1、1White PaperTen Cross-domain Innovation DirectionsAffecting the Future Information andCommunication Development2022.6(2022)PrefaceOver the past few years,fifth generation(5G)commercialization hasaccelerated consumption upgrading,production efficiency andsocialharmony,takingthetransformationofdigital

2、intellectualization and achieving high-quality development as themain objective.The rapid development of 5G has led to theadvancement and integration of big data,cloud computing,artificialintelligence(AI)as well as other cross fields,such as chips anddevices.The deep integration of communication tec

3、hnology(CT),information technology(IT)and data technology(DT)has becomean inevitable trend in the future.Compared with 5G,cross-domain and interdisciplinary IT(i.e.,cross-domain technology supporting or influencing the informationand communication development)will accelerate integration insixth gene

4、ration(6G)development,and play an important catalystrole.It is expected to break through the key challenges of ultra-largebandwidth,deterministic delay,low power consumption,low cost,high security,easy deployment and operation in the developmentof 6G,which will promote the society to the digital twi

5、n andwisdom ubiquitous,and truly realize the Metaverse where thevirtual and real world are integrated and interacted.Recently,for the traditional information and communication field,China Mobile has released a series of white papers such as the ones,entitled 2030+vision and requirement,technology tr

6、end,networkarchitecture prospect,etc.The aim of this white paper is to selectten cross-domain innovation directions affecting the developmentof information and communication in the future according to theblueprints of information and communication technology(ICT)in2035.The white paper analyzes the r

7、elated challenges,and calls onthe academia and industry for more attention and investment toaddress these challenges together,in order toensure thesustainable,healthy and great-leap forward development of theinformation and communication industry in the future.The ten cross-domain innovation directi

8、ons are key areas that mightaffect the development of information and communication industryin the future from the perspective of China Mobile.They involveterminals,management,cloud,computing,security,low carbon aswell as paradigm,which are only for enlightenment,even notcomprehensive.It is necessar

9、y to keep making calibration anditerative updating during research and exploration with colleaguesfrom academia and industry.ContentPreface.11.New Perception-Sensing the Physical World Comprehensivelyand Intelligently.12.New Terminal-A New Form of Human-computer Interaction.63.New Computing Power-En

10、abling High-Efficient and MassiveSecure Data Processing.94.New Security-A Built-in Security System for the Future Network145.Low-Carbon Innovation-Help to Accomplish Carbon Peaking andCarbon Neutrality Goals.196.New Materials-the Basis for the Development of FutureInformation Networks.237.New Bionic

11、s-Combining Biology Technology and InformationTechnology.268.New Networking-A Connection+Computing+CapabilityNetwork with Open,Flexible and Building-block-like Architecture.309.New Infrastructure-Integration of Communications and PublicInfrastructure.3310.New Paradigm-Innovative Industrial and Busin

12、ess Models.36Conclusion.411The peoples desire for a better life will keep going.By 2035,humanbeing is expected to build a digital intelligence society with digitaltwin and wisdom ubiquitous,and live in the meta cosmic world ofvirtual and real interaction.In the future,the information andcommunicatio

13、n network will expand to more scenarios and enablemore industries,and be more closely integrated with cross fieldssuch as new perception,new terminal,new networking,newcomputing power,new bionics,new security,Low-carbonInnovation and new materials.The communication infrastructuremay be integratedwit

14、h new facilitiessuch asmunicipaladministration,transportation and property management.This willresult in more blurred boundary between the future informationand communication industry and these cross fields so thatcross-border integration will become a new paradigm and a newnormality.Thus,more disru

15、ptive innovations will emerge into thebusiness model and industrial ecology.1.New Perception-Sensing the Physical World Comprehensivelyand IntelligentlyWith the advancement of internet of things(IoT)technology andapplications,the sensing technology in the future will be developedfromsingle-functiona

16、ndlowintelligencetomultifunction,collaborative,and high intelligence to support emerging application2scenarios such as integrated sensing and communication,andhuman digital twins.1.1 Sensing TechnologyAs a bridge that connects between the physical and the digitalworld,the sensing technology has beco

17、me the cornerstone ofindustrial digitalization,and is increasingly important to thedevelopment of the whole industry.In the future,the developmentof sensors will feature the fusion of various technologies,thesensors will integrate with technologies such as communication,energy harvesting,and heterog

18、eneous integration,thus achievemore accurate,reliable,and powerful sensing capabilities.Integrated sensing and communicationIn the future,the sensing technology will no longer be limited tohardware devices,instead,it can realize environmental perceptionby employing the ubiquitous network signals.Whe

19、n an objectmoving in radio frequency(RF)field,the position,moving speed,moving direction,and attitude of the object can be identified byanalyzing the signal variation of the radio electromagnetic wave,thus a large-scale environment perception can be realized.Theintegrated sensing and communication t

20、echnology is the fusion ofcommunication and radar,and therefore,the base station needs toallocate sensing resources in the time-domain and spatial-domain,and needs to support sensing and data processing functions,3although the current relevant research is still in the early stage.Passive-sensing tec

21、hnologyBy integrating environmental energy harvesting technology withsensors,the sensors can employ vibration,temperature,light,andcommunication signals as energy sources,thus the sensors requireno power supply,and can greatly expand the application scenarios.The key of passive sensing technology is

22、 to improve energyconversion efficiency,which is related to the physics and materialsscience,and it also involves low-voltage energy storage mechanisms,efficient energy management algorithms and software.Therefore,the biggest challenge of passive sensing is to combine basic science,electronic circui

23、t,and software together,and meanwhile achieveoptimal balance among those technologies.Microsystem technologyWith emerging technologies,the performance of integrated circuit(IC)is gradually approaching to the limit of Moores law.Therefore,the industry integrating microelectronics with other disciplin

24、es toproduce a new technology microsystems,thus to achieve Morethan Moore.The microsystem technology is a combination ofmicroelectronics,micro-electro-mechanical system(MEMS)andoptoelectronics,featuring miniaturization and systematization.Microsystems are miniaturized by using advanced integrationsc

