesignoftechnicalandprototypesolutionsandareundergoingtesting.IntermsofthenetworkarchitectureofICDTintegration,service-orientednetworks,nativeAI,digitaltwinnetworks,multidimensionalnetworking,anduser-centricnetworksarekeyresearchdirections.ThenewnetworkarchitecturehelpspropeltheICDTintegrated6Gnetworkstowardsanewgenerationofmobileinformationnetworkswithsensing,computation,andintelligence,providingcomprehensivecoverageacrossair,land,andspace.TheresearchonICDTintegrated6Gtechnologyisgraduallytransitioningfromisolatedtechnicalresearchtowardssystematicconstructionandindustryconsensus.Thisarticleadvocatesforstrengtheninginterdisciplinaryresearchandcollaborationinthefieldofinformationandcommunicationtechnology,drivenbydemand,toacceleratejointtacklingofbasictheoriesandbottlenecktechnicalissuesandfostercollaborativeinnovation.Simultaneously,emphasisshouldbeplacedonthetransformationofinterdisciplinaryinnovationtowardsapplicationandservices,nurturingindustrialecosystemsattheapplicationlayer,andincubatingnewbusinessesandapplications.Finally,furthereffortsshouldbemadetoenhanceinternationalcooperation,withcountriesworkingtogethertosolidifythefoundationof6Gandpromoteitsglobaldevelopment.Contents 1 42.NewTrendsinICDT 62.1IntegratedCommunicationandCo 6 15 15 162.2.3ApplicationsandCh 19 212.3.1CurrentResearchStatusofLarge 222.3.2EmpoweringtheCommunicationInd 232.3.3IntelligentEndogenousCommunication 26 273.ICDT-integratedWirelessAirInt 29 29 29 31 33 33 35 37 37 39 44 44 45 49 49 50 61 61 65 694.ICDT-integrated 73 734.1.1Overviewof6GNetworkArchitecture 73 74 75 75 75 77 78 78 78 80 82 82 83 854.5Multi-dimensionalNe 86 86 87 91 92 92 93 94 94 95 971.IntroductionTherapiddevelopmentofinformationtechforceforsocialprogress.Inrcommunication,AI,bigdata,systemcontrol,andsensICDTIntegrated6GNetwork1.0whitepaperwasofficiallyreleased,regarding6Gasexpandedfromsimpleinfoinformationcomputing,andinformationapplication,providinguserswithmorepowerfulendogenouscapabilitiesiundergoescontinualevolution.TheICDTIntegrated6GNetwork2.0whitepaperwasreleasedattheGlobal6GConfereAdditionally,inMarch2023,theICDTIntegrated6GNetwork3.0whitepOverthepastyear,therehasbeextendingitsscopeandpositiointerest.Theacademicandindustrialcirclesarecoltechnologiesintheoreticalresinitiatives.InJune2023,theInternationalTelecommproposalfortheframeworkanddevelopmentgoals,trends,typical6G.ThereleaseofthisproposalsignifiesconsensusandguidingprinciplesforfuturecommunicationtSeptember2023,theNextGenerationMobileNetworks(NGMN)Allianceaofthisstatementhighlightstheglobalconsensusonindustry-levelguidancetodriveitsinnovationForceintendingtoadvancethecommunicationtechnostandardizing6Gtechnology,drivingicharacteristicsofcross-borderintegrationandmulti-layerbreakthroughs.BuildingonthelatestadvancementtheFuTUREForumhasledthedevelopmentoftheICDTIntegrated6GNetwork4.0ICDT-integratednetworkarchitecture.Italsoprovidesaconctechnicalconcepts,applicationscsupportingitscontin2.NewTrendsinICDTIntegrationTheICDTintegrationisundergoingasignificaintegrationofcommunicationandcomputing,semanticcommunication,anintegrationofAIlargemodwayforafutureofintelligent,efficient,andreliablecommunicatFigure2-1ICDTIntegrationTrends2.1IntegratedCommunicationandCanewtechnologicalfieldthathasemergedduetothediincludingtechnologicalevolutionrequirements,businessdevelopmenttransformationandupgradingofnetworkinfrastructure.technologicalcapabilitiessuchastheiandcomputingresources,functionalintegpowerandtheupgradeofintegratedcommunication,computing,andintelligentresourcesandfactors,providingsolidsupportforthegrowthoftWiththerapidonsetofthedigitalera,characterizedbyconnectivity,andintelsignificantnationalstrategicinitiative.AsacriticalnationalinformationinfrastrucRANspossessalargenumberofbasestations.Theexplorationandutilizationoftheirachievingnetwork-widecomputationalupgrading.TheintegrationandprogressionofFromatechnologicalstandplikecommunication,cloudcominevitabletrendintheevolutionofnext-generationnetworks.ResearchintokeytechnologieslikeAIendogeneityandintegratedsensincommencedfor6G,bringingforthfreshdemanFuturenetworkdesignsmustholisticallyaddressthecommunicationandcomputingtofacilitatetheintegrateddevelopmentoftechnologiesandmaximizetheiradvantFromabusinessexpansionstandpointlocaldataoffloading,anddetermiqualityinspectionandremotecontrol,isfacilitatedbycloud-bintegratingitwithintetailoredsolutions,empoweringthefutureof6G.endof2023,thetotalnumberof5GbasestationsinChinahasreached3.377million,representingtheheaviestinvestmentandthelarThankstoitsuniqueticomputationalpower,formingablueoceanofcomputationalvalue.Sieconomicvalue,yetrequiresinnovativetechnologyforefficientutilization.Likewise,upcoming6Gnetworkswillconstituteavastcomputingpowerresourcepool,contributingsubstantialcomputtargetstypical6Gusecaseslikenetworkintecoexistenceofwirelesscommunicationandcomputing.Byconsolidatingcomputingpowerofbasestations,itenablestheimplementationofAI-drivennetworkoperationandoptimization,effectivelysupportingtheenhancementofheavy-loadcellperformance,intelligentdetectionoffdemands.Furthermore,leveragiidentificationatbasestationsensurestheassusuchaspaymentapplicationswithinhospitalenvironments,andensuresseexperiencesforhospital-ownedmobileapps.Additionally,applicationslikebstationAIoptimizatiscenariosfortheintegrationandcoexistenceofwireleTheintegrationofcomputing,networks,andservicesbasedonbasecapacityexpansionrepresentsakeyaspectoftheintegrationandcoexistenceoflocalcomputingneedsofultra-lowlatencyandhigh-bandwidthapplicationsbymetaverse.Thesecondaspectfocuseson2B,provideploymentwhileeffectivelyreducingtheoverallcostsofindustryapplications.intelligentcollisionpindustryorganizations,haveundertakenresearchontheintegrationofwirelessZTEbeginbyanalyzingtop-levelbusinessoperationsandscenavarioustask/applicationrequirements.Then,theybreakdowntheserequirementsintoandnetworkorchestration,AI/sensing/computingKPIdefinitions,andcapabilityvalidations,andpromotingenterprprimarilyfocusontheadvancementofTheyemphasizetheintegrationofmaturescenarios,includingAIandedgecomputinintothistechnology.Thesecompaniintegrationandcoexistenceofwirelesscommunicationandcomputingmusttransitionfromindependentfunctionalityancoexistenceofcommunicationandcomputing.ThisinvolvesmovingawayFigure2-2ReferenceNetworkArchitectureoftheIntegrationandCoexistenceofWirelessCommunicationandComputingThereferencenetworkarchitectureasshowninFigurlayers:infrastructure,networkfunction,andmanagementandorc1)Theinfrastructurelayterminaldevicestoofferv2)Thenetworkfunctionlayerbuildsuponthenewcomputingresourceefromtheinfrastructurelayer,enhancingtraditionaluserandcontrdelivercomputationalexecutionandcommunicationservices.Thewirelesscomputingpower,computingtaskdemands,jointlyschedulingandclosed-loopmonitoringtoguaranteeQoSforcommunicat3)BuildinguponthetraditionaRAN,themanagementandorchestrorchestrationofwirelesscomputingresources,jointmanagementandorchestrationofcommunicationandcomputing,andmanagemcomputingservices.Themanagementandorchestrationlayerprovidesbothlonetworkbraintoincorporatewireless-sidecomputingpowerintothebroadernetwork,therebydeliveringintegratedcommunicationandcomputingservemphasizestheintegrationofwirelesscommunicationandcomputingresources,aspects:thejointoptimizationofcomputingandcommunicationresourceprovisionofcomputiResearchReporton6GISACSystemDesignwasreleased.Itanalyzedpotentialcomputingkeyperformanceinincludingcollaborationbetweensensingandcomputingnodes,high-performanceterminalcomputingservices,communication-computingcollaboration,WhendiscussingthenexindustryorganizationintheUnitedStates,communicationandcomputing.ItinvolvesthetransmissionandprocesintegrationofcommunicationandcomputingthroughvtransmissionandprocessingforavarietyofnewimprovingnetworkefficiencyandreliaprocessingofnetworkdatathroughtheuseofAIandmachinelearningCloudEvolution,outliningresearchdirectionsincludingjointorchestrationofcommunicationandcomputing,dynamicorchesttechnologiesandprotocolstodrivetheadoptionofnetworkvirtualization,efficiencyindatatransmissionandprocessing.totherequirementsofdiversenewapplications,ultimatelyenhancingtheuserserviceThe6GflagshipresearchorganizationHexa-X,undertheEuropeanCommission,hasoutlinedanddelvedintotheprojectedapplicationsanddemandsfortheyemappingintonewdigitalservices.Additionally,theyhavepublishedtheJointDesignSchemeofAI-DrivenCommunicationandComputingtolaythegroundworkthroughearly-stageresearchonsystemarchitecture,technologies,followedbyadvancingtheproposedschemesintostandardizationandcooAccordingtoindustryresearch,theintegrationandcoexistenceofwirelessFigure2-3ThreeMajorTechnicalDirectionsfortheIntegrationandCoexistenceofWirelessCommunicationandComputing(I)Regardingresources,thekeychubiquitousRANinfrastructuretoadequatelyUponintegratingcomputationaltasksintothecomputinghardware.Furthermore,theunifiedhardwsupportwirelessnetworksanddchallengesinreal-timesharing,allocation,andmanagementofhecomputingresources.Theintegrationofcommunicatisupportreal-timeallocationofcomputingresourcesencompassesthreekeyschedulingdemands;2)multidimincorporatingfactorssuchplatformcapableofevolvingsuchasspace-timevariatiodynamicwirelessnetworkenvironments,alongsideconstraintsoncommucomputingresources.ThecentralissuepertainscommunicationandcomputingserviceswhileensuringtheirrelTheconceptoffunctionalintegrationandcoexistencommunicationprotocolsastcommunicationprocesses,enablingreal-timeprecisecontroloverbothcomputingservicesarebeingimpleAIcapabilitiesandcomputingservicesareplcontrolplaneanduserplaneofRANs.ResearchEnhancingthecontrolplaneprimarilyaimstofacilicomputingresourcesandtaskswithintheRAN,ensuringclosed-loopqualityofserviceforcomputingtasks.Keytechnologiesencompass(1)designofwirelessfunctions,interfaces,andprocresourcesundermulti-dimensionalobjectivesincludingendcomputingqualityofservice,andenergyefficiency,alongsidevRegardinguserplaneenhancement,theenclosedpipelinedesigncannolongermeettherequirementsfouserplane'subiquitousopenness,dynamicrouting,andintwillbecomethedirectionofevoluopenness.Keytechnologiesincludediversedataoffldynamic,on-demandorchestrationandopennessofwirelesscommunicaticomputing,andintelligentresourcesTheprimaryresearchgoaloftheorchestrationserviceistoestablishadistributehierarchicalintegratedorchestrcomputing,enablingtechnologiesencompass1)jointorchestrationofwirelesscommunicationandserviceintentions,intentiontranslation,andservice(locTheintegrationandcoexistenceofwirelesscommunicationandcomputitechnologyrepresentthedeepintegrationoftechnologiesdrivingtheevolutionofRANcomputing,andintelligenceparadigmof6G.2.2SemanticCommunThedesignoftraditionalcommuinformationtheory,utilizingbit-basedmetricincludingbiterrorrateandtransmissionrate-whilefocusingontechnicalchallengesassociatedwithaccuratelytransmittingbits/symbols.Ascommunicationtechnologyadvancesrapidly,systemcapacityhasnotheoreticallimitproposedbyShannon'stheory.Furthermore,ofcommunicationtechnologyandAI,intelligentterminalscannowcomprehendtransmissiontasksandscenarios,task-drivencommunicationsystemsprioritizeovershallowbit-levelprecision.Neverthelessthecontentandmeaningoftransmittedinformation,posingchallengesinmeetingthediverseandefficienttransmissionneedsofmulti-modaltasksinfuture6GneSemanticcommunicationisanewtypeofcommunicationonthetransmissionofmeaningsandsignificance,ratherthanthesymbolsthemselves.Itemphasizestheaccuratetransmissionofthemeaningofthetrthecommunicationprocess.fulfillingfuturecommunicationdemandattentionfromglobalactypes.Onetypeisrootedinfundamentaprobability.Drawinginspirationinformationtheoryincorporateschallengesassociatedwithmeasuringsemanticinformationtransmission.Inrecentyears,asdataprocessingcapabilitieshfeatureextractionmethapplicationsacrossvarioustypesofinformationsources.Forexample,semanticfeaturesoftextinformationareextractedusingnaturallanguageprocessingtechniques,whereassemanticfeaturesofimageinformationareextractedusingcomputervisionandpatcommunicationsystemsconductedonreinforcementlearning-basAIadvancesrapidly,semanticcommunicationpresentsextensandreplacesthetraditionalsourcecodingmodulewithsemaextractsemanticfeaturesassociatedwithunderlyingtransmissiontasks.features.Insemanticcommunication,apaknowledgebasesorsemanticmismatchesatboththetransmitterandreLeveragingknowledgebases,semanticcommunicationselectivelytransmitsthemostpertinentsemanticfeaturesforagiventask,rallowsfortheremovalofasignificantportionofFigure2-4SemanticCommunicationSystemModelThesemanticcommunicationsystemmainlyencompassecodingtechnologiesnearingtheirtheoreticallimits,therehasbeenextensiveresearchonjointsource-channeloutperformsseparatemodulardesign.Thefundamentalconceptofjointsignal-to-noiseratio(SNR)conditionstoenhancesystecounteractthedetrimentaleffectsofnoise.Insemanticcommunicommunicationtraffic.AsshowninFiguinspiredbyjointsource-channelcoding,hasrececonsideredoneofthemostimportantteincorporatingtheSNRinenvironments.Thankstotherapiddevelopmentofdeimages,andaudio,especiallFigure2-5JointSemantic-ChannelCodingModeltransmissionofcontinuousoutputsignalsthroughanalogmodulation,eliminatinghardwarelimitations.Candtransmissiontechnologybecomescrucial.Currently,theprimarymethodinvolvesquantizingcontinuoussignals,discretelytransmittingthem.Insomestudies,auniformquantizautilizedtoequidistantlymaptheoutputofDNNsintosymbols,whiletheotherstudiesemployedtomapcontinuoussigtransmissioninsemanticcommunicationsysteproblem,aimingtoachievee[11].Itisnolongerenoughtopursuetraditionalbit-basedoptimizaerrorrateminimization,toensufieldofresearchisdevelopingnewoptimizationgoalstosemantics.Thenewindicatorsofsemanticcommunicatisemanticspectralefficiency,andsemantictransmissionsystemsandoptimizedthesemanticQoEinmulti-modal2.2.3ApplicationsandC1)Cansemanticinforframework,alongwithassociatedtheoreticallimits,andhasachsuccessinthedesignofcommunicationsystems.Inspiredbyit,numerousresearchershavesoughttoadvancesemanticinfotheconceptofsemanticentropybyreplacingstatisticalprobabilitywithlogicalthefieldofsemanticinforandtask-oriented.Therefore,theremaynotbeauniversalmathematicalframeworktodescribethesemanticcommunicationsystemofmulti-modaltransmissiontasksandbackgroundknowledgebases.Furthermore,theyrequireadaptationtodiverseenselectingparametersduringthetrainingofneedformoreappropriateandsystematicperformanceevaluapracticalimplementationofstraditionalobjectiveindicatorstomeasuresystemspecificdatasources,emphasizingthenecessityofdevelopinguniversalmeasurementstandards(similartothebiterrorrateintraevaluatingtheperformanceofsystemsintegratingmulti-moincorporateAIintowirelesscommunicationsystems.communicationsystemsarepredominantlypoweredbyDNNs,necessitatingadditionalcomputingresourcesandpotentiallyresultinginsignificantcomputingoverhead.Therefore,thereisasubtlebacomputingpowerduringthepracticalimplementatiavailableonsomenewterminaldevices.Furtheintelligenceofferssupplementarycomputingpower.Innetwork(CPN),asanewarchitecfacilitatetheircollaborativesharingpossibilitiesforenablingsemanticcommunicatio4)PrivacyandsecurityinseMaintaininginformationprivacyandsecurityrepresentknowledgebasefordecodprivacyandsecurityprotectiontosomeextistransmitted,theeavesdroppingofsemanticinformationcouldhaveprseriousimplications.Insemanticcommunicationsystems,whichphysicalandsemanticchannels,securingcommunicationnecessarytodynamicallymanagethetrade-offbetweendatasecurityandtransmissionefficiencydependingonvarioustra5)Applicationandstandardizationofsemanticcommunicationin6Gandimmersivecommunicationinthecontextof6G,facilitatingtheconnectionbetweenthevirtualandphysicalworlds.Aswirelesscommunicationtechnologycoverage,andenableintelligentailsstringentrequiremandvoiceinputs.Removingredundantinformationreducesthevolumeoftransmitteddatawhilemaintainingseamlessdatainterchange.Semanticalleviatedownlinkpressureandcatalyzetheimplementationofaugmentedrealityandimmersivecommunicationtec2.3LargeModelandWTherapidadvancementofAI,drivenbybigdataandcloudcomputinghighlyintelligenthuman-machineinteractionexperiencesandicontentgenerationcapabcommunicationnetworksadvance,thearchitectureofwirelessnetworksiscontinuouslyevolving,accompaniedbyanincreasingvarietyofterminalsandserviceapplications.Foralocommunicationthatarediftraditionalmethods.ThedemonstratedpotentialoflargeAImodelshasledtowidespreadbeliefintheirbroadapplicabilityinareassuchasnaturallanguageprocessing,computervision,andmulti-modalfields,aswellastheirabiladvanceintelligentautonomousnetworkcapabilitieswithincommunicationnetworks.Inaddition,therapiddevelopmentandbroadapplicationoflargeAImodelsahighreliability.Therefore,researchingtheintegrationoflargeAImodelswithwirelesscommunicationisofsignificantimportance.2.3.1CurrentResearchStatusofLargeOriginatingfromnaareabbreviatedasLargeLanguageModels(LLM)whenmenlarge-scalepre-trainingmodels,andultra-large-scalepre-trainingmodels.Thenumberofparametershasadvancedfromthescaleofbillionstotrillions,markingsignificantnaturallanguageprocessing,machinevision,andscientificcomputingindustry-specific.General-purpoabilities,resembling"generaleducation,"whereasinduadjustthemodelswithindustryknowledge,similarto"professionaleCurrently,majorinstitutionsabAnthropic,Google,andMeta,withmobillionsandhundredsofbillions.Uptonow,thetopGPTlargemodelsabroadinnativemultimodallargemodelGemini,BardwasofficiallyrenOnaglobalscale,prominentinstitutionsmainlyincludeChatGPT,Claude,Bard,Llama2,Hugglargemodelscontinuetoememulti-modalprocessing,enablingittohandlevarioustypesofdata.Inthisglobalcompetition,Chinaisalargemodelsarerapidlyprogressing,trtoreal-worldtesting.NotaChinaMobile's"Jiutian."DataindicatesthatbytheendofNovember2023,morethan300largemodelshavebeenintroducedtot2.3.2EmpoweringtheCommunicationIndustrywithLaservicecapabilities.Forexample,massivedatafromcommunicutilizedtotrainlargecommunicationnetworkmodels,andtherobustnaturallancapabilitiesoflargemodelscanbeemployedtoenhanceoperatiolikeintelligentcustomerservice.Moreover,theoperationandprovisionofservicesbylar