1、1/35Executive SummaryIn 2023,ITU-R issued the IMT-2030 framework highlighting sustainability,security,and resilience,connecting the unconnected,and ubiquitous intelligence as overarching aspects which act as design principlescommonly applicable to all usage scenarios.In another recommendation about

2、the future development of IMT for2030 and beyond,ITU-R mentions that quantum technology with respect to the RAN is a potential technology toensure security,and resilience when allowing for a legitimate exchange of sensitive information through networkentities.Therefore,the target is becoming more cl

3、earer to apply quantum technology in achieving secure andresilience in the 6th generation(6G)communication and beyond.To this end,in this annually revised white paper,we introduce research progress in applying quantum information technologies(QITs)to communication andnetwork and computing over the p

4、ast year and propose some expectations of quantum technology research in 2024.Chapter 2 focuses on quantum secure communication aiming at safeguarding critical information by applyingquantum mechanisms.The introduction starts with various theories and experiments continuously carried out inquantum k

5、ey distribution(QKD),quantum random number generator(QRNG),and quantum information network(QIN),followed by state-of-the-art standardization activities for QKD all over the world.In the implications for 6G,quantum encryption demonstration deployed on the internet of vehicles;integrated continuous va

6、riable QKD(CV-QKD)with G.698.4 device;and deploying quantum cryptography in the 6G network are introducedrespectively.Chapter 3 gives insight into the research of how to satisfy the dramatically increased communication systemperformance and rich diversity of innovative services expected by 6G by app

7、lying quantum computing.Firstly,computing scenarios and key issues for communication are analyzed,including signal processing,networkoptimization,service processing,and network intelligentization.Secondly,a Classical+Quantum hybridcomputing platform with a robust computational foundation is proposed

8、 to provide computational support servicestailored to different domains,facilitating research innovation and product implementation.Thirdly,the implicationsof quantum computing for 6G are introduced with three examples,which apply quantum computing to solveclassical communication issues,respectively

9、.2/35Based on the barrier-breaking achievements in 2023,2024 will probably mark a significant year for quantumcomputing technology,from when the field of quantum computing is expected to transition from physical qubits toerror-correcting logic quantum bits,and anti-quantum cryptography research is e

10、xpected to speed up as well.3/35Table of ContentsExecutive Summary.11.Introduction.42.Quantum Communication and Network.62.1.Key Technologies.62.1.1.Quantum Key Distribution.62.1.2.Quantum Random Number Generator.82.1.3.Quantum Information Network.92.2.Standardization Activities for QKD.102.2.1.Chin

11、ese Standardization Progress.102.2.2.International Standardization Progress.122.3.Implications for 6G.162.3.1.Quantum Encryption in the Internet of Vehicles.162.3.2.Quantum Encryption Integration with Bearer Network Equipment.172.3.3.Quantum Communication Security.183.Quantum Computing.203.1.Computi

12、ng Scenarios and Key Issues for Communication.203.1.1.Signal Processing.203.1.2.Network Optimization.213.1.3.Service Processing.223.1.4.Network Intelligentization.223.2.Quantum Hybrid Heterogeneous Computing.233.3.Implications for 6G.263.3.1.Single-Cell Massive MIMO Antenna Optimization.263.3.2.MIMO

13、 Beam Selection of Multiple Cellular.283.3.3.Phase Correction of Millimeter Wave Signals.314.Future Expectation.345.Acknowledgement.354/351.IntroductionThe scope of this annually revised white paper is to introduce the latest research progress about quantuminformation technologies(QITs)fulfilling st

14、ringent demands of communication and computing envisaged in 6G orbeyond 6G.In addition to benefits expected from QITs to communication and network and computing,this versionof 2024 white paper proposes some expectations of quantum technology research in 2024.Chapter 2.Quantum Communication and Netwo

15、rkChapter 2 focuses on quantum secure communication aiming at safeguarding critical information by applyingquantum mechanisms.In 2023,various theories and experiments have continuously been carried out in the following key technologies.For quantum key distribution(QKD),progress has been made in new

16、protocols and classical quantumco-transmission studies,etc.,and the performance of QKD systems has been further improved.Quantum randomnumber generator(QRNG)technology is currently being developed and improved to achieve more efficient andstable QRNGs.Many laboratories and research institutes have c

17、onducted a number of experiments to verify thefeasibility and stability of quantum Information Network(QIN).About the standardization activities for QKD,major standardization organizations have actively carried out thepreparation of QKD related standards,covering terminology definitions,application

18、scenarios and requirements,network architecture,equipment technical requirements,QKD security,testing and evaluation methods,and otheraspects.In the last,the implications of quantum technologies for 6G are discussed from the following three aspects:quantum encryption demonstration deployed on the in

19、ternet of vehicles;integrated continuous variable QKD(CV-QKD)with G.698.4 device to converge QKD into classical communication network and thus make full use ofexisting telecom infrastructure;and deploying quantum cryptography in the 6G network to achieve the overallsecurity management of the communi

20、cation system are introduced.Chapter 3.Quantum ComputingTo satisfy the dramatically increased communication system performance and rich diversity of innovative services5/35expected by 6G,Chapter 3 gives insight into the research of how to enhance communication by applying quantumcomputing.Firstly,co

21、nsidering that the essence of communication is a series of mathematical calculations,a hierarchicalcommunication network from a computing perspective is described to facilitate the analysis of computing scenariosand key issues for communication including signal processing,network optimization,servic

22、e processing,andnetwork intelligentization.Secondly,a Classical+Quantum hybrid computing platform with a robust computational foundation is proposedto provide computational support services tailored to different domains,facilitating research innovation and productimplementation.Especially,the archit

23、ecture design of this hybrid computing platform considers principles andconcepts of modularity,standardization,wide compatibility,autonomous security,and intelligence efficiency.Thirdly,the implications of quantum computing for 6G are introduced with three examples,which apply quantumcomputing to so

24、lve classical communication issues,respectively.The three examples include:solving single-cellmassive MIMO antenna optimization by using the Filtering Variational Quantum Algorithm(FVQE),solvingMIMO beam selection(MBS)by designed quantum algorithms based on Coherent Ising machines(CIM),solvingphase

25、correction of millimeter wave signals by applying a phase offset correction model obtained with QuantumSupport Vector Machine(QSVM)algorithm on the terminal side and thus reducing reference signaling overheads.Chapter 4.Future ExpectationIn the last quarter of 2023,we witnessed an industry milestone

26、 in the quantum area,i.e.,breaking the 1,000-qubitbarrier,giving quantum computers more computing power than ever before.Meanwhile,specialists from academiacreated a quantum computer with the largest-ever number of logical quantum bits i.e.,48 logical qubits,whereinthe logical qubits rather than the

27、 hardware-based qubits are promising to reduce the massive amounts oferror-correcting suffered by quantum computers.Consequently,2024 will probably mark a significant year forquantum computing technology,from when the field of quantum computing is expected to transition from physicalqubits to error-

28、correcting logic quantum bits,and anti-quantum cryptography research is expected to speed up aswell.6/352.Quantum Communication and Network2.1.Key Technologies2.1.1.Quantum Key DistributionQuantum communication is based on quantum superposition or entanglement to realize key distribution orinformati