25、hemes to produce new functions at the system level,which4greatly increases the function density of systems.The difficulty ofmicrosystems technology is to ensure high integration density whilerealizing coupling andcoalition between different modules,andthe microminiaturization also causes challenges

26、to robustness,reliability,and validation technologies of microsystem technology.Smart sensingThe advancement of IoT,big data and AI technologies facilitates thedigital and intelligent transformation of society,enabling all devicesto be connected and integrated to achieve data-driven andaugmented Int

27、elligence.The future sensing technology will havenew features:first,the intelligence,including strong sensing abilityand advanced understanding ability;second,the autonomy,whenoperating in an unattended environments,the sensors have strongenvironmental adaptability and self-organization ability;Thir

28、d,thecollaboration,devices shall dynamically share resources(data,knowledge,computing,communication and power resources)witheach other,and provide complex services through cooperation.However,to achieve those features,smart sensing needs to focuson the following challenges:First,the intelligence lev

29、el of IoTservices needs to be improved.The system requires the ability tolearn from small sample data,which can be used to build an IoTknowledge graph,realize resource virtualization,collaborativecomputingenvironmentsharing,andthustoachievehighintelligence level.Second,the learning ability of IoT se

30、rvice needs to5be improved.Through the data interaction between wearabledevices,environments,and personnel,the IoT devices shouldrecognize and predict human behavior and emotions,and provideservices on demand to meet individual needs.Third,the contextlinkage ability of IoT service needs to be enhanc

31、ed,for instance,themany-to-many mapping between user demand and data requirescomplex computing,and the behavior of sensors and actuatorsneeds to be designed with corresponding coordination models.1.2 Human Digital TwinDigital twins is one of the typical use cases of the new perception,which is based

32、 on emerging sensing technologies and intelligentperception capabilities,that can perform somatosensory perception,emotional perception,and consciousness perception of the humanbody,and understand the neuroplasticity of the human brain bylearning human somatosensory data and data of the devices ande

33、nvironments used by human beings,so that the human body andthe virtual twin can be mapped to each other.The digital twin alsoinvolves acquisition and transmission technologies in the humandomainsuchasbrain-computerinterfaceandmolecularcommunication,aswellasinformationprocessingsuchascomputation and

34、presentation.The realization of digital twins faces many technical and ethicalchallenges.The first is how to effectively fuse data from differentsources and use AI and other algorithms to achieve accurate6diagnosis and prediction for the somatic twin;the second is how toread and quantify human emoti

35、ons and five sensory sensations forthe empathic twin.In vivo information transfer requires a complexhuman channel environment,and related technologies such asmulti-dimensional heterogeneous transmission system based onmolecularcommunicationstillneedtobeimprovedandsystematically validated.Third,brain

36、 science is still in its infancy,andthere is need for breakthroughs in brain-computer interfacetechnologies for consciousness perception.Fourth,many newethical and privacy security risks willemerge,and correspondingpolicies and regulations need to be formulated.2.New Terminal-A New Form of Human-com

37、puter InteractionThe new human-computer interaction will shorten the distancebetween man and machine.In the future,new terminals should becomfortable,convenient,bio friendly,low power consuming andubiquitous.Flexible electronic devices will be closer to the humanbody and even integrated into the bod

38、y;holographic display cangenerate more immersive and interactive imaging effects,and bringricher sensory experience to users.The new terminal may beinvisible,enabling users to obtain a richer and non-perceptualexperience,by ubiquitous sensing equipment,cloud computingpower,friendlyhuman-computerinte

39、ractiontechnologyandubiquitous communication ability.72.1Flexible ElectronicsFlexible electronics can be summarized as an emerging electronictechnology that makes organic/inorganic electronic devices onflexible plastic and thin metal substrates.It can be used for flexibleelectronic display,organic l

40、ight emitting diodeorganic light-emitting diode(OLED),printing radio frequency identification(RFID),thin film solarpanel,with unique flexibility and ductility,as well as efficient andlow-cost manufacturing process.Like traditional IC technology,manufacturing technology andequipment are the main driv

41、ing force for the development offlexibleelectronictechnology.Flexiblethinfilmtransistortechnology is one of the important technologies.Compared withrigid silicon-based ICs,flexible non silicon-based chips are realizedby making thin-film transistors on plastic or metal foil substrates.Itis not only c

42、heaper to manufacture,but also thinner and moreflexible.The key of flexible electronic manufacturing is how tomanufacture smaller flexible electronic devices on a larger substrateat a lower cost.Its technical level can be measured by thecharacteristic size of the chip and the area of the substrate.O

43、nlywhen the transistor contained in the chip reaches a certain density,the performance of flexible chip can be compared with that oftraditional silicon-based chip.8Flexible electronic technology will break through the intrinsiclimitationsofclassicalsilicon-basedelectronicsandprovideimportant opportu

44、nities for future industrial development such asdevicedesignintegration,energyrevolutionandmedicaltechnology reform in the post molar era.Currently,there are twomainchallenges.Thefirstismechanicalproblems.Flexibleelectronic components will continue to bear alternating stress whenthey are folded and

45、bent repeatedly,which is easy to crack after along time.It is mainly overcome by structural design.The secondchallenge is the problem of electronic packaging,that is,how topackage the components integrated on the flexible substrate tightlyand achieve the expected function.2.2Holographic DisplayHolog

46、raphic communication will combine AI,big data and othertechnologies to build a multi-agent,realize the integration andinteraction between the physical world and the digital world,andprovide a new life experience of blending virtual and real.Holographic communication will integrate AI and a variety o

47、finteraction modes,and use new holographic display terminals andequipment to provide users with a two-way circular couplingexperience from multiple dimensions such as vision,hearing andtouch.The key technologies of holographic communication includeholographic display,sensing interaction and data com

48、munication.9Holographic display may carry a variety of new intelligent terminals,involving dynamic computing,three-dimensional display,opticalencryption and compression coding and other technologies;Sensinginteraction needs to support multi-channel virtual reality fusioninteraction ability(such as g

