1、1/84AbstractMassive access technology refers to the technologies developed to address the challengeof accommodating extensive connectivity.6G sets higher demands for connection density,specifically targeting scenarios with massive connections.The envisioned connection densityfor 6G is tens of millio

2、ns of terminals per square kilometer.6G can bring about a fully digitalsociety by integrating Internet of Everything(IoE),allowing individuals to engage in natural,inspiring,and pervasive interactions with their surroundings.Each individual is anticipated toactivate tens to hundreds of terminals con

3、nected to the 6G network,with the interaction cyclefor machine-type terminals expected to decrease from 1 day or 2 hours in 5G to mere seconds.In the context of 6G,there are increased demands for low latency,particularly for the sporadictransmission of small packets by machine-type terminals.This ne

4、cessitates a further reductionin the delay from initial access to establishing a Radio Resource Control(RRC)connection,aiming to achieve simplified access procedures.6G advances the Massive Machine Type Communications(mMTC)scenario byintroducing the Massive Communication scenario,which involves a po

5、tentially enormousnumber of access terminals,reaching up to tens of millions per square kilometer.Additionally,in the context of 6G,the Immersive Communication and Hyper Reliable and Low-LatencyCommunication(HRLLC)scenarios stemming from Enhanced Mobile Broadband(eMBB)and Ultra-Reliable Low-Latency

6、Communications(URLLC)scenarios bring forth increaseddemands for indicators such as connection density and latency.To satisfy these demands,extensive research into massive access technology is essential.This white paper aims to lay out the framework of massive access technology,investigate its applic

7、ation scenarios and demands,pinpoint and illustrate the technicalproblems that need to be fixed,offer a path forward to solve these problems,and suggestspecific technical fixes.Firstly,it examines the many applications of massive accesstechnology and outlines the requirements for indicators like con

8、nection density,latency,datatransmission frequency,and data packet size in each application.It then examines therequirements for massive access technology,including enhancing connection density,2/84decreasing communication latency,and optimizing communication processes,in conjunctionwith the particu

9、lar requirements of use cases.It then turns to study the key technologies forenabling massive access,explores the industrys prevailing technical approach of UnsourcedMultiple Access(UMA),and suggests Uncoordinated Random Access and Transmission(URAT)technology,efficient connectionless transmission t

10、echnology,sparse InterleaveDivision Multiple Access(IDMA)for UMA,and multi-user encoding and decoding schemesbased on On-Off Division Multiple Access(ODMA).Based on this,it refines the designschemes for transmitters and receivers and presents resource hopping,pattern divisionrandom access(PDRA),and

11、Virtual User Splitting as the foundation for multiple accessschemes;iterative receivers based on Sparsification Transformation;capacity-optimized andlow-complexity iterative receivers;and multi-user encoding schemes.Finally,it integratesand categorizes the proposed key technologies,refining them to

12、create a technical roadmapfor enabling massive access.3/84Contents1.Introduction(CICT Mobile).42.Application Scenarios of Massive Access Technology(CICT Mobile,China Unicom).73.Important Use Cases.103.1 ToC Digital Twin World(CICT Mobile).103.2 Critical Connection-Intensive IoV(ZTE,CICT Mobile).123.

13、3 Ultra-Low-Power IoT(vivo).144.Basic Requirements(CICT Mobile).175.Communication Process of Massive Connections(CICT Mobile,vivo).196.Key Technologies for Enabling MassiveAccess.246.1 UMA Schemes.256.1.1 URAT(CICT Mobile).256.1.2 Efficient Connectionless Transmission(ZTE).276.1.3 Multi-User Encodin

14、g and Decoding Schemes Based on ODMA(XDU).426.1.4 Sparse IDMA for UMA(China Mobile).456.2 Transmitter Design Schemes.516.2.1 MultipleAccess Scheme Based on Virtual User Splitting(DOCOMO).516.2.2 MultipleAccess Scheme Based on Resource Hopping(BJTU).556.2.3 PDRA(USTB).596.3 Receiver Design Schemes.62

15、6.3.1 Iterative Receiver Based on Sparsification Transformation(USTB).626.3.2 Capacity-Optimized and Low-Complexity Iterative Receiver and Multi-User EncodingScheme(XDU,ZJU).667.Technical Roadmap(CICT Mobile).728.Conclusion and Outlook(CICT Mobile).77References.79Contributing Unit.82Abbreviations an

16、d Acronyms.834/841.Introduction(CICT Mobile)As emerging technologies like IoT and AI rapidly advance,the demand forcommunication technologies is also on the rise.The arrival of 5G technology has givenpeople access to faster,more stable,and more reliable communication services.However,asthe number of