29、on transmission,which is unconditionally secure at the theoretical level.Quantum key distribution(QKD)is the most developed quantum communication technology based on the basic principles of quantummechanics,combined with the encryption method of one encryption at a time to transfer the key betweenco

30、mmunication users.In 2023,various QKD theories and experiments have continuously been carried out,progress has been made in newprotocols and classical quantum co-transmission studies,etc.,and the performance of QKD systems has beenfurther improved.A joint team led by Tsinghua University gave a secur

31、ity proof of the device-independent QKD(DI-QKD)protocol by linking complementarity to quantum nonlocality and provided a new theoretical tool for thepractical implementation of DI-QKD1.A joint team led by the Australian National University(ANU)proposed ameasurement DI-QKD protocol that requires the

32、preparation of high-dimensional quantum states to be measuredusing the coherent total photon number method,and simulations shown that it can break the PLOB limit at shorterdistances than Twin-Field protocols when encoded in a 7-dimensional state2.A collaborative effort spearheaded bythe China Academ

33、y of Telecommunications Research(CATR)has successfully demonstrated a remarkable totaltransmission data capacity of 1Tbps within an optical transport network.This achievement was realized over100.96km through co-fiber transmission employing few-mode fiber,generating a quantum security key rate(SKR)o

34、f 2.7kbps3.QKD experiments using solid-state single-photon emitters are attracting increasing attention due totheir rapidly improving performance and compatibility with future quantum networks.The joint team led byHeriot-Watt University(UK)conducted QKD experiments using InGaAs quantum dots as a sin

35、gle-photon source,generating a finite key of 13 kbps at 100km,in one-minute acquisition time4.These research results are helpful inexploring QKD applications and realizing large-scale QKD network.1https:/doi.org/10.1103/PhysRevLett.131.1408012https:/doi.org/10.1038/s41534-023-00698-53https:/doi.org/

36、10.1364/OL.5004064https:/doi.org/10.1038/s41467-023-39219-57/35Currently,quantum communication systems,relying on QKD and other technical solutions,have beencommercially launched and implemented both domestically and internationally.Nonetheless,commercial QKDsystems still encounter numerous challeng

37、es concerning secure key rates,transmission distances,device size,andhigh costs.In commercial QKD systems,transmission is often achieved using prepare-and-measure QKD,whichcan be further classified into two types:continuous variable QKD(CV-QKD)and discrete variable QKD(DV-QKD).The advantage of CV-QK

38、D is that it could achieve high SKR over metro transmission distances using the classicalcommunication detection schemes.In 2023,Shanxi University adopted the discrete modulation CV-QKD togenerate 2.11 Mbps SKR over 80km5.Shanghai JiaoTong University used a transmitter-side light sourceintegration s

39、ystem to generate 0.75 Mbps SKR at 50 km6.The Technical University of Denmark used areceiver-side integrated scheme system to achieve 300 Mbps SKR at 10km7.The University of Waterloo gave asecurity proof of the finite key length of the discrete modulation CV-QKD and experimentally demonstrated thatt

40、he QKD transmission distance can be longer than 72km with 1012 key length8.The DV-QKD experimental system has undergone continuous development,resulting in certain enhancements toboth the SKR and transmission distance.In 2023,the group of applied physics from Geneva realized a SKR of 64Mbps over 10

41、km via time-bin encoding QKD using multipixel SNSPDs9.The research team led by University ofScience and Technology of China made achievements on both aspects,taking advantage of multipixel SNSPDs,anew-record SKR of 115.8 Mbps over 10 km fiber channel was obtained using a deceptive state based BB84 Q

42、KDprotocol10;adopting the 3-intensity sending-or-not-sending TF-QKD,relay-less QKD was realized over a 1002 kmfiber channel.These studies demonstrated that current techniques can satisfy the encryption requirements for highbandwidth communications and the feasibility in long distance communications.

43、For QKD industrialization,low-cost,mass-manufactured and practical QKD devices are required.From acommercial utilization perspective,the core devices of quantum communication,including the QKD encoder and5https:/doi.org/10.1364/OL.4920826https:/doi.org/10.1364/PRJ.4733287https:/arxiv.org/abs/2305.19

44、6428https:/doi.org/10.1103/PRXQuantum.4.0403069https:/doi.org/10.1038/s41566-023-01168-210https:/doi.org/10.1038/s41566-023-01166-48/35decoder,are moving towards miniaturization and cost-effectiveness.National Information OptoelectronicsInnovation Center from China Information and Communication Tech

45、nologies Group Corporation developedsilicon-based polarization state modulator and demodulator.Relying on the two modules,the qubit-based clocksynchronization and chip-based polarization compensation were demonstrated over 150km distance to achieve866bps SKR11.Researchers at the University of Geneva

46、,Switzerland,and the Institute of Photonics andNanotechnology,Italy,demonstrated a chip-based QKD system using a silicon-based transmitter chip supportinghigh-speed modulation and a polarization-independent low-loss receiver chip in aluminum borosilicate glass,toachieve a 1.3kbps over 151km12.2.1.2.

47、Quantum Random Number GeneratorQuantum Random Number Generator(QRNG)is a device that utilizes the principles of quantum physics togenerate true random numbers.Unlike traditional random number generators,QRNG generates true randomnumbers based on quantum optical principles,such as vacuum state noise,

48、quantum phase noise of laserspontaneous radiation,and photon number statistics.It stands as the sole genuinely theoretically defensible randomnumber generator to date,leveraging quantum mechanical uncertainty to guarantee the generation of highlyunpredictable and uncorrelated random numbers.QRNG has

49、 important applications.In cryptography,true randomnumbers are crucial for key generation,encryption algorithms and authentication,etc.QRNG can provide highersecurity against password cracking.However,it should be noted that QRNG only guarantees the true randomnessof the generated sequences and does

50、 not include the security of the distribution process.QRNG technology is currently being developed and improved.Many research institutes and companies arecommitted to researching and developing more efficient and stable QRNGs.In 2023,researchers from a joint teamled by Ghent University experimentall

51、y demonstrated an ultra-fast random number generation rate of 100 Gbit/s,setting a new record of an order of magnitude increase in the rate of QRNG based on vacuum fluctuation13.Quantum Dice(UK)announced the launch of its latest generation of APEX QRNG with post-processing randomnumber generation ra

52、tes of up to 7.5 Gbps14,which can also be integrated into existing infrastructures and havehigh security features.The German Federal Ministry of Education and Research funded the Chip-Based Quantum11https:/doi.org/10.1364/PRJ.48294212https:/doi.org/10.1364/PRJ.48147513https:/doi.org/10.1103/PRXQuant

53、um.4.01033014https:/www.quantum- Number Devices project15,which will develop a high-speed generation of random numbers based on thequantum photonic effects within a compact chip,meeting the Common Criteria for IT product security.With thefurther development of quantum technology,it is expected that

54、QRNGs will be utilized in a wider range ofapplications and contribute significantly to information security and scientific research.2.1.3.Quantum Information NetworkQuantum Information Networks(QIN)is a communication network system based on the principles of quantumphysics.It utilizes key technologi

55、es such as quantum entanglement manipulation,quantum teleportation,quantumrelay,etc.,aiming at realizing the functions of quantum long-distance communication,quantum computation,andquantum information interconnection network.QIN currently stands as a research hotspot within the quantuminformation fi