49、esture,posture,eye movement,voice,smell and touch).The multi-dimensional information is very easy tobe distorted by noise,jitter,packet loss and other factors,which is agreat challenge to maintain the naturalness,realism,immersion andsynchronization of interaction.A large number of complex datagener

50、ated by the interaction between holographic display andsensing have high requirements for communication.These data maycome from video,audio,touch,or people,physical objects andbackground environment.The required network transmissionbandwidth is very large.Efficient encoding/decoding technologymust b

51、e used for compression processing.At the same time,itneeds technical support such as ultra-large bandwidth,ultra-lowdelay,cloudedgeintegrationcomputingpower,highsynchronization and strong security.3.New Computing Power-Enabling High-Efficient and MassiveSecure Data ProcessingAs ICT pushes the world

52、into the digital economy society,themassive data has become as an increasing important factor ofproduction.As a new production tool,computing has penetrated10into many aspects of economy and society.In other words,computing power is productivity.The demand for computing poweris growing rapidly,leadi

53、ng to a big change in global computingtechnology and industries.And its system presents a trend ofdiversificationandcoexistenceofmultipleevolutions.Thenon-classical computing,such as quantum computing,opticalcomputing,lightweight computing,and privacy computing,isgradually moving from theory to prac

54、tice.3.1 Quantum ComputingQuantum computing is a new computing mode that followsquantum mechanics to control quantum information units forcomputation.It uses quantum bits as the basic unit,and realizesdata storage and calculation through the controllable evolution ofquantum states.It has the advanta

55、ge of strong parallel computingcapability and can provide exponential speedup on some specificproblems.Quantum computing can be used for quantum simulations,quantum optimization algorithms,quantum cloud computing,andquantum-enhanced AI,where the quantum optimization is aneffective approach to solve

56、many optimization problems includingthe direction of information processing.The high computing powerand high parallel computing requirements of AI make it possible tocombine quantum computing with AI.The model of quantumcomputing+AI has been used in data clustering and other fields.11However,the app

57、lication of quantum computing in informationcommunication still faces some great challenges.First of all,thecombination of AI and quantum computing is still in the stage oftheoretical exploration,which is far from practical application,andthe improvement of quantum computing on machine learning is n

58、otclear.Secondly,since quantum computing can only solve specificoptimization problems,it is necessary to explore the feasibility ofquantum computing on general optimization problems.Thirdly,there are some bottlenecks in the miniaturization and stability ofquantum computer.No matter what type of quan

59、tum computer is(superconductivity-based,optical-based,ion trap-based or nuclearmagnetic resonance-based),there are limitations in maintainingquantum coherence time and in producing coherent quantum,bothof which require extremely high material technology.3.2 Optical ComputingDue to the resistance,cap

60、acitance and induction,significantamount of heat would be generated when electric charges movefrequently inside the devices.As the circuit integration level keepsgrowing for communications and computing devices,the energyconsumption associated with the moving of charges increaseexplosively and becom

61、es unsustainable.Hence,it is quite necessaryto explore a computing system that is more energy efficient.Optical computer is a potential candidate with the aim to achievefull-opticalcomputingwhichincludes:opticalprocessors,12transmitter/receiver of optical data,optical memory,etc.Amongthem,optical pr

62、ocessors are able to fulfill the functions such ascontrolled amplification and optical logic-gates,based on thenonlinear diffractions of optical material/devices.Optical datatransmissioncanbethroughopticalfibersorspatiallightmodulators(similar to optical lenses).Optical memory can be madeof phase-ch

63、ange or thermal-change based optical disks.The biggestadvantage of optical computing is the low energy consumption.Apart from the optical-to-electro and electro-to-optical conversionsat the input and output ports,there is no electric charges movinginside the devices.Another big advantage of optical

64、computing isthat it can achieve massive inter-connection very efficiently,forinstance,a spatial light modulator is essentially an analog antennawith very large aperture which can distribute optical signals frommultiple branches to different receivers.Currently,the main challenge of optical computing

65、 is the maturityof optical devices which involves optical processing,optical datatransmission and optical memory.Significant breakthrough isneeded in the field of optics and material science.3.3 Lightweight ComputingIt has been envisioned that there will be trillions of AI-enabled IoTdevices globall

66、y.Computing on resource-constrained IoT devicesmust be energy-efficient to minimize the energy needs.Differentapproaches are needed given the diversity of IoT applications.13Using energy harvesting with low power chips in the IoT devices issuch an approach.For example,low-power microcontrollers(MCUs

67、)may consume only a few milliwatts to microwatts.Energy may beharvested from many sources,e.g.,radio waves,temperatures,vibrations,or sunlight.Thus,the research on technologies of energyharvesting is required.While the IoT devices may harvest energy,the energy-gain functionmay be challenged to keep

68、up with the computing applications.Thisis especially hard for sophisticated applications that need muchenergy.It may result in software problems if the device energy runsout during the application processes.In order to maintaincomputing correctness on such devices,additional measures areneeded to de

69、al with the outages and more researches on efficientcomputing in IoT devices are needed.Further,IoT devices are usually subject to resource constraints,e.g.,data,knowledge,energy,computing and communication resources.This requires IoT systems to be resource-aware,which cansimultaneously take into ac

70、count of status of sensor battery,thestatusofactuatorpowersource,andthe possibleenergyrequirement of the applications.Researches on resource-aware,energy-efficient computing technologies are requiredto realizelow-cost,small and efficient future IoT systems.143.4 Privacy ComputingAs the realization o

71、f cross-domain data fusion and data valuemining in a secure and compliant manner has become an urgentdemand in the era of big data,privacy computing technologiesaround multi-party secure computing and federated learning areattractingwidespreadattention.Privacycomputingutilizescutting-edge cryptograp

72、hy technologies such as secret sharing,oblivioustransfer,garbledcircuit,homomorphicencryptioncomputing,and zero-knowledge proof to make data available andinvisible.However,privacy computing needs further research andbreakthroughs in distributed computing systems,the improvementof trusted and secure