17、 IoT devices continues to rise,accommodating the massive influx of devices hasemergedasapivotalchallengeintodayscommunicationlandscape.Traditionalcommunication technologies are no longer capable of meeting the access demands imposedby this vast volume of devices.Therefore,investigating and developin

18、g massive accesstechnology has emerged as a crucial focal point in the era of 6G.Presently,numerous standardization organizations are actively engaged in 6G researchand have made substantial advancements in novel multiple access technologies.For example,ITU has examined an application-oriented dynam

19、ic access scheme that,within an overarchingframework,employs diverse multiple access schemes tailored to specific applicationscenarios.This facilitates the potential convergence of different multiple access schemes andsupports the standardized advancement of new multiple access technologies.The IMT

20、2030(6G)Promotion Group has reached a consensus on technical concepts,research progress,challenges,and research directions.The scope of research encompasses fundamentaltransmission methods at the transmitting end and the signal processing process at thereceiving end,providing a guiding framework for

21、 future research on new multiple accesstechnologies.This facilitates the resolution of outstanding issues and the attainment ofsignificant breakthroughs.In November 2022,a crucial technical research report was issued,introducing the potential of massive access technology without user identification

22、based onNon-Orthogonal Multiple Access(NOMA)to greatly enhance terminal connection density.This advancement offers robust support for 6G scenarios such as low power consumption andultra-large-scale connectivity and serves as the groundwork for thorough exploration intophysical layer air interface te

23、chnology.FuTURE Forum issued the White Paper on 6G Visionand Technology Trends 1 in 2020 and the White Paper on the Evolved Random Access andMultiple Access Transmission Technologies 2 in 2022.These documents cover grant-freetechnology,which does not necessitate full network coordination,and also in

24、troduce a designapproach for integrating random access and multiple access transmission technologies.Inlarge-scale connection scenarios,6G Flagship emphasizes the significance of NOMA as a keytechnology.It highlights the need to address challenges associated with user activity detectionand data deco

25、ding,while also proposing research on NOMA technology that operates withoutthe necessity for full coordination.Meanwhile,there have been significant advancements in the academic field of NOMAtechnology,leading to the introduction of massive access technology 3.All terminals encodeinformation sequenc

26、es using the same codebook,and the coded bits are then transmitted overshared resources using Slotted ALOHA after modulation.Its primary characteristic is the lackof necessity to allocate terminal IDs for transmission,leading to its designation as unsourcedor uncoordinated multiple access,which has

27、emerged as a prominent research focus inacademic circles.Recently,academic research has extensively explored novel multiple accesstechnologies for 6G.There is a widespread consensus that mMTC represents a pivotal5/84scenario for 6G,requiring enhanced performance indicators beyond those of 5G.Consequ

28、ently,the introduction of innovative multiple access technologies becomes essentialto tackle the ensuing challenges.In tackling significant challenges such as pilot/preambledesign,channel estimation,and multi-user activity detection,many studies have incorporatedtechnologies like compressed sensing,

29、multi-user encoding,and deep learning to proposesolutions such as Unsourced RandomAccess(URA)and UMA.The core concept of URAliesin its ability to enable concurrent access and data transmission without requiring coordinatedscheduling or a separate access process,thus simplifying the access protocol a

30、nd reducingsignaling overhead.Widely recognized as the prevailing multiple access scheme in 6Gacademic research,it forms the basis for the majority of multiple access investigations.TheURA scheme currently proposed in academic circles can be primarily categorized into twotypes.One type involves the

31、use of concatenated inner and outer codes to transmit pilots anddata concurrently,thereby integrating the access and data transmission processes.The othertype emphasizes reducing the coupling between inner and outer codes.The concatenation ofinner and outer codes often necessitates additional redund

32、ant information,leading to adecrease in spectral efficiency.Therefore,this scheme primarily relies on specific correlatedinformation,such as spatial and channel correlations,to eliminate connection coding.This white paper aims to explore the research and development of massive accesstechnology in th

33、e 6G era.It seeks to establish a comprehensive framework for this technology,encompassing application scenarios,technical requirements,key challenges,viable technicalroadmaps,and several pivotal technologies.Ultimately,it aims to provide strong support forthe advancement of 6G technology.Chapter 2 o

34、utlines the two primary application scenariosof massive access technology,tailored to varying traffic model and performance indicators.Chapter 3 details the important use cases of massive access technology in 6G,including theToC digital twin world,the critical connection-intensive IoV(Internet of Ve

35、hicles),andultra-low-power IoT(Internet of Things).Chapter 4 analyzes the requirements for massiveaccess technology in 6G based on scenario use cases,focusing on indicators like connectiondensity,latency,and terminal data transmission frequency,and aiming to reduce latency bysimplifying communicatio

36、n processes and enhancing connection density through optimizedtechnical solutions.Chapter 5 outlines the communication process for massive connections,encompassing simplified security and transmission mechanisms.Chapter 6 thoroughlyexamines the primary challenges and explores key technologies to ful

37、fill these demands.These encompass URAT,efficient connectionless transmission technology,multi-userencoding and decoding schemes based on ODMA,sparse IDMA for UMA,multiple accessschemes based on resource hopping and Virtual User Splitting,PDRA,iterative receiversbased on Sparsification Transformatio

38、n,capacity-optimized and low-complexity iterativereceivers,and multi-user encoding schemes.Chapter 7 offers a forward-looking perspectiveon technological evolution to shape the roadmap for future massive access technology.Thisencompasses non-orthogonal,uncoordinated,integrated random access and mult

39、iple accesstransmission,and considerations for receiver design.Chapter 8 concludes and provides afuture outlook.This white paper holds vital significance in driving the research and development ofmassive access technology in the era of 6G.Firstly,this white paper introduces the6/84fundamental framew

40、ork and universal requirements of massive access technology,offeringguidance and reference for future research.Next,this white paper provides a detailedoverview of the key technical solutions for massive access technology,offering insights andguidance for research in relevant fields.Finally,this whi

41、te paper extensively deliberates onthe demand challenges and their corresponding technical roadmaps,establishing a consensusand groundwork for future endeavors.We believe that,through collective endeavors,massiveaccess technology will progress and contribute significantly to the IoT and associated f

42、ields,establishing the groundwork for an advanced intelligent society and delivering increasedconvenience and innovation to peoples lives.7/842.Application Scenarios of Massive Access Technology(CICTMobile,China Unicom)6G is expected to demonstrate a notably substantial increase in connection densit

43、ycompared to 4G and 5G.While 4G boasts a connection density of 2,000 connections persquare kilometer and 5G reaches one million connections per square kilometer,the prevailingindustry expectation is that 6G will achieve a staggering ten million connections per squarekilometer or possibly even higher

44、.In massive connection scenarios,a diverse range ofterminal device types is anticipated.Alongside traditional IoT devices,the introduction ofnovel unsourced,low-power terminal devices is expected.In June 2023,the ITU finalized a proposal 4 outlining the framework and overall goalsfor IMTs 2030 and f

45、uture development,and introduced six primary application scenarios for6G.On the one hand,it evolved the eMBB,mMTC,and URLLC scenarios within 5G,presenting Immersive Communication,Massive Communication,and HRLLC scenarios.Onthe other hand,it expanded upon the existing 5G scenarios to propose Ubiquito

46、usConnectivity,Integrated Artificial Intelligence and Communication,and Integrated Sensingand Communication scenarios.Massive Communication scenarios encompass a wide range of applications,such assmartcities,transportation,logistics,healthcare,energy,environmentalmonitoring,agriculture,and many othe

47、r fields.These applications frequently require support for a varietyof battery-free or long-life-battery IoT devices.High connection density is required for suchscenarios,and the document proposes a density requirement ranging from 106to 108devicesper square kilometer.Additionally,tailored to specif

48、ic use cases,there is a requirement toaccommodate varying data rates,power consumption,mobility,and coverage range,as wellas considerations for security and reliability.In existing technology,terminals must establish network access before initiating datatransmission.The number of terminals that can

49、be supported is limited by data transmissionresources and coordinated signaling resources of the network.For Massive Communication in6G,connection density is required at a much higher level than in 5G,possibly 100 timeshigher.Compared to 5G,6G scenarios will support a wider range of typical use case

50、s andterminal device types,placing additional demands on the traffic model.For example,therewill be an increase in the terminal data transmission frequency.One message per device everytwo hours is the recommended terminal data transmission frequency under the 5G mMTCtraffic model 5.According to the

51、IMT-2030(6G)Promotion Group,terminal datatransmission frequency in ultra-large-scale connection scenarios might be as low as once aday or as high as once every few milliseconds 6.There is a greater need for concurrent useraccess within a particular period of time due to the increase in terminal quan

52、tity and datatransmission frequency.It is expected to support simultaneous access of hundreds of terminalsper millisecond.The challenge that needs to be overcome is supporting communication formassive terminals while staying within the limits of network signaling and data transmissionresources.There