56、eld,representing the forefront of development in both communication and computation for the future.In recent years,many countries have been actively promoting the research and application of quantum informationnetworks.Many laboratories and research institutes have conducted a number of experiments

57、to verify thefeasibility and stability of QIN.In 2023,researchers at the University of Science and Technology of China andPeking University realized 51-qubit entanglement on the Zuchongzhi superconducting quantum computer platform,using high-fidelity parallel quantum gates,and realized 51-qubit one-

58、dimensional and 30-qubit two-dimensionalcluster states and achieved fidelities of 0.6370.030 and 0.6710.006,respectively16.A joint team from PekingUniversity has constructed a chip-based multi-dimensional quantum entanglement network.The network consistsof a central chip connected to three end chips

59、 by optical fiber,and the entanglement recovery and full connectivityhave been effectively realized at the end chips by using hybrid multiplexing technology,which lays the foundationfor the construction of large-scale and practical entanglement network17.NIST constructed the NG-QNet(NISTGaithersburg

60、 Quantum Network)testbed to characterize the function of the QIN base components18.The researchteam led by Lincoln Laboratory constructed a 50km three-node quantum network experimental bed(BARQNET)for testing quantum state signal transmission characteristics and compensation mechanisms19.The Univers

61、ity ofWaterloo will collaborate with Europe research team aiming at connecting Canada and Europe via a quantum15https:/www.ipms.fraunhofer.de/en/press-media/press/2023/Photonic-quantum-chip.html16https:/doi.org/10.1038/s41586-023-06195-117https:/www.science.org/doi/10.1126/science.adg921018https:/ww

62、w.nist.gov/programs-projects/quantum-communications-and-networks19https:/doi.org/10.48550/arXiv.2307.1569610/35satellite link20.The University of Florida,in collaboration with the University of Calgary,Canada,proposed andlaunched a quantum information network based on satellite relay21.Meanwhile,som

63、e companies are activelyengaged in the development of QIN.For example,Qunnect,in cooperation with New York University,testedsuccessfully a 16-kilometer QIN link using highly entangled quantum photons22.These efforts and collaborationsare expected to promote the development and application of QIN.2.2

64、.Standardization Activities for QKDIn recent years,major standardization organizations have actively carried out the preparation of QKD relatedstandards,including the China Communications Standardization Association(CCSA),the China CryptographyIndustry Standardization Technical Committee(CSTC),and t

65、he National Information Security StandardizationTechnical Committee(TC260);Internationally,there are the International Organization for Standardization(ISO),the International Telecommunication Union(ITU),and the European Telecommunications Standards Institute(ETSI).The content of the preparation has

66、 covered terminology definitions,application scenarios and requirements,network architecture,equipment technical requirements,QKD security,testing and evaluation methods,and otheraspects.2.2.1.Chinese Standardization ProgressChina Communications Standardization Association（CCSA）The China Communicati

67、ons Standardization Association(CCSA)is a standardization organization engaged in thefield of information and communication technology in China,conducting research on communication standardsystems.CCSA has established the 7th Special Task Group(ST7)for Quantum Communication and InformationTechnology

68、,which includes two sub working groups:the Quantum Communication Working Group(WG1)and20https:/uwaterloo.ca/news/science/connecting-canada-and-europe-through-quantum-satellite?utm_source=miragenews&utm_medium=miragenews&utm_campaign=news21https:/journals.aps.org/prapplied/abstract/10.1103/PhysRevApp

69、lied.20.02404822https:/www.nyu.edu/about/news-publications/news/2023/september/nyu-takes-quantum-step-in-establishing-cutting-edge-tech-hub-in-.html11/35the Quantum Information Processing Working Group(WG2).ST7 has initiated 25 standard development projectsin terms of terminology definition,applicat

70、ion scenarios and requirements,network architecture,equipmenttechnical requirements,QKD security,and testing and evaluation methods.Among them,the national standardGB/T 42829-2023“Basic requirements for quantum secure communication applications”was officially issued inAugust 2023.12 other communicat

71、ion industry standards have also been officially promulgated and implemented:YD/T 4632-2023 Technical requirements for quantum key distribution and classical optical communication cofiber transmissionYD/T 3835.2-2023 Test methods for quantum key distribution（QKD）systems Part 2:QKD system based onGau

72、ssian modulated coherent state protocolYD/T 4410.1-2023 Quantum Key Distribution(QKD)Network Ak Interface Technical Requirements Part 1:Application Program Interface(API)YD/T 3834.2-2023 Technical requirements for quantum key distribution(QKD)systems Part 2:QKD systemsbased on Gaussian modulation co

73、herent state protocolYD/T 4303-2023 Technical specification of quantum secure communication application equipment based onIPSec protocolYD/T 4302.1-2023 Technical specification for quantum key distribution(QKD)network management Part1:NMS system functionYD/T 4301-2023 Quantum secure communication ne

74、twork architectureYD/T 3907.2-2022 Key components and modules for Quantum Key Distribution(QKD)based on BB84protocolPart 2:Single photon detectorYD/T 3907.1-2022 Key components and modules for Quantum Key Distribution(QKD)based on BB84protocolPart 1:Laser sourceYD/T 3907.3-2021 Key components and mo

75、dules for Quantum Key Distribution(QKD)based on BB84protocol-part 3:Quantum Random Number Generator(QRNG)YD/T 3835.1-2021 Test methods for Quantum Key Distribution(QKD)system-Part1:Decoy state BB84protocol QKD systemYD/T 3834.1-2021 Technical requirements for quantum key distribution(QKD)system-Part

76、1:Decoy stateBB84 protocol QKD system12/35China Cryptography Industry Standardization Technical Committee（CSTC）QKD technology involves the generation,management,and use of passwords.The China Cryptography IndustryStandardization Technical Committee(CSTC)has conducted research on password industry st

77、andards such asQKD technical specifications and evaluation systems.At present,the following two quantum related cryptographyindustry standards have been officially released:GM/T 0108-2021 Decoy-state BB84 quantum key distribution product technology specificationGM/T 0114-2021 Decoy-state BB84 quantu

78、m key distribution product test specificationNational Information Security Standardization Technical Committee（TC260）The National Information Security Standardization Technical Committee(TC260)is a technical organizationengaged in information security standardization work in the field of information

79、 security technology in China,responsible for organizing and carrying out standardization technology work related to domestic informationsecurity.TC260 undertakes the corresponding business work of information security related internationalstandardization organizations such as ISO/IEC JTC1/SC27.The

80、two international standard proposals in the field ofquantum secure communication,ISO/IEC 23837-1 Security Requirements,Testing and Evaluation Methods forQuantum Key Distribution Part 1:Requirements and ISO/IEC 23837-2 Security Requirements,Testing andEvaluation Methods for Quantum Key Distribution P

81、art 2:Testing and Evaluation Methods,driven by TC260 andled by China,have been officially released.2.2.2.International Standardization ProgressInternational Organization for Standardization（ISO）The International Organization for Standardization(ISO)is currently the largest and most authoritativeinte

82、rnational standardization agency in the world,and the International Electrotechnical Commission(IEC)isresponsible for international standardization work in the fields of electrical and electronic engineering.ISO/IECJTC1(First Joint Technical Committee of the International Organization for Standardiz