73、computing efficiency,the guarantee of modelaccuracy,the standardization of data elements,and the complianceof business development.4.New Security-A Built-in Security System for the FutureNetworkWith the openness of the network and the emerging of newtechnologies,the traditional security system based

74、 on patching andplug-in ideologies can no longer meet the needs of the increasinglycomplex network environment,which demands built-in securitycapabilities with network trust,secure data sharing and proactiveimmunity support,and ultimately realizes the transformation from15network security to secured

75、 network.4.1 Secure Quantum CommunicationThe development of quantum computers has been accelerating andtech giants including Google,IBM,Microsoft,and Intel haveinvested heavily into the research and development(R&D)ofquantum computing,which can solve mathematical problemsfundamentally challenging fo

76、r current computers,for example,Quantum computing may greatly reduce the time of crackingcommonly used asymmetric encryption algorithms such as the RSAalgorithm.Asymmetric encryption relies on the difficulty offactoring large numbers into primes,but the prime factorizationalgorithm based on quantum

77、computing(e.g,Shor)can find primefactors of large numbers in a much shorter period of time,whichshakes the foundation of asymmetric cryptography systems basedon the complex mathematical problems.Moreover,quantum entanglement is crucial for long-distancequantum communication and scalable quantum comp

78、uting,but thestorage of quantum entanglement,especially high-dimensionalquantum entanglement is currently a challenge and must be solvedin order to build a quantum communication network capable ofhigh volume and long-distance transmission.As a result,it is needed to carry out researches on quantum-b

79、asedsecurity technologies in the context of mobile communication,16including:Quantum attack resistant cryptographic algorithms;Thelastone-kilometertransmissionproblemofquantum crypto key;Mathematically proven secure crypto key by using thetime-varying,reciprocal and spatially irrelevant properties o

80、fwireless channels;The storage of quantum for long-distance and securequantum communication etc.4.2 Dynamic Data Coloring TechnologyAs data gradually become critical assets in business operations andcarry great values,the risks of unauthorized data access willincreaseexponentiallywhendataflowwithino

81、routsideorganizations.Traditional data security technologies are insufficientin solving data pollution,data privacy violation,and data leakageintractability problems exacerbated by increasing data fluidity.Inthe future,data security capabilities need to adapt to the datafluidity to ensure dynamic se

82、curity protection for flowing data.Therefore,it is necessary to study the data security technologieswith data fluidity awareness,built-in support of data asset labeling,data flow path management and risk detection,which will provide17the capabilities of data pollution control,data value assessment,a

83、nd data tracing required for data protection in future dynamicnetwork environment.4.3.Trusted Built-in Security SystemThe traditional security systems are no longer able to effectivelydefend against security risk uncertainty and potential attacksbecause of their plug-in and patching based mechanisms

84、.There aretwo fundamental challenges:Transform network from risk uncertainty to certainty ofsecurity trust with the support of trust relationship establishment,transmission and evaluation;Evolve from passive protection of network to activeself-immunity security to meet the requirements of predictabl

85、enetwork behavior,strict management,and extreme end-to-enddifferentiation for different services.Zero trust is a security concept based on continuous verification,never trust,which will play an important role in the future mobilenetworks and break the traditional practice of providing securitymostly

86、 at the network border.Facing the future open networkenvironment,its of great importance to study the securityarchitecture with deep integration of zero trust and communicationnetworks,to build a trusted endogenous security system,and to18plant security into the networks genes to provide ultimate se

87、curityfeatures such as self-immunity and self-evolution.4.4 Unified Identity Authentication based on Biometric and DeviceCharacteristicsThedevelopmentofthedigitaleconomyandmobilecommunications has brought a growing number of economicactivities with high security requirements on mobile networks,sucha

88、s comprehensive government administration,financial payment,medical care,etc.In these scenarios there are demandingrequirements for users identity authentication;however,currentlymobile phone numbers are the single commonly used real nameidentifiers for users,which cant meet the needs of consistency

89、 ofauthentication between business users and mobile device owners.In recent years various biometric authentication technologies haveemerged followingthe development of sensorand featurerecognition technologies,such as fingerprints,palm prints,face,iris,voiceprint,etc.The biometric based authenticati

90、on has theadvantageofconveniencewithoutsettingandmemorizingpasswords,and is rapidly being used in various applications.On theother hand,the device authentication based on RF fingerprints canperform individual identification of wireless devices at the physicallayer,which has also become a future rese

91、arch direction.A trusted digital identity authentication system should be built for19mobile networks with unified identity authentication which shouldcombinebiometrics,devicefingerprintandtraditionalauthenticationtechniques.Forthisgoalmoreresearchoncutting-edge intelligent authentication mechanisms

92、incorporatingmultiplebiometricfeatures,multipleauthenticationfactors,continuous assessment and verification are needed.5.Low-Carbon Innovation-Help to Accomplish Carbon Peakingand Carbon Neutrality GoalsIn September 2020,China announced at United Nations GeneralAssembly(UNGA)that it will strive to p

93、eak carbon dioxide emissionsby 2030 and achieve carbon neutrality by 2060,which is an arduoustask.Complex and large scale communication systems are seen ashigh energy consumption(EC)infrastructures,therefore need todecrease carbon emissions by focusing on two possible solutions:reducing expenditure

94、and broadening sources of income,i.e.tocurb energy consumption and switch to green energy.In terms of reducing expenditure,due to the huge cost pressure onnetwork operators given by the high EC of network devices,ZeroWattsZero Bits is necessary to be realized both in EC andcommunicationvolumessimult

95、aneouslyforfuturenetworks.Omni-directionalimprovementofenergyefficiencyandECreduction relies on the following technologies of high efficiency20radiosignalingtransmission,highefficiencyhardwareimplementation,deepsleeptechnology,integrationofcommunication technology and other applications,etc.5.1 High

96、 Efficiency Radio Signaling TransmissionFuture network will adopt minimal signaling structure,reducingenergy radiation from multi-dimension including source domain,time domain,frequency domain and spatial domain.For example,new encoding method can be used in source domain to transmitmore information