53、fore,it is essential to explore massive access technology to accommodate a8/84greater number of terminal connections.Massive Communication scenarios put forward higher requirements for connectiondensity,while the requirement for low latency is usually less strict and can often toleraterelatively hig

54、h latency,enabling end-to-end communication to be achieved within mereseconds.In practical applications,it is important to consider not only the number of supportedterminals,but also the data packet size,end-to-end communication latency,and the reliabilityof data transmission.In some cases,the requi

55、red connection density is less than that ofMassive Communication scenarios,which call for 108devices per square kilometer.It issufficient to have about 106devices per square kilometer.However,applications such as IoVand smart factories require millisecond-level end-to-end communication latency.Furth

56、ermore,in some application scenarios such as digital twins,larger data packet sizes are required.Examples of these include the transfer of packets of hundreds to thousands of bytes.On thebasis of meeting higher latency requirements and larger transmission packets,furthersupporting the access of mass

57、ive terminals faces significant challenges.Therefore,there is animperative need to conduct research on massive access technology capable of meeting highlatency requirements and facilitating larger data packet transmissions.According to different traffic model and indicator requirements,the applicati

58、on scenariosof massive access technology can be divided into two categories:one type is MassiveCommunication,which requires support for a large number of terminals(e.g.,108devices/km2),is not sensitive to latency,usually transmits bursty small data packets,and hashigh demands for connection density

59、but low requirements for data transmission frequency,latency,data packet size,and reliability,as indicated by the orange curve in Figure 2-1.Theother type is the scenario that supports a large number of terminals(e.g.,106devices/km2)while placing higher demands on the size of transmitted data packet

60、s,end-to-endcommunication latency,and data transmission reliability.This scenario integrates thecharacteristics of Massive Communication,Immersive Communication,and HRLLCscenarios,representing a combination of these three scenarios,with its key indicatorcapabilities shown by the blue curve in Figure

61、 2-1.It has lower requirements for connectiondensity compared to the first type of application scenario,but higher requirements for datatransmission frequency,latency,data packet size,and reliability.Figure 2-1Application Scenarios of MassiveAccess TechnologyMassive access technology enables the con

62、nection of massive terminals,meeting the9/84demands of the mentioned scenarios.On the one hand,it can satisfy the 6G requirements forlatency,connection density,and other indicators.On the other hand,in use cases like digitaltwins and ultra-low-power IoT that necessitate the deployment of massive dev

63、ices,massiveaccess technology can be employed as an underlying technology to support terminal accessand data transmission.10/843.Important Use Cases3.1 ToC Digital Twin World(CICT Mobile)Byintegratingsensing,computing,modeling,simulation,andcommunicationtechnologies,digital twin enables real-time in

64、teraction and seamless connection between thephysical and digital worlds.Its extensive applications span across intelligent manufacturing,smart cities,smart agriculture,and healthcare 7.Digital twin presents many challenges forthe architecture and capabilities of 6G networks,demanding substantial de

65、vice connections,high throughput,low-latency transmission,and more.This is essential for the precisereal-time capture of subtle changes in the physical world and the transmission of interactioninformation.For example,the following technical specifications must be met:connectiondensity of 107108devic

66、es per square kilometer,air interface latency of less than 1 ms,transmission rates ranging from Mbps to Gbps,and error probability of less than 10-56.Digital twin is widely used on both the consumer side(C-end)and industrial side(B-end),stimulating creative activities and communication in the virtua

67、l world,while enhancingindividuals understanding of the laws governing the physical world.The ToC digital twinworld expands upon the physical and digital worlds by introducing a human-centricdimension encompassing sensory perception,physicality,intellect,and values.Thisintegration facilitates intera

68、ctive information exchange among these three worlds,with afocus on delivering personalized,real-time,and immersive experiences for individuals.It ischaracterized by all-inclusive connectivity,real-time quality,and accuracy.Figure 3-1 ToC Digital Twin WorldUsing Immersive Tour of the Forbidden City a

69、s an example,the following illustratesthe typical features of the ToC digital twin world:Personalized experience:By constructing a ToC digital twin world of the ForbiddenCity,the virtual scenery may be adjusted to visitors own tastes,improving theviewing experience.To provide visitors with personali

70、zed experiences,it needs a11/84high connection density 106devices/km2.Real time experience:The virtual scenery must seamlessly integrate with thearchitecture and cultural relics,providing real-time presentation to visitors for aseamless fusion of the virtual and the physical.This requires stringent