83、ation/InternationalElectrotechnical Commission)is an international standardization committee in the field of information technology,responsible for the international standardization of information technology.Among them,the Information13/35Technology Security Technology Subcommittee under ISO/IEC JTC

84、1-Information Security,Network Security andPrivacy Protection Subcommittee(ISO/IEC JTC1 SC27)is responsible for the standardization of general methodsand technologies for information technology security,including determining general requirements for informationtechnology security services,developing

85、 security technologies and mechanisms,proposing security guidelines,andpreparing supportive documents and standards for management.SC27 officially released two quantum keydistribution related technical standards led by China in 2023:ISO/IEC 23837-2:2023 Information security Security requirements,tes

86、t and evaluation methods forquantum key distribution Part 2:Evaluation and testing methodsISO/IEC 23837-1:2023 Information security Security requirements,test and evaluation methods forquantum key distribution Part 1:RequirementsInternational Telecommunication Union（ITU）The International Telecommuni

87、cation Union is a United Nations agency responsible for information andcommunication technology affairs,responsible for developing global telecommunications standards.Thetelecommunications standardization department under ITU maintains a high level of attention to the developmentand evolution of qua

88、ntum information technology and its future impact on information communication networksand industries.The standardization work related to quantum communication in ITU-T is currently at the forefrontof the world.The quantum communication related technical standards of ITU-T include the Q series-Switc

89、hing and signaling,and associated measurements and tests,X series-Data networks,open system communications and security,and Yseries-Global information infrastructure,Internet protocol aspects,next-generation networks,Internet of Thingsand smart cities.The quantum communication related parts in the Q

90、 series standards are Q.4160-Q.4179:Protocols and signaling forQuantum key distribution networks.The following standards have been officially released so far：ITU-T Q.4160(12/2023):Quantum key distribution networks Protocol frameworkITU-T Q.4161(12/2023):Protocols for Ak interface for the quantum key

91、 distribution networkITU-T Q.4162(12/2023):Protocols for Kq-1 interface for the quantum key distribution network14/35ITU-T Q.4163(12/2023):Protocols for Kx interface for the quantum key distribution networkITU-T Q.4164(12/2023):Protocols for Ck interface for the quantum key distribution networkThe q

92、uantum communication related parts in the X series standards are X.1700-X.1729:Quantum communication.The following standards have been officially released so far：X.1702:Quantum noise random number generator architectureX.1710:Security framework for quantum key distribution networksX.1712:Security re

93、quirements and measures for quantum key distribution networks key managementX.1714:Key combination and confidential key supply for quantum key distribution networksX.1715:Security requirements and measures for integration of quantum key distribution network and securestorage networkThe quantum commu

94、nication related parts in the Y series standards are Y.3800-Y.3999:Quantum key distributionnetworks.The following standards have been officially released so far：Y.3800:Overview on networks supporting quantum key distributionY.3801:Functional requirements for quantum key distribution networksY.3802:Q

95、uantum key distribution networks Functional architectureY.3803:Quantum key distribution networks Key managementY.3804:Quantum key distribution networks Control and managementY.3805:Quantum key distribution networks Software-defined networking controlY.3806:Quantum key distribution networks Requireme

96、nts for quality of service assuranceY.3807:Quantum key distribution networks Quality of service parametersY.3808:Framework for integration of quantum key distribution network and secure storage networkY.3809:A role-based model in quantum key distribution networks deploymentY.3810:Quantum key distrib

97、ution network interworking FrameworkY.3811:Quantum key distribution networks Functional architecture for quality of service assuranceY.3812:Quantum key distribution networks-Requirements for machine learning based quality of serviceassuranceY.3813:Quantum key distribution network interworking Functi

98、onal requirementsY.3814:Quantum key distribution networks functional requirements and architecture for machine learning15/35enablementY.3815:Quantum key distribution networks Overview of resilienceY.3816:Quantum key distribution networks Functional architecture enhancement of machine learning basedq

99、uality of service assuranceY.3817:Quantum key distribution network interworking Requirements for quality of service assuranceY.3818:Quantum key distribution network interworking ArchitectureY.3819:Quantum key distribution networks-Requirements and architectural model for autonomicmanagement and cont

100、rol enablementEuropean Telecommunications Standards Institute（ETSI）The European Telecommunications Standards Institute(ETSI)is an independent non-profit regional informationand communication technology standardization organization in Europe.ETSI established the ISG-QKD standardgroup as early as 2008

101、 to explore QKD standardization.ETSI has released 12 technical specifications,includingterminology definitions,protection profiles,system components,application interfaces,security certificates,deployment parameters,etc.Among them,the second version of application interfaces,components and internali

102、nterfaces,and control interfaces has been released:ETSI GS QKD 016 V1.1.1(2023-04)Quantum Key Distribution(QKD):Common Criteria Protection Profile-Pair of Prepare and Measure Quantum Key Distribution ModulesETSI GS QKD 018 V1.1.1(2022-04)Quantum Key Distribution(QKD):Orchestration Interface for Soft

103、wareDefined NetworksETSI GS QKD 015 V2.1.1(2022-04)Quantum Key Distribution(QKD):Control Interface for SoftwareDefined NetworksETSI GS QKD 004 V2.1.1(2020-08)Quantum Key Distribution(QKD):Application InterfaceETSI GS QKD 014 V1.1.1(2019-02)Quantum Key Distribution(QKD):Protocol and data format ofRES

104、T-based key delivery APIETSI GS QKD 012 V1.1.1(2019-02)Quantum Key Distribution(QKD):Device and CommunicationChannel Parameters for QKD DeploymentETSI GR QKD 007 V1.1.1(2018-12)Quantum Key Distribution(QKD):VocabularyETSI GR QKD 003 V2.1.1(2018-03)Quantum Key Distribution(QKD):Components and Interna

105、l Interfaces16/35ETSI GS QKD 011 V1.1.1(2016-05)Quantum Key Distribution(QKD):Component characterization:characterizing optical components for QKD systemsETSI GS QKD 008 V1.1.1(2010-12)Quantum Key Distribution(QKD):QKD Module Security SpecificationETSI GS QKD 005 V1.1.1(2010-12)Quantum Key Distribut

106、ion(QKD):Security ProofsETSI GS QKD 002 V1.1.1(2010-06)Quantum Key Distribution(QKD):Use Cases2.3.Implications for 6G2.3.1.Quantum Encryption in the Internet of VehiclesChina Unicom has built a Quantum key cloud platform which obtains quantum keys from QKD or QRNG,storesand manages the key safely.Th

107、rough the security mechanism,the quantum keys can be distributed to the usersecurity terminal and provide high-level security protection.Utilizing this Quantum key cloud platform,we havecarried out some quantum encryption demonstrations such as quantum encrypted call,quantum public networkcluster in

108、tercom,quantum video conference,and quantum UAV patrol.Recently,we demonstrate the quantum encryption in the internet of vehicles scenario.In this scenario,thecommand and dispatching signals require high-level security to ensure the safety of vehicles under autonomousdriving or remote driving.We dep

109、loy the quantum key cloud platform in the internet of vehicles,and connect to theintelligent transportation service cloud platform,autonomous parking system and remote driving cockpit throughprivate network.The vehicles and road side unit(RSU)can be connected through bearer network.We can integratet