97、 with fewer bits;innovation waveform is beingdeveloped in time domain to increase sleep time duration ofdevices;more technologies in spatial domain such as very largescaleactivemeta-materialphasedantennaandlasercommunication can help to form more precisely directed beams,and lower energy waste of ra

98、dio radiation.5.2 High Efficiency Hardware ImplementationFirst,we must improve the efficiency of RF devices.RF devices arethe main EC components of traditional base station system,amongwhich RF power amplifier(PA)has the highest proportion of powerconsumption.In terms of material and process,the fou

99、rthgeneration semiconductor will be applied in PA;and new circuit andalgorithm design can be used to further simplify the architecture,for instance,direct RF sampling technology and fewer interfacesbetween modules will increase RF efficiency to 80%,90%or even21100%.Second,we must reduce digital proc

100、essing EC.Future network willuse passive optical devices instead of active chip.Thanks torefraction,scattering and other optical properties,series of complexlogical operations in digital computing can be completed withalmost zero EC.Furthermore,improving chip process will be ofbenefit to gradually a

101、pproach the power density limit,andtechnologies like chiplet encapsulation are also expected to breakMoores Law to further lower the EC of digital signal processing.Third,we must promote heat dissipation.Let the heat out of therunning system through liquid cooling at the plate levelandmaterial of fa

102、st thermal conductivity,to increase the power densitylimit of the chip.The main challenges here are high price of newmaterials,increase of deployment and maintenance cost,difficultyof production process,and low degree of supply chain autonomy,etc.5.3 Deep Sleep TechnologyTraffic-awarebasedmodulariza

103、tionhardwareandminimalistprotocol design are introduced to switch on/off functionalitiesaccording to traffic load conditions.When network is out-of-service,most hardware modules transit into deep sleep mode with almostno power consumption.As soon as a terminal detects traffic arriving,dormanthardwar

104、ewillresumerapidlytoprovidealmost22zero-latency loading,and ensure high quality of network service.The main challenge in this case is to completely redesign thehardware and software of the chip.5.4 Integrating with Other ApplicationsCommunication integrating with other applications is a kind ofeffic

105、ientwaytosaveenergy,whichcanachievedoubledachievementwithhalvedeffort.Forexample,visiblelightcommunication combines semiconductor lighting technology withoptical communication technology.Combined,the light-emittingdiodes(LEDs)light intensity is used for information modulation.Itneeds to overcome the

106、 problems of limited LED bandwidth,limitedcoverage,and susceptibility to interference.In addition,the scale ofnetwork nodes will grow sharply in the future.Free electromagneticenergy and friction power generation can also be employed asenergy sources to achieve coordinated transmission of informatio

107、nand energy and recycling energy saving.In terms of finding more resources,green new energy should bevigorously promoted.However,the new energy consumption rate islow,due to the problems of curtailment.In order to improve theutilization rate,it is necessary to build energy storage facilities.Many ba

108、se stations in mobile communications are located in top ofmountains and rural areas where new energy resources arerelatively abundant.Other than running base station directly by thenearby green energy,excess electricity can also be stored in backup23batteriesinthebasestation.Theintelligentendogenous

109、communication network can be integrated with the intelligentnetwork to maximize the utilization rate of green new energy.Compared with wind and solar energy,nuclear fusion is stable andenvironmentally friendly.It is considered to be the ultimate energysource that can solve human energy problems.The

110、difficulty ofachieving nuclear fusion lies in controllability.Controllable nuclearfusion methods include magnetic constraints,inertial constraints,gravitational constraints,etc.Magnetic constraints are close tosuccess under laboratory conditions,which is also the mainstreamresearch direction.But the

111、re is still a long way to go in terms ofcommercialization,andthereshouldbemorecontinuousexploration and breakthroughs.6.New Materials-the Basis for the Development of FutureInformation NetworksThe development of information and communication networks withhigherfrequencybandsandlowerEChasputforwardtr

112、ansformative demands for new materials.New antenna materials,new chip materials and new optical fiber materials have becomethree directions worthy of attention.246.1 Nanoprinted AntennaNanoprinting is an emerging technology for the preparation offlexible electronic devices by printing new materials

113、on differentflexible substrates.On a flexible fabric substrate,flexible wearablefabric antennas can be prepared by printing nano-conductiveparticles or coating conductive polymers according to the antennastructure.This makes flexible wearable communication systemspossible.Currently,printing electron

114、ic ink to prepare flexibleelectronic devices has become a new direction,but the printing inkmarket is less mature and requires huge amounts of polymericnanocomposites to assist development.6.2 Carbon NanotubeCarbon Nanotube(CNT)is an alternative to silicon transistorspotential technology for the pro

115、duct.At the same level ofintegration,carbon nanotube chips are smaller than silicon-baseddevices.Carbonnanotubeshave extremelyhigh toughness,strongerelectrical conductivity,shorter electrical signal transmission delay,high thermal conductivity,and low EC.In theory,compared withsilicon transistors wi

116、th the same characteristics,the energyutilization rate of carbon nanotube chips can be increased by 60%to 70%,the operating speed can be increased by 5 to 10 times,andthe power consumption can be reduced by 10 times.However,25large-scale production of high-purity and uniform carbon nanotubesand fixi

117、ng them on the substrate to make field effect transistorsaccording to the design requirements still face huge challenges.6.3 Anti-resonant Hollow FiberWith the explosive growth of information traffic,the inherentdefects of quartz as a fiber material gradually appear,such asnonlinearity,inherent time

118、 delay,limited bandwidth,maximumcapacity of single-mode fiber(bandwidth distance product),photo-induced damage,etc.barriers to technological development.New optical fibers such as hollow photonic crystal fibers,hollowanti-resonant fibers,few-mode fibers,and multi-core fibers havebecomeactivefrontier