71、latency toachieve millisecond-level end-to-end communication.Immersive experience:Like the mixed reality(MR)effect,this technique restoresand displays the palaces grandeur,color aesthetics,historical relevance,and otherelements by superimposing genuine buildings and cultural relics onto virtual scen

72、ery.This provides visitors with an incredibly immersive experience.To produce acompletely immersive experience,hundreds to thousands of bytes of data packetsmust be transferred.Compared to 5G,6G has significant differences and higher requirements in terms ofimmersive experience.6G will provide highe

73、r performance in transmission speed,networklatency,connection density,and reliability.Transmission speed:While 5G aims for a transmission speed of 20 Gbps,6G isprojected to reach 1 Tbps,which will make the immersive experience smoother andmore realistic.Network latency:6G is expected to achieve ultr

74、a-low latency as low as 1ms.Compared to the latency of 1-10ms in 5G,it will greatly improve the performanceof real-time interaction in immersive experiences and reduce the impact of latencyon user experience.Connection density:The connection density supported by 5G is 1 millionconnections per square

75、 kilometer,and 6G is expected to support 10 millionconnections per square kilometer.This means that 6G will be able to provide higherquality immersive experiences in high connection density scenarios.Reliability:To meet the requirements of immersive experiences,6G is projected tomake significant adv

76、ancements in reliability.For example,it aims to provide99.9999%service reliability for critical tasks and real-time interactive applications.Table 3-1 illustrates the estimated traffic model for this use case,including the terminaldevice type,terminal device quantity,and status reporting frequency.T

77、able 3-1 Traffic model for Immersive Tour of the Forbidden CityTerminal DeviceTypeTerminalDeviceQuantityStatusReportingFrequencyTerminalQuantityper SecondEstimation BasisTourists smartdevices104105About 1/min103104Number of visitors to theForbidden CityBuilding protectionsensor104105About 1/min10310

78、4Use cases in TR 22.84012/84Cultural relicdisplay sensor104105About 1/min103104Number of cultural relics ondisplay in the ForbiddenCityReal environmentsensing andimaging devices103104About 1/s103104Area of the Forbidden CityCultural relicstorage sensor105106About 1/hour103104Use cases in TR 22.011To

79、tal(number of terminals accessed andtransmitted per second)104105Approximate estimationresultsThe ToC digital twin world prioritizes the human experience.It aims to deliverpersonalized,real-time,and immersive experiences tailored to user needs by supportingmassive device communication.Compared to th

80、e Massive Communication scenario,higherrequirements are placed on data transmission latency and packet size.This entails supportinga connection density of approximately 106devices/km2,facilitating communication fromhundreds of terminals per millisecond,and ensuring millisecond-level end-to-end laten

81、cy.Additionally,across various use cases,the packet sizes vary from hundreds to thousands ofbytes.The existing method of establishing a connection before data transfer leads tosubstantial transmission latency,impacting user experience within the ToC digital twin world.Therefore,it is necessary to ex

82、plore massive access technology,simplify the signalinginteraction from initial access to multiple access transmission,and reduce latency.3.2 Critical Connection-Intensive IoV(ZTE,CICT Mobile)High-density IoV information transmission exhibits both massiveness and burstiness,thus necessitating low lat

83、ency and high reliability.Meeting the requirements for low latencyand high reliability in scenarios of massive and bursty information transmission poses aconsiderable challenge.Moreover,the rapid movement of vehicle nodes results in swiftchanges in the network topology of the IoV,making it even more

84、 challenging tosimultaneously meet the requirements for massive,bursty,low-latency,and high-reliabilitycommunication.The primary reason for the decline in eV2X performance during large-scale vehicleinformation exchange in high-density vehicle areas is the conflict arising from resourcecompetition.Fo

85、r example,TR 22.886 outlines a specific scenario featuring high-density U.S.freeways,assuming 10 lanes on each.At intersections where three freeways meet,the vehicledensity ranges from 3100 to 4300 vehicles per square kilometer 8.This challenge is beingencountered by both LTE-V and IEEE802.11p/IEEEl

86、609.In particular,as shown in Figure 3-2 in scenarios with frequent information exchangeamong numerous vehicles,the situation may arise where multiple vehicles broadcast theirmessages simultaneously.According to TR 36.885,the defined transmission packet size forV2V is 190/300/800 bytes 9.Without col

87、laboration between vehicles,each vehicle13/84autonomously determines itssignal generationand transmission resourceselection.Consequently,instances arise where different vehicles opt for the same transmission resourceand utilize identical signal generation methods,resulting in transmission collisions