110、he quantum SDK to the platform and systems or deploy the quantum box by the terminals side which can obtainthe quantum keys from the quantum key cloud platform and perform quantum encryption to the command anddispatching signals.The quantum SDK can also be integrated to the wireless communication mo

111、dule whichprovides network access for the vehicles.The security terminals and platform use the pre-charged quantum keys forthe identity authentication and initial encryption.Internet of vehicles is very sensitive to network transfer delay because of the high speed of vehicles.So we test theextra del

112、ay due to encryption/decryption and demonstrate no big influence to the original service.17/35Figure 1 Quantum encryption in the internet of vehicles scenario2.3.2.Quantum Encryption Integration with Bearer Network EquipmentUsually QKD network needs establishment of extra fibers which is also very e

113、xpensive to deploy in the bearernetwork.We want to converge QKD into classical communication network,which can save the fiber resource andutilizing existing telecom infrastructure.The G.698.4 system is a multichannel bi-directional DWDM system whichcan be used in fronthaul network,metro bearer netwo

114、rk.Low-cost C-band tunable laser is utilized,andbi-directional OD/OM/OADM is used.For metro or access application,the maximum reach is about 20km,and nooptical amplifier is used.The G.698.4 system doesnt contain optical amplifier,which is not allowed in the QKDsystem.Besides,CV-QKD is a novel techno

115、logy which is more robust to the photon noise induced by the classicalcommunication and much cheaper than DV-QKD.So its very convenient to integrate CV-QKD with G.698.4device.We integrate CV-QKD with G.698.4 device as shown in Figure 2.This device includes QKD module,opticalcommunication module,WDM

116、module,encryption module and control module.and The quantum signal ofCV-QKD is multiplexed with wavelengths of G.698.4 system with WDM module and transmit in the same fiber.After co-propagation,the quantum signal is separated by WDM module at the receiver.The quantum key producedby QKD module is sen

117、t into the encryption module which can transform the data in plaintext to ciphertext.We alsointegrate the QKD control function into G.698.4 to achieve the encryption parameter setting such as encryption18/35method,quantum key valid period.Figure 2 Integration of CV-QKD with G.698.4 device2.3.3.Quant

118、um Communication SecurityQuantum communication is a new type of communication that utilizes the principle of quantum mechanics forinformation transmission,and utilizes the special properties of quantum states,such as quantum superpositionstates and quantum entanglement,to realize the secure transmis

119、sion of information.Quantum cryptography is the key technology of quantum communication security as the basis of quantumcommunication technology.Quantum cryptography refers to the use of quantum states as the key technology forencryption and decryption of information,in which quantum key distributio

120、n(QKD)combines quantum mechanicsand cryptography of quantum cryptography.Quantum cryptography ensures the security of the key through the useof quantum states to generate and distribute encryption keys.Quantum cryptography mainly realizes confidential transmission through quantum keys,and data confi

121、dentiality isrealized through corresponding single photon quantum exchange.In the quantum key layer,the quantum key willprovide each user with the corresponding quantum key communication service.In the actual transmission process,the quantum form of single photon is difficult to be observed and copi

122、ed,and in the transmission process onesecret at a time method is used.Therefore,the decryption of data information can be realized only after bothparties use their own keys at the same time,which can improve the confidentiality of data information in thetransmission process and ensure that the data

123、information will not be stolen and tampered with.By deploying quantum cryptography in the 6G network,the overall security management of the communicationsystem is realized.Quantum cryptography taking quantum key as the main means of confidentiality,distributes19/35quantum key for users with correspo

124、nding authority in 6G communication,carries out non-sequential substitutionand morphological changes during data storage and transmission,so as to realize end-to-end security guarantee of6G communication.20/353.Quantum Computing3.1.Computing Scenarios and Key Issues for CommunicationThe essence of c

125、ommunication is a series of mathematical calculations.From a computational perspective,communication networks are simply divided into physical layer,network layer,and application layer,as shown inFigure 3.Among them,the physical layer is mainly responsible for communication signal processing,the net

126、worklayer is responsible for topology,access,routing,resource management,and the application layer is mainlyresponsible for service optimization and traffic management.In order to enhance processing performance,machinelearning(ML)is introduced in each layer and becomes a special and important comput

127、ing scenario incommunication.In addition,the security of each layer has always been the default computing scenario.Figure 3 Hierarchical Communication Network from Computing Perspectives3.1.1.Signal ProcessingCommunication signal processing is the underlying computation of communication.Taking wirel

128、ess signalprocessing as an example,signal processing involves the transformation,filtering,encoding,decoding,modulation,demodulation,transmission,detection,estimation,and interference coordination of wireless signals at both ends ofthe transmitter and receiver.For large-scale Multiple Input Multiple

129、 Output(MIMO)systems,Large BandwidthOrthogonal Frequency Division Multiplexing(OFDM)systems,and large-scale terminal access systems,thecomputational complexity of signal processing such as channel estimation,precoding,signal detection,and channelencoding and decoding significantly increases.Large sc

130、ale MIMO signal processing.Large scale MIMO signal processing exists in large-scale antenna array21/35systems,distributed antenna systems,and cellular free systems,involving channel estimation,precoding,signaldetection and other processing processes,including matrix multiplication,inversion,tensor p

131、roduct,conjugatetransposition,decomposition and other high-dimensional matrix operations.These basic operations require asignificant amount of computing resources,posing significant challenges to system design.At present,in order tosolve this computational problem,methods such as compressive sensing

132、 or key parameter estimation are usuallyused to achieve effective signal processing with relatively small computational costs,based on fully exploring thesparse characteristics of high-dimensional signals.Massive terminal access signal processing.The number of terminals(users)connected in wireless c

133、ommunicationsystems is increasing day by day.In the case of multiple terminals sharing access resources,the dimension ofwireless access signals will increase with the increase of the number of terminals,which brings difficulties tochannel estimation,multi-user signal detection,and interference coord

134、ination.High frequency and large bandwidth signal processing.Millimeter wave,terahertz,and visible light frequencybands can provide larger bandwidth,but large broadband wireless signals bring larger matrix operations andchannel encoding and decoding,especially the complexity of long codes.On the oth

135、er hand,high-frequency highbandwidth wireless systems will also be used for target ranging,speed measurement,angle measurement and otherpositioning scenarios,bringing the need for radar signal calculation.3.1.2.Network OptimizationThe overall goal of network optimization is to improve customer satis

136、faction through optimization methods such asnetwork topology,functionality,service,parameters,and resources.Network topology optimization refers to minimizing the overall network construction cost while meeting therequirements of overall traffic transmission and disaster recovery backup.Network topo

137、logy optimization occurs inevery link of network planning,network construction,and network operation and maintenance.Related to this isrouting optimization.Network coverage optimization refers to maximizing network coverage through network parameter configuration,mainly including blind spot,weak spo

138、t,overlapping spot,and pilot interference spot optimization.For large-scaleantenna systems,there are many parameters to be optimized for signaling beams and data beams,with a large22/35optimization space and complex problems.Network capacity optimization refers to the rational allocation of user tra

139、ffic in network resources to maximizesystem capacity.Wireless network capacity optimization mainly includes single station multi-user access control,multi-user scheduling,and load balancing.Network energy efficiency optimization refers to meeting given service requirements with minimal energyconsump