119、researchfields.Amongthem,theanti-resonant hollow fiber has achieved breakthrough developmentin recent years due to its low loss and large bandwidth.In the future,itisanticipatedthathollowfibercanreducetheopticalcommunication delay by 1/3,and its ultra-low nonlinearity,largetransmission bandwidth win

120、dow and other characteristics areexpected to greatly increase the optical amplification distance andoptical system capacity.At present,the basic theories such as the light-guiding mechanism,bandwidth limit and loss limit of anti-resonator hollow fibers havenot yet been thoughtfully studied.The resea

121、rch of end-to-endanti-resonator hollow fiber communication system has just started.Thesubversivenewhigh-speedlong-distanceoptical26communication system based on hollow fiber needs to be exploredfrom scratch,and the industry chain needs to promote the researchof hollow fiber and its ultra-high-speed

122、long-distance opticalcommunication system.7.New Bionics-Combining Biology Technology and InformationTechnologyThe development of brain science and material science paves theway for fast progress of bionics in the field of information&communications and spawns two new directions:brain-inspiredcomputi

123、ng and DNA-based storage.7.1 Brain-inspired ComputingArtificial neural networks(ANN)was proposed in 1940s.The inputto each neuron in an ANN is a static electric signal.Its nonlinearresponses to excitations can be modeled by various nonlinearmappings.However,this is quite different from the transmiss

124、ionand processing of neural signals inside human brains which are inthe form of signal pulses.Compared to the cognitive process ofhuman brains,ANN merely utilizes some topology structures ofneurons,which is essentially the pile-up of digital logic gates.Manykey mechanisms of human brains are still u

125、nknown.27By further inspired by the bionics,models for neurons and synapsesthat are most closely resemble their biological counterparts havebeen proposed,the so called the third-generation neural networktechnology or brain-inspired computing.Its potential is huge:accurate modeling of space-time char

126、acteristics of biological signalswhere neural signal pulses are transmitted and processed.Thiswould drastically reduce the activity of neurons and ensurelow-energy consumption with high efficiency.By passing and distributing neural pulses across neurons andsynapses,brain-inspired computing can learn

127、 faster than deeplearning when it is used for characterization and signal processing.Hence,it is more suitable for high-efficient space-time signalprocessing and data mining.Brain-inspired computing has graduallyshown its benefits in the fields of intelligent decision-making,robot-control and comput

128、ational neurology.It is anticipated thatthe architecture and chips for brain-inspired computing will be oneof the key areas of computer architectures in post-Moore Era.Still there are several challenges for brain-inspired computing.Firstof all,the study is lack of basic support.Brain-inspired comput

129、ingrequires chips of special purposes,as well as a series of hardwareand software support that is based on neural science.Such chiparchitecture is based on new programable architecture that followsthe principles of neural science and has bio-neuron like analogcircuits to achieve fuzzy logic.This is

130、totally different from the28Von-Neumann architecture which is based on general-purposeprocessors and heterogeneous accelerators.In Von-Neumannarchitecture,memory and processors are separated,and the entiresystem operates with instruction flows.The study on brain-inspiredchips is still in its early s

131、tage,hampered by the lack of coredevices/materials and implementation complexities.While thegeneral processor-based simulator for brain-inspired computing canmitigate the deficiency of chips,the study and development of basicmodules is still in exploration stage and quite incomplete.Second,the human

132、 brains are not adequately understood andsimulated.Right now we can only achieve quite coarse simulation ofneural signal transmissions.There is still a long way to go inexploring how the information is transmitted,exchanged andintegrated via intricate coordination among hundred billions ofneurons in

133、side a brain.In addition,the study on brain-inspiredcomputing requires deep knowledge and expertise.Yet,such talentsare quite rare,not enough to hold up a big industry.7.2 DNA-based StorageEfficient data storage is an indispensable part of informationtechnology.Benefitingfromthefastdevelopmentinsemi

134、-conductor storage and magnetic storage,the data recordingdensity has kept growing at a pace similarly to Moores law.However,assemi-conductortechnologyisapproachingthequantum limit,it is necessary to look for alternative media to store29information,to meet the demands by the development ofinformatio

135、n technology.The idea of DNA-based storage wasproposed in 1960s,not long after DNA was found.After about 30years of a quiet period,the study on DNA-based storage hasbecome active,triggered by the projects of gene sequencedecoding.DNA-based storage is a way of information storage based on thebase seq

136、uences on the chromosomes of nucleus inside cells.Thereading and recording of base sequences on DNA segments are viathe chain reaction of polymerase.As the size of each DNA is atmolecular level,the information unit has two bits(A,G,T and C),and the structure of DNA is double helix and compressible,e

137、xtremely high storage density can be achieved within a very smallspace.Compared to the traditional solid-state memory devices andmagnetic disks/tapes,DNA-based storage has the advantage of lowpollution,light weight and organically degradable.Nevertheless,currently the data rate for the reading and r

138、ecording of DNA-basedstorage is very slow,much lower than that of solid-state devices ormagnetic based storage.Also it is quite expensive to prepare DNAsamples suitable for data storage and the cost of read/write is veryhigh.These challenges require breakthroughs in the fields of biology,chemistry,e

139、tc.to solve.308.New Networking-A Connection+Computing+CapabilityNetwork with Open,Flexible and Building-block-like ArchitectureFor the long-term business needs in 2035,the communicationnetwork will support the access from space,sky,earth and sea.Thescope of business access will greatly expand.The bu

140、siness data willhave a huge growth and the business types will be extremelyenriched.The current networking mode which include centralizeddeployment and centralized management and control is difficult tomatch these above-mentioned needs of future business changes.Therefore,the future communication wi

141、ll evolve towards an open,flexible,building-block-like connection+computing+capabilitynetwork.In the future,the integration of network abilities includingcomputing ability,perception,intelligence and other capabilities willbe significantly improved.Eventually,the network,computing andintelligence wi

142、ll be omnipresent.In the future,communicationnetworking needs to perform flexible and real-time integratedcontrol of multi-standard access,resource sharing and networkstructure expansion according to the needs of various types of usersso as to achieve the optimal management of spectrum,connection,co