88、.Thesecollisions present substantial challenges for other vehicles attempting to receive thesebroadcast messages.Moreover,traditionalV2VinformationtransmissionschemeslikeLTE-VandIEEE802.11p/l609 are fundamentally rooted in a half-duplex transmission mechanism.Thismeans that when a vehicle transmits

89、data on a specific channel,it is unable to simultaneouslyreceive data on the same channel.Consequently,this leads to the issue of missed reception:ifmultiple vehicles broadcast their data packets on the same channel,they will be unable toreceive each others transmissions,as shown in Figure 3-2.This

90、situation significantlycompromises high reliability.While enhancements in the sensing mechanism and increasedtime-frequency resources can decrease the missed reception rate,ensuring zero missedreceptions remains a challenge.This undoubtedly leads to a significant decrease in thereliability of V2V tr

91、ansmission.Figure 3-2 Scenario of Frequent Information Exchange among Numerous Vehicles14/84The critical connection-intensive IoV scenario encompasses specific operations such asformation driving,advanced driving,expanded sensor services,remote driving,and vehiclequality service support.Table 3-2 pr

92、ovides a detailed breakdown of the specific use casesand their typical requirements for these operations 8.Table 3-2 Requirements for Typical IoV Use CasesScenarioClassificationDataPayloadDataArrivalRateDevice DensityLatencyRequirementsAutonomousdriving6,000 bytes10/sHigh connectiondensity10-100 msA

93、utomaticformation6,000 bytes50/sHigh connectiondensity20 msDrivingcontrol1,600 bytes-9,840vehicles/km2100 ms3.3 Ultra-Low-Power IoT(vivo)Within the overarching vision of Intelligent Connection of Everything and Digital Twin10 in the 6G era,IoT is poised for broad development and strong demands for c

94、onnectivity.As an illustration,market analysis firm IoT Analytics has projected 11 that the demands forIoT terminals will exceed 27 billion by 2025.Although the existing standands like NarrowBand Internet of Things(NB-IoT)and RedCap can fulfill certain requirements of IoT,theyconfront some challenge

95、s,particularly in terms of high deployment costs and reliance onbatteries,which hinders long-term,maintenance-free operation and limits the applicationscenarios.Therefore,it is crucial to develop ultra-low-power IoT technology to reduceterminal costs,enable passive deployment and minimize maintenanc

96、e,while still benefitingfrom the strengths of cellular networks regarding wide-area coverage,manageability,andtrusted security.Ultra-low-power IoT terminals can be categorized as follows:Passive terminals:Without any active components,these terminals cannotindependently generate carrier signals.They

97、 transmit signals by adjusting theantennas impedance to change the amplitude and/or phase of the excitation signalfrom the network,i.e.,backscatter communication.Passive terminals operate atmicrowatt-level power consumption.Semi-passive terminals:Similar to passive terminals,semi-passive terminals c

98、annotindependently emit carrier signals either and use backscatter communication forsignal transmission.However,these terminals have energy storage units(ESUs)likecapacitors,albeit with less capacity than batteries,and thus can amplifybackscattered signals using stored energy.Their operating power c

99、onsumptionranges from tens to hundreds of microwatts.Active low-power terminals:With larger ESUs,these terminals can harvest andstore environmental energy,such as radio frequency,solar,light,and thermal energy.They have carrier generation capabilities similar to traditional IoT terminals but are15/8

100、4much less complex than NB-IoT devices.Active low-power terminals operate atmilliwatt-level power consumption.In terms of applications,ultra-low-power IoT can be found extensive uses in industrialmanufacturing,smart transportation,warehouse logistics,and smart homes,performing taskssuch as control i

101、nformation dissemination,data collection,and positioning in various fields.Industrial manufacturing:Industry 4.0 aims to establish a data-driven and intelligentcollaborative platform for automated planning,production,and maintenance.Thisentails integrating ultra-low-power IoT terminals to manage mat

102、erial storage,transportation,and consumption on production lines,as well as deploying numeroussensing-enabled ultra-low-power IoT terminals to gather real-time insights about theproduction environment,enabling prompt adjustments across manufacturing andprocessing stages.Furthermore,it is essential t

103、o integrate ultra-low-power IoTterminals into diverse production and processing devices to promptly retrievecontrol commands for various manufacturing operations.Smart transportation:Essential for building smart cities,it involves deployingultra-low-power IoT terminals with traffic and pressure sens

104、ors on roads to assist thecloud-based city infrastructure in gathering dynamic data about roads,vehicles,andpedestrians.This enables optimal road planning,congestion warning and relief,andrisk mitigation,enhancing overall traffic efficiency.Warehouse logistics/Asset management:Both warehouse logisti