140、tion costs,with a focus on optimizing power control under minimum rate constraints,base station/carrierswitching,and computing task offloading/migration.3.1.3.Service ProcessingService processing mainly refers to source signal processing and service optimization in the network.The growthof large-sca

141、le multimodal services has put forward higher demands for computing power.Source signal processing is a series of computational processes that involve sampling,quantization,representation,encoding,compression,transmission,and reconstruction of source content such as images,videos,speech,and text.Wit

142、h the gradual growth of the metaverse service,higher computational power is required for 3D videopre-production or real-time rendering.AI services such as natural language processing,computer vision,andspeech recognition,especially those based on large models,have recently driven explosive growth in

143、 computingpower demand.At the same time,semantic communication technology replaces symbol representation withsemantic representation,providing a new source coding transmission method and bringing new computationalscenarios for source signal processing.service optimization refers to the adjustment of

144、 network and service equipment,functions,and parameters to matchnetwork status with service status,ensuring end-to-end service quality.The focus of service optimization includestraffic prediction,traffic optimization,user behavior prediction,content distribution,cache optimization,servicemigration,a

145、nd service parameter optimization.3.1.4.Network IntelligentizationMachine learning transforms the computation of original problems in application scenarios into computation inmachine learning,providing a new algorithmic paradigm for signal processing,network optimization,and service23/35optimization

146、.Intelligent signal processing.The application of AI technology for signal processing has been widely carried out inthe field of communication.Among them,supervised learning with regression and classification capabilities is usedfor channel parameter detection and estimation,modulation mode detectio

147、n and classification,spectrum sensingand detection.Unsupervised learning that can cluster and reduce the dimensionality of signals can be used toreduce the dimensionality of high-dimensional communication signals;Reinforcement learning is good atdecision-making and prediction,and can be used for spe

148、ctrum perception and sharing;Deep learning can classify,estimate,and eliminate interference between communication signals,as well as complete numerous signalprocessing related tasks such as channel estimation,signal detection,and beam management.These AI methodshave unique advantages in big data ana

149、lysis,efficient parameter estimation,and interactive decision-making,butthere are problems with high complexity model training and large parameter estimation,which pose highrequirements for the computing power of communication systems.Intelligent network optimization.Transform optimization problems

150、into model training and inference calculationsbased on machine learning models and algorithms.Network and AI can also achieve deep integration at thearchitecture level.Intelligent service processing.Almost all service optimization problems can be solved and enhanced by introducingmachine learning,su

151、ch as traffic detection,classification and prediction,content distribution and cacheoptimization,user behavior feature analysis,service parameter optimization,and so on.3.2.Quantum Hybrid Heterogeneous ComputingComputational power stands as one of the pivotal factors driving the advancement of the a

152、rtificial intelligence(AI)industry.The training of large-scale deep learning models entails substantial computational costs,rendering itchallenging for numerous enterprises and research organizations to sustainably access such resources fordevelopment.Quantum computing,on the other hand,holds the po

153、tential to augment AI computation capabilitiesacross various dimensions including theory,paradigms,hardware,algorithms,and applications.This augmentationsignificantly enhances training efficiency while mitigating computational expenses.Furthermore,in specific orcomputationally intractable problem do

154、mains such as combinatorial optimization,simulation,and machine learning,24/35quantum computing naturally possesses advantages,enabling the effective harnessing and expansion ofcomputational resources beyond classical computing clusters.Therefore,in response to the pervasive demand forcomputational

155、resources from a diverse array of application-driven enterprises and research institutions,aClassical+Quantum hybrid computing platform is poised to deliver technologically advanced and economicallyaccessible computational services.The Classical+Quantum hybrid computing platform comprises two compon

156、ents on the hardware level:quantumcomputing and classical computing.On the software front,it encompasses both quantum software platforms andclassical computing software platforms.The overall architecture of the solution should adhere to principles and concepts of modularity,standardization,wide comp

157、atibility,autonomous security,and intelligence efficiency.Effectively designing and configuring thehardware structure of the system ensures alignment with the requirements and characteristics of the software,maximizing the capabilities of the hardware,enhancing computational efficiency,and meeting t

158、he demands forfuture development and system upgrades.The core system architecture comprises both the hardware infrastructure and software infrastructure of the coresystem.Firstly,the core hardware infrastructures main chips and accelerator chips can adopt a combination ofcommercially available chips

159、 and domestically developed chips to ensure the security of the chip supply chain andapplication ecosystem in the complex and dynamic international environment.The main chips utilize architecturescompatible with x86(including domestically produced Haiguang x86 processors),while the accelerator chips

160、employ architectures compatible with mainstream GPU ecosystems.This approach balances internationallyrecognized hardware with domestically developed,controllable core hardware that boasts excellent compatibilitywith ecosystems.It can widely accommodate a vast array of mature application software and

161、 AI frameworks,facilitating the integration of various types of computational resources to meet diverse computing modes such as AItraining,inference,numerical simulation,big data processing,quantum computing,and others.This seamlesscompatibility across a multitude of application scenarios reduces ap

162、plication development costs.Additionally,thehardware infrastructure layer employs a diverse range of computing devices capable of supplying computationalresources as per the differentiated computational characteristics of various applications,thus enabling flexible25/35resource allocation.Furthermor

163、e,the core systems software infrastructure layer needs to integrate various computing frameworks suchas artificial intelligence,high-performance computing,and big data.Through the management of the computingservice middleware,it achieves functions such as workspace management,resource management,res

164、ourcescheduling,application center,billing management,permission management,and user management.In terms ofcomputing,the AI computing platform needs to support two different application scenarios:training and inference.It should support various management functions including dataset management,hyper

165、parameter tuning,modelmanagement,model development,container services,image repositories,task measurement,and applicationdeployment.It should support mainstream AI computing frameworks such as TensorFlow,PyTorch,PaddlePaddle,as well as AI algorithm development platforms.For supercomputing,it require

166、s a rich set of basic softwareenvironments,including compilers,math libraries,automated configuration tools,and software tuning tools.Cloudservices need to support elasticity scaling and other features.In the field of quantum computing,it is necessary toutilize next-generation quantum simulation tec

167、hnology to provide efficient and reliable quantum computingsimulation services on traditional computing hardware,thus improving computational efficiency.A unifiedtechnical architecture is conducive to supporting a diverse range of application scenarios,promoting innovation inupper-layer applications

168、,fostering a thriving application ecosystem,and attracting related enterprises and talents.The Classical+Quantum hybrid computing platform,based on its robust computational foundation,providescomputational support services tailored to different domains,facilitating research innovation and productimp

169、lementation.The computing service middleware offers integrated solutions for both hardware and software,featuring AI processing chips at the hardware level and pre-installed various AI frameworks and toolkits within theplatform.This setup enables users to conveniently complete algorithm porting,adap

170、tation,development,andtesting.Adapted applications can be packaged and deployed via container images through the applicationmanagement platform,reducing dependencies on deployment environments.Additionally,it allows for theallocation and scheduling of underlying computational resources based on appl

171、ications computational demands.Furthermore,it provides full lifecycle management capabilities for applications,including creation,upgrade,pause,and termination,ensuring efficient and stable application operation.26/353.3.Implications for 6GThe key impact of quantum computing lies in three aspects:ne