143、mputing and capability.The future networking also can supportthe ways of network services leap-forward improvement fromconnection to computing and finally to capacity and support theflexible expansion and on-demand deployment of network servicecontent for user-centric.31Plug-and-playconnection:Enabl

144、ednetworkhasflexiblearchitecturesandprovidesbasicarchitecturesupportforleapfrogging network services from connection to computing.Design a minimalist end-to-end protocol functions can enable fast,robust,and ubiquitous connections between networks,users andnetworks,andbetweendifferentfunctions.Themin

145、imalistend-to-end protocol functions also can achieve resource sharingbetween multi-standard and multi-function,fusion of the individualdomains,local domain and wide domain,space-air-earth-seaintegration,and integrated ubiquitous connection and ubiquitouscoverage of the fusion multi-dimensional netw

146、ork which supportedby all services.When the terminal accesses the network,only thekey identity authentication and security authentication are neededto realize basic communication.When other access points,includingaccess points from space,air,earth or sea,access the anchor pointto expand the network

147、coverage,the security authentication isperformed between other access points and the anchor point toachieve wireless coverage.Building-block-like service-oriented network function:Enablednetwork has flexibility functions and realizes user-centric networkservices.A network functional component is cre

148、ated withservice unit for user needs.Enabled network can activate thecorresponding service component according to user needs,andrealize thebuilding block like network combining,arranging,configuring and deploying service components base on demand.32Eachservicesoftwarefunctionalcomponentrealizestheun

149、iversality of service components by standardizing its relatedfunctions,input and output,and realizes end-to-end consistencythrough signaling control,the last but not the least,realizesend-to-end compatibility through unified orchestration on thenetwork side.Crowdfunding network with open capabilitie

150、s:Based on flexiblenetwork architecture,network functions build an open system ofdistributed intelligent cloud network,and build a capable opencrowdfundingnetworkthatintegratesprotocols,perception,intelligence,third-party applications and other multi-dimensionalcomputing networks to achieve network

151、service capabilities for alltypes of users.We build an end-to-end digital twin autonomicalsystem,realize multi-level autonomical network with collaborationof centralization+distribution and offline+online,and providenetworkcapabilityservicesfornetworkmanagementandintegration of intelligence and powe

152、r for third-party applications.Through the cloud,edge,terminal,industry integration can bemade with distributed and centralized cooperation.On the onehand,more network functions will be extended to the networkedge to realize the regional autonomy of edge network;On theother hand,the global-oriented

153、orchestration and schedulingfunctions will be centralized to support complex cross-domainservices.33Althoughtheindustryhasreachedanagreementontheabove-mentionedtechnologyevolutiondirection,theimplementation paths and methods of specific technologies havenot been unified yet.It is necessary to contin

154、ue to gatherconsensus of industrial,jointly carry out innovation by multi-fieldintegration,build a brand new technology system,and achievetechnological breakthroughs.In addition,the industry has formedan independent ecosystem around their respective fields with theclosed-form nature.It needs to rely

155、 on standardization,open sourceorganizations,and industry alliances to reshape the industrialpattern.9.New Infrastructure-Integration of Communications and PublicInfrastructureThe coordinated construction and integrated development ofinformation infrastructure,fusion infrastructure and innovationinf

156、rastructure have become a new trend.In the future,municipalpublicinfrastructure,buildinginfrastructure,andwide-areainfrastructurecanadoptvariousintegrationsolutionswithcommunication systems,which lay the foundation for a newinfrastructure with Connection+Computing+Capability.349.1IntegrationofCommun

157、icationandMunicipalPublicInfrastructureCommunication technology is an important means to support thetransformation and upgrading of traditional infrastructure,whichhelps to form a new type of integrated infrastructure and improvetheintelligence,efficiencyandconvenienceoftraditionalinfrastructure.The

158、 integration of new information infrastructure,for example,communication capabilities with municipal light polesand so on to achieve simultaneous construction of lighting andcoverage systems.The transparent antennas are integrated with thebus billboard.The intelligent transformation of the manhole c

159、overcanrealizeblindandheatsupplementationorcoverageenhancement at the site and along the line.A wide range of IoTapplicationsmaybeintegrated,suchascombiningsensortechnology with water,electricity,and gas pipeline networks toachieve intelligent collection and reporting of IoT information,disaster war

160、ning,etc.;A communication management platform maybe built for smart cities with smart city operation capabilities;vehicle-road-cloudcommunicationinteractivenetworkintegration infrastructure can be constructed for special scenariossuch as smart transportation,etc.9.2 The Combination between Communica

161、tion and BuildingsInfrastructureWemaydesignandconstructbuildingstogetherwith35communication network,by introducing new material,new processand new technology.For example,LED lights are utilized in buildingsto modulate information,to develop visible light communication.Reconfigurable Intelligent Surf

162、ace(RIS)may be embedded on thesurfaceofbuildingswithbeamformingtoadjusttheelectromagneticwavebeamintelligently,promotingtheoptimization ofsignal coverage.Antenna technology and metalgrid printing&etching technology may be combined basedon transparent dielectric substrate to realize the integrated de

163、signof transparent antenna and building glass,which update theappearance of communication equipment and construction facilities.Itdramaticallyincreasestheinstallabilityandconcealment.Ubiquitous wireless technology helps to completeintelligent buildings,fully utilizing the information transmissioncap

164、acity of indoor communication equipment.The integration ofcommunication network,location network and IoT could providenew application for buildings.9.3 Integration of Communication and Wide-area InfrastructureAn space-air-ground-sea wide-area information network may bebuilttocreateamulti-level,three

165、-dimensional,all-round,all-weatherinformationspace,includingspace-air-ground-seawide-area ubiquitous communication system architecture,meetingthe requirements of communication immediacy,stability,security,and compatibility;satellite communication system and satellite36Internet at various orbital alt

166、itudes of high,medium and low;floating communication platforms such as balloons and airships;broadband wireless access system for high-speed rail,aircraft andother civil transportation vehicles;maritime and underwatercommunication technology and networks;targeted and customizedwide area information