105、cs and assetmanagement applications necessitate precise identification and tracking of goods orassets,with particular consideration given to the dynamic nature of goodscirculation in warehouse logistics.Low-value goods can be managed usingultra-low-power IoT terminals installed on pallets or contain

106、ers,whereas high-valuegoods may require the installation of an ultra-low-power IoT terminal for each item.The low-cost advantage of ultra-low-power IoT terminals is expected to lower thedeployment barrier,allowing for also installing the terminals individually on thelower-value goods,further increas

107、ing the management granularity.To this end,dueto goods stacking,the terminal deployment density could be remarkably high.Smart home:The essence of smart homes is to provide a personalized,healthy,andconvenient lifestyle,ultimately enhancing quality of life.This requires deployingultra-low-power IoT

108、terminals to connect numerous environmental sensors andhousehold appliances for automated control based on environmental data,such astemperature and humidity regulation,gas leak warnings,intrusion detection,andautomatic alarms.Moreover,localizing personal items is another significantapplication of s

109、mart homes.Many valuable but small items are often carelessly leftaround,making them challenging to locate promptly,such as keys and ID cards.Installing ultra-low-power IoT terminals on these items enables real-time tracking oftheir locations,facilitating quick retrieval of lost items.The specific s

110、ervices involved in the aforementioned important application scenariosinclude asset or goods inventory checks,sensing data services,positioning services,and16/84control services.Table 3-3 details the typical requirements for these types of connections12.Table 3-3 Requirements for Typical Use Cases o

111、f Ultra-Low-Power IoTScenarioClassificationDeploymentScenarioDataPayloadDataArrivalRateDeviceDensityLatencyRequirementsAssets orgoods inventorycheckIndoor oroutdoor 256bits 1/s 1.5/m2 1sSensing dataIndoor 800bits 20/hour 1.5/m2 30soutdoor 0.01/m2PositioningIndoor 1000bits 1/hour 0.25/m2 1soutdoor 1/

112、hour 0.01/m2ControlIndoor 800bits0.53/hourN/A 1.5/m2 10soutdoor 200bits 2/m2Figure 3-3 IoT Development Trends 13In general,the ultra-low-power IoT technology offers lower power consumption,reducedcosts,and slower data rates compared to existing cellular IoT technologies like NB-IoT andenhanced Machi

113、ne-Type Communication(eMTC).These characteristics significantly expandthe potential deployment of application scenarios,driving networks to support increasedconnection density and quantity.17/844.Basic Requirements(CICT Mobile)With the rapid development of mobile Internet,Internet of Things,big data

114、 and otherrelated technologies,the demand for data transmission rate,delay and connection density isbecoming higher and higher.In the existing access technologies,terminals need to completeinitial access,and then carry out data transmission.The whole process involves multiplesignaling interactions,w

115、hich requires a lot of time and resources.Chapter 3 outlines the important use cases of massive access technology,along with theirrespective communication requirements,as listed in Table 4-1.Table 4-1 Communication Requirements for Important Use Cases of MassiveAccessTechnologyScenario Use CaseConne

116、ctionDensityDataTransmissionFrequencyAverage DataPacket SizeAverageCommunication LatencyToC Digital Twin WorldAbout106/km21/s-1/hHundreds tothousands of bytesMillisecondlevelCriticalConnection-IntensiveIoV105-106/km210/s-1/minHundreds tothousands of bytesMillisecondlevelUltra-Low-Power IoT106-107/km

117、21/s-1/hTens of bytesSecond levelIn application scenarios demanding extensive connection like Massive Communication,existingaccesstechnologiescancausesubstantiallatencyduetoconstraintsontime-frequency and signaling resources.In application scenarios requiring real-time,low-latency performance,such a

118、s the ToC digital twin world,the connection density ofaround 106/km2is required.Moreover,terminals must be capable of transmitting atfrequencies ranging from 1/s to 1/h and processing relatively large data packets,usuallyranging from hundreds to thousands of bytes,while maintaining millisecond-level

119、 latency.Meanwhile,certainapplicationscenariosdemandhigherconnectiondensity.Inultra-low-power communication scenarios,data packets are typically small,around tens ofbytes,with data transmission frequencies ranging from 1/s to 1/h.These scenarios exhibit lowsensitivity to latency,typically at the sca

120、le of seconds,while necessitating support for aconnection density of 106to 107per square kilometer.Based on the analysis of the connection density and data transmission frequency fromthe use cases listed in Table 4-1,assuming the connection density is 107/km2and the terminaldata transmission frequen