172、twork,machine learning,and security,as shown inFigure 4.Quantum optimization,quantum search,quantum signal processing,and quantum machine learning canenhance network capabilities and service quality.Focus on designing quantum computing algorithms for networkoptimization and network intelligence.Figu

173、re 4 The Impacts of Quantum Computing on Communication3.3.1.Single-Cell Massive MIMO Antenna OptimizationRequirement analysisConsider the case of static cell optimization,the optimization objective of this problem is to maximize the coverageratio within a single cell,and the variables to be optimize

174、d include horizontal azimuth angle,elevation azimuthangle(or tilt angle),horizontal beam width,and vertical beam width.Consider a single-cellular model,assuming a base station is located at(0,0).The users in the cellular are distributedwithin a sector range centered on the base station,with a radius

175、 of 300 meters and a 120.There are some samplingpoints in the cellular,randomly and uniformly distributed within the area.Coverage ratio is the ratio of the numberof sampling points(i.e.users)with RSRP greater than the threshold value to the number of all sampling points inthe cellular.If RSRP is gr

176、eater than the threshold value,it is considered that the sampling point can be effectively27/35covered by the antenna beam.direction of beamBSuserbeamxyzFigure 5 Single-cell massive MIMO antenna optimizationThe range of variable values is as follows:-Horizontal beam width:15,25,45,65,90,105,110-Vert

177、ical beam width:6,12,25-Horizontal azimuth angle:-30:1:30-Elevation azimuth angle:-15:1:15This problem is a NP-hard combinatorial optimization problem that can be solved with quantum algorithms.Solution designAbove problem is solved using the Filtering Variational Quantum Algorithm(FVQE).By introduc

178、ing thefiltering operator,the quantum state is evolved intoF f H,|where f(H,)is a monotonically decreasing function,which achieves probability amplification of the ground state.Using a variational algorithm with parameters(such as hardware efficient simulation),approximate the filteringoperator with

179、 the help of the parameter shift rule.If the weight optimization problem of large-scale MIMOantennas is re modeled as a Hamiltonian ground state problem,it can be solved using this algorithm.28/35This algorithm uses hardware efficient simulation,so there is no need to worry too much about line optim

180、izationand line mapping.The basic quantum gates supported by the original quantum chip Wukong are CZ,R,where is the angle between a rotation axis on the XY plane and the X positive direction.Through pyqpandas built-incompilation algorithm,any SU(2)gate can be converted into two gates with any rotati

181、on axis,as well as a Vrital Zgate.More details about FVQA are shown in the reference23.Performance Simulation Verification and AnalysisThe simulation results show that the algorithm can optimize the coverage ratio,and the effect is comparable toclassical optimization algorithms such as quantum parti

182、cle swarm optimization.The algorithm was tested using theOriginal Wukong 72 bit quantum computer.To ensure reliable performance,6 quantum bits coupled in a chain inthe chip were selected for testing.The probability of the algorithm obtaining the optimal solution reaches 70.34%,achieving the predeter

183、mined goal of maximizing coverage.3.3.2.MIMO Beam Selection of Multiple CellularRequirement AnalysisMassive MIMO can provide multiple data streams simultaneously by utilizing a large number of antennas andbeamforming,thereby achieving higher throughput and better signal quality.This is expected to i

184、mprove thecoverage and capacity of cellular networks.However,due to the high mobility of users and inter cell interference,traditional relatively static beamforming settings can no longer meet the dynamic requirements of networkcoverage.The problem of MIMO beam selection(MBS)is becoming increasingly

185、 prominent.MBS refers toselecting a set of beams under given constraints to maximize network performance,such as improving signalquality and system throughput.Specifically,in the MBS problem,the target geographic area is usually divided intogrids,and each beam has a reference signal receiving power(

186、RSRP)value on the corresponding grid.The RSRP ofa grid is defined as the maximum RSRP value received on that grid.The MBS problem is to find a set of beams foreach cell to maximize the RSRP of each grid.The MBS problem is an NP-hard combinatorial optimization problem,23Amaro,David et al.“Filtering v

187、ariational quantum algorithms for combinatorial optimization.”Quantum Science&Technology(2022)7 015021.29/35especially in 5G systems with a large number of cells and antennas.For example,when there are hundreds ofbeams in multiple units,it is difficult to find the best solution from billions of beam

188、 combinations.For combinatorial optimization problems,there are classic algorithms such as greedy algorithms,branch and boundalgorithms,and simulated annealing algorithms.Greedy algorithms are simple and efficient,but they may fall intolocal optima.The branch and bound algorithm can ensure global op

189、timality,but the computational cost may behigh.Simulated annealing is a metaheuristic optimization algorithm that gradually cools down the temperature toencourage optimization convergence to the global optimum,but this cannot guarantee.Coherent Ising machines(CIM)has been applied in multiple scenari

190、os,including compression sensing and job scheduling problems.Theresearch group will design quantum algorithms based on CIM to solve the MBS problem.Problem and SolutionIn the MBS problem,the target coverage area is divided into grids,as shown in Figure 6,with each grid covered byseveral small cells.

191、Each cell has a set of MIMO beams,and the MBS problem is to select a certain number ofbeams from each cell to maximize the number of grids that meet specific constraints.If the maximum RSRP in thegrid exceeds the given threshold and the difference between the maximum RSRP and the second maximum RSRP

192、in the grid exceeds the given value.The RSRP from the cell to the grid is determined by the maximum RSRPamong all beams.The reason for setting a threshold for the difference between the maximum signal strength andthe second maximum signal strength is that in MIMO systems,there is mutual interference

193、 between beams.If thesignal strength of multiple beams is similar,it may cause signal interference and reduce the performance of thereceiver.30/35Figure 6 Problem of MBSThis problem can be formulated in a quadratic unconstrained binary optimization(QUBO)form and the moredetails are shown in the refe

194、rence24.Export the ising matrix based on the QUBO model,input the ising matrix intoCIM,and run CIM to solve the optimization problem.CIM Simulation Verification and AnalysisThe purpose of CIM simulation is to evaluate the feasibility of using CIM as a solver for MBS problems,that is,whether the qual

195、ity of the solution and the size of the problem are suitable for CIM.As shown in Figure 7,the working mode of CIM is different from traditional computers that rely on semiconductorintegrated chips.On the contrary,it uses laser pulses in optical fibers as the basic computing unit,called qubits.Theini

196、tial research focused on the idea of injecting synchronous laser Ising machines.As the number of coupled lasersincreases proportionally to the square of quantum bits,an improved scheme using nonlinear optical crystals isproposed based on a degenerate optical parametric oscillator(DOPO).Two DOPO base

197、d methods have beendeveloped,namely optical delay line CIM and measurement feedback CIM.The requirements for load and precisecontrol in the first solution are unacceptable.The research group adopts the second approach,which is provided byBeijing Bose Quantum Technology Co.,Ltd.for CIM.24Huang,Yuhong

198、 et al.“Quantum Computing for MIMO Beam Selection Problem:Model and Optical ExperimentalSolution.”GLOBECOM 2023-2023 IEEE Global Communications Conference(2023):5463-5468.31/35Figure 7 Principal diagram of CIMTable 1 presents a comprehensive performance comparison between CIM physical machine,simula

199、ted annealing(SA),and Tabu search.The CIM physical machine consistently achieved the optimal value of the objectivefunction for the algorithm implementation in all cases.In addition,the Hamiltonian of the solution discovered bythe CIM physical machine is very close to the Hamiltonian of the optimal