167、network application for various verticalindustries(suchasagriculture,forestry,fisheryandanimalhusbandry industry).In order to promote the healthy development of new informationand communication infrastructure in the future and realize theubiquitous deployment of new infrastructures that integrate an

168、dsynergize,it is necessary to build a broad consensus and addressdifficulties and challenges:The design of communication networkarchitecturefornewinfrastructurerequirescomprehensiveinnovation.The deep integration of infrastructure needs to bepromotedthroughcross-industrycooperation.Therelevantsuppor

169、ting policies need in-depth research,bold formulation,andactive implementation.business scenarios towards consumers andbusinesses and related industrial chains need to be activelysupported and cultivated.10.New Paradigm-Innovative Industrial and Business ModelsMobile communication will involve more

170、cross-border fields and37disciplines.The data,information,and communication technologies(DICT)will cultivate a new open source ecosystem and create newindustrial and business models.With the continuous maturity oftechnology,the increasingly rich scenes and data will help the realworld map to the dig

171、ital world more accurately,and realize theinteraction between the virtual and real world through digitalidentity,digital twinning and digital reconstruction,forming a newrule paradigm.10.1 Fostering New Industrial Models via DICT ConvergenceFutureInformationTechnologiesarerapidlyevolvingandconvergin

172、g along with emerging technologies such as big data,cloud computing,and AI.The deep convergence of DICT hasbecome an inevitable trend in the mobile communications industry.ThestandardizationprocessinthetraditionalCTindustrydemonstrate itself as a long cycle with slow pace of evolution,and itsets a h

173、igh bar for vendors to participate.For example,the 3GPPoften takes 2 to 3 years for a given version of standard to completefrom its origination to finalization.It is hence extremely hard forsmall and medium enterprises to afford and follow through such along R&D cycle,except for a handful large oper

174、ators,globalequipment vendors,and chip manufacturers.Broad innovation istherefore hindered,preventing the growth and prosperity of theeco-system.In contrast,the modern IT industry leverages open source38communities to promote the creation and growth of eco-systems.Typically,open source projects evol

175、ves rapidly,and become the defacto standard after productization,which in turn pushes thestandard forward.Furthermore,widely used agile developmentmethods in R&D in recent years,along with various deploymentautomation,testingframeworks andtools,closed-loopdatautilization,life cycle management of AI/

176、ML algorithms and models,they together spawned a wild variety of software-defined modelsand eco-system.The adoption of the microservice architecture andthe momentum of cloudification have effectively helped thesoftwareandhardwaredecouplingandthesoftwaresegmentation.Combined with flexible APIs and SD

177、Ks provided bythe community,it sets a low bar so that start-ups,even individualdeveloperscanparticipateandcontributeinopensourcecommunity,resultingsignificantgrowthoftheeco-systems.Enterprises participating in open source and innovation can alsobenefit from faster go-to-market pace,forming positive

178、feedback tobuild up a healthy structure of the eco-system.The deep convergence of DICT also brings new challenges to thefuture information and communication industry and standardizationprocess.The synergy of the three is not only reflected in thetechnical perspective,but will also extend to others s

179、uch aseco-system,security,standardization,open source community andetc.It remains a challenge to the information and communicationindustry to foster new industrial models when it sorts out to learn39from past lessons,leverage on each others strengths,and manageorganic growth among the three.10.2 New

180、 business models are born from distributed technologyWith the popularization of IoT technology and the wide applicationof related equipment,massive data will greatly promote thedevelopment of distributed technology,and it will be improvedaccordingly for the subsequent industrial application,division

181、 forsocial responsibility and economic mechanism.The traditional centralized data center can no longer meet the newchallenges brought by the changes as above.Instead,it is realizedby point-to-point,agile and flexible,data privacy,low-cost andefficient distributed solutions.With the whole life cycle

182、includingdata generation,collection,storage,circulation,collaboration,analysis,intelligence,etc.,its needed by technical support andsustainable development,such as blockchain,edge computing,privacycomputing,federallearningandotheremergingtechnologies in this field.Distributed technology allows to re

183、cognizeeach others identities,share data and establish collaboration,generates new business and operation modes among individuals,and organizations.With the new multi-party trust and coordinationmechanism,the digitalized and AI-oriented world can greatlyimprove flexibility and response efficiency.En

184、terprises can explorenew markets with new methods,while the government andrelevant organizations should make new policies and industrial40eco-system standards.In addition to the digitalization and intellectualization for industrialeconomics,themetauniverseshouldalsohavestrongcharacteristicssuchasdiv

185、isionandcoordinationonsocialresponsibility,and reorganization on social relationship,so as toform a new safe,stable and orderly economic system for productionand operation.In this process,the establishment of industrialstandards and regulatory order is accompanied.It needs tostrengthen the overall p

186、lanning of distributed coordination,guidescientific and technological giants and industrial leaders to carry outstandardization cooperation,and support the entity economy toformulatevariousstandardssuchasdistributedtechnology,infrastructures,services and contents.In terms of laws andregulations,from

187、 monopoly to coordination,from supervision toautonomy,from risk to security,all kinds of new problems andchallenges have also emerged accordingly.The formulation of lawsand regulations to promote the development and application ofdistributed technology should start with basic elements such asalgorit

188、hms,data and transactions.41ConclusionFocusing on the vision of information and communication in2035,this white paper selected ten cross-domain innovationdirectionsthataffectthedevelopmentofinformationandcommunication technology in the future from the perspective ofChina Mobile.It is expected to att

189、ract more attention and motivateto work together on these directions,and calibrate and iterate theviews constantly with the colleagues of academia and industry.It isbelieved that,with the joint efforts of global innovation chain andindustrial chain,the breakthrough of these ten cross-domaininnovationdirectionswillhaveaprofoundimpactonthesustainable,healthy and great-leap forward development ofinformation and communication industry in the future,which willhelp the bright vision of digital twin and wisdom ubiquitous to bea reality!Digital Twin and Wisdom Ubiquitous