121、cy is 1 message/60 seconds/device,it is estimated that 167 datapackets need to be received within one millisecond.Increasing the connection density andterminal data transmission frequency will further raise the number of data packets that need tobe supported per millisecond.With the given assumption

122、s,the 6G connection density is tentimes that of 5G and the 6G terminal data transmission frequency is 1/120 of 5G,theestimated connection capacity of 6G wouldbe 1,200 times of 5G.5G technology is difficult to18/84support 6G massive connection scenarios.Therefore,in order to support hundreds oftermin

123、als accessing simultaneously every millisecond,new massive access technologies needto be designed in 6G,including simplifying communication processes,optimizing transmitterand receiver solutions,etc.19/845.Communication Process of Massive Connections(CICT Mobile,vivo)UE must go through several basic

124、 procedures in the current communication systemsbefore transmitting data to the network.These procedures include accessing the gNB,networkregistration,authentication,networksecurityconfiguration,UEcapabilityreporting,establishing an RRC connection with the gNB,establishing a PDU session connection w

125、iththe UPF,making scheduling requests,accepting uplink transmission schedules,data transfer,and receiving confirmation messages 14.At least 26 air interface transmissions will occurbetween the UE and gNB,with an even greater number expected during retransmissions.Inresponse to the massive connection

126、 demands expected in 6G,driven by a surge in terminalquantities,existing multiple access technologies are inadequate.There is a compelling need tooptimize access technologies,thus simplifying multiple access procedures,reducing latency,improving system efficiency,and accommodating a larger number of

127、 terminals.In massive connection scenarios,the communication process can leverage the URATscheme.Thisinvolvescomprehensivedesignconsiderationsencompassingnetworkarchitecture,protocol stack,data flow,interfaces,security mechanisms,and multiple accessprocedures.Regarding network architecture,the URAT

128、scheme incorporates an application proxy onthe gNB,delivering Non-Access Stratum(NAS)functionality to the gNB-served terminalsand facilitating uplink data forwarding.With the application proxy in place,terminals arerelieved from direct interaction with the core network for signaling and data exchang

129、e.Similarly,the new network architecture necessitates corresponding adjustments to protocolstacks,data flows,and interfaces.Regarding security mechanisms,the application proxyprolongs the life of integrity protection and data encryption keys utilized by terminals.Figure5-1 shows the communication pr

130、ocess of massive connections.In current technology,oncethe terminal has accessed the gNB,various procedures such as network registration,authentication,configuration of network security,UE capability reporting,RRC connectionestablishment with the gNB,and setting up PDU session connections with the U

131、PF can besimplified into signaling interaction steps(7)to(9)between UE and gNB+app proxy usinglow-complexity security mechanisms,as depicted in Figure 5-1.Similarly,procedures likemaking scheduling requests,accepting uplink transmission schedules,and handling datatransmission can be simplified into

132、signaling interaction steps(10)and(11)between UE andgNB+app proxy employing low-complexity transmission mechanisms,as indicated in Figure5-1.It is important to note that the processes represented by signaling interaction steps(10)and(11)in Figure 5-1 may recur multiple times throughout the lifespan

133、of integrityprotection and data encryption keys.20/84Figure 5-1 Uplink Signaling Interaction Process of Massive Connected TerminalsFigure 5-1 illustrates URAT Transmission,a massive access technology that integratesrandom access and data transmission processes.The related key technologies are detail

134、ed insection 6.1.1.Figure 5-2 shows the signal transmission process of the URAT air interface.Figure 5-2 Communication Process of URATAir Interface TransmissionDuring data transmission,the terminal can initiate access by sending preambles andsimultaneously transmitting data without waiting for the a

135、ccess process to conclude.In theevent of data transmission failure,a retransmission can be initiated.The detailed signaltransmission process is outlined below.Preparation Stage before Terminal Signal Transmission(t1):Before datatransmission,the terminal must initially generate the data for transmiss

136、ion,including preambles and communication data packets.Subsequently,it undergoessignal transmission pre-processing,involving encoding and resource mapping.Terminal Time Synchronization and Frame Alignment(t2):Upon completion of21/84data transmission preparation,the terminal initiates time synchroniz

137、ation and framealignment procedures.Terminal Air Interface Data Transmission(t3):Following time synchronizationand frame alignment,the terminal initiates data transmission over the air interface,transmitting signals comprising preambles and communication data packets.Network Device Signal Detection

138、and Decoding(t4):Upon receiving data packetsfrom terminals,the network device conducts the detection and decoding of thereceived data packets.Before moving on to the detection and decoding ofcommunication data packets,it is imperative to finish the detection of the preamble