200、solution,which confirms theeffectiveness of this method in simplifying the model and searching for solutions close to the lowest energy value.Table 1 Comparison of CIM physical machine,simulated annealing,and Tabu search performance3.3.3.Phase Correction of Millimeter Wave SignalsRequirement Analysi

201、sWireless transmission on the 5G millimeter wave band is more sensitive to phase shift.At this point,a TrackingReference Signal(TRS)signal is needed to track the phase rotation of the data signal and perform correction andcompensation.This method can achieve more accurate performance compensation.Bu

202、t it requires continuoussending of reference signals for measurement,which requires significant signaling overhead.With the continuous development of machine learning,AI algorithms can be used to solve signal correction32/35problems,such as support vector machines.However,support vector machines req

203、uire a significant amount ofstorage and computation time when the sample size is large.Quantum Support Vector Machine(QSVM)algorithmis a machine learning algorithm based on quantum computing,which has the same functions as classical algorithms,lower computational complexity,and faster computing spee

204、d.This algorithm requires the corresponding capability reporting and parameter signaling notification.The basestation issues a phase offset correction model based on the computing power of the terminal,and there is no need tocontinuously transmit the TRS.The terminal can directly correct the signal

205、phase offset based on the model,thereby reducing dependence on the TRS and reducing RS overheads.Method DescriptionThe method description is as follows:Step 1:The UE receives the control signaling/configuration rule sent by the BS to determine the frequency offsetcorrection method used at the UE.The

206、 UE determines the correction method used and performs phase correctionaccording to the instructions.The correction methods include:using reference signal correction,using receiveralgorithm correction,and jointly correcting the above two algorithms.By configuring the correction method,thenetwork can

207、 consider service requirements and terminal capabilities,select the most suitable correction method,consider signaling costs and latency while considering the complexity of the receiving end,and select the mostsuitable correction method.Step 2:UE reports the terminal capability information to the BS

208、.The configuration of network correction methodsbased on first ability information;The first capability information refers to the frequency offset correction schemethat UE can support,and can also include one or several types of terminal capability information:1)the processingtime required by the te

209、rminal correction algorithm;2)The size of the processing problem at the terminal(in termsof quantum bits);3)Decoherence time;4)quantum gate fidelity.The correction method is determined according to the control signaling,which is used to directly notify the UE.According to the configuration rules,the

210、 correction method can be determined by the correspondence between theUE receiving the first information configured on the network side and the correction method,and then thecorrection method can be determined based on the first information and the correspondence.The first information33/35can be MCS

211、 index or processing latency requirements.Step 3:The training model is completed on the network side,and the network issues a suitable algorithm model tothe UE through RRC signaling based on the first capability information.The specific method is:There is a training dataset on the network side.The t

212、raining set data includes feature vectors and labels.Forexample,the feature vector represents the signal strength and phase of the modulated signal through the channel,and the label indicates the class to which the modulated signal belongs.It can be used as training set data byreceiving M known upli

213、nk signals,such as known uplink signals.Construct M linear equation systems based on thetraining set data,implement an SVM classifier,and then use the HHL algorithm to solve the equation system,trainthe SVM classifier,and obtain the training model;Train multiple models with different sample sizes,an

214、d distributeappropriate training models to the terminal side based on UE capability information;Input the received signal intothe trained classifier model,classify the signal and perform offset correction,output its corrected signal,andcomplete signal correction.Step 4:The number of repeated measure

215、ments in the receiving algorithm affects the accuracy of phase correction,which needs to be determined based on the indicated information.The receiving end can confirm the number ofrepeated measurements based on the first indication information.The first indication information can be a signalingdire

216、ctly indicating the number of repeated measurements.Alternatively,it can be configured in conjunction withother signaling systems,such as MCS.(RRC needs to issue a new MCS indicator table for joint configuration)34/354.Future ExpectationIn the last quarter of 2023,we witnessed an industry milestone

217、in the quantum area,i.e.,crossing the 1,000-qubitthreshold,giving quantum computers more computing power than ever before.In November of 2023,Quantum startup Atom Computing25announced that it has created a 1,225-siteatomic array in its next-generation quantum computing platform.In December of 2023,I

218、BM announced that it has produced a quantum system based on a chip namedCondor,which is the largest transmon-based quantum processor yet released,with 1,121 functioningqubits26.In addition,a joint research team of specialists from Harvard University,QuEra Computing Inc.,the University ofMaryland and

219、 MIT,has created a quantum computer with the largest-ever number of logical quantum bits27,i.e.,48logical qubits.An approach based on logical qubits rather than hardware-based qubits is promising to reduce themassive amounts of error-correcting suffered by quantum computers.Wherein,logical qubits ar

220、e groupings ofqubits connected via quantum entanglement.Instead of relying on redundant copies of information as anerror-correcting protocol,logical qubitbased machines rely on the built-in redundancy of entanglement.In theirpaper published in the journal Nature,the group shows that while executing

221、calculations,their quantum computerhad fewer errors than other larger machines based on physical qubits.The above barrier-breaking achievements mark a milestone toward fault-tolerant quantum computers capable ofsolving large-scale problems.This is expected to have a breakthrough impact in fields suc

222、h as medicine,materialsscience,pharmaceuticals,energy,and so on,leading the fields to a stage that is not yet feasible using currentcomputer technology.Meanwhile,the disruptive potential of quantum computing also raises critical issues aboutdata security and threatens existing encryption standards,t

223、hus moving up the timetable for anti-quantumcryptography research.From the year of 2024,the field of quantum computing is expected to transition from physical qubits toerror-correcting logic quantum bits,and anti-quantum cryptography research is expected to speed up,marking asignificant year for qua

224、ntum computing technology.25https:/atom- thanks to the following contributors for their wonderful work on this whitepaper:Editors:Xin GUO(Lenovo),Chih-Lin I(China Mobile)Contributors:National Information Optoelectronics Innovation Center(NOEIC)&China Information and CommunicationTechnologies Group C

225、orporation(CICT)Yanxin HAN,Xin HUA,Xi XIAOChina Telecom Quantum GroupJun LUOChina Unicom Research InstituteChunxu ZHAO,Wenxiu QU,GuangquanWANG,Haijun WANGChinaAcademy of Information and CommunicationsTechnology(CAICT)Qi YUAN,Xiaoli LIUChina Mobile Research InstituteChengkang PAN,Shuai HOU,Xinying LI

226、,Xian LUTURINGQ Co.,LtdXianmin JIN，Lin YANG，Daquan YANG，Feng YIN，Junjie HELenovoXin GUOFuTUREFORUMiscommittedtocuttingedgetechnologiesstudyandapplications.Controversies on some technical road-maps and methodologies may arise from time to time.FuTURE FORUM encourages open discussion and exchange of i

227、deas at all levels.The White Paperreleased by FuTURE FORUM represents the opinions which were agreed upon by all participatingorganizations and were supported by the majority of FuTURE FORUM members.The opinionscontained in the White Paper does not necessarily represent a unanimous agreement of all FuTUREFORUM members.FuTURE FORUM welcomes all experts and scholars active participation in follow-on workinggroup meetings and workshops.we also highly appreciate your valuable contribution to the FuTUREWhite Paper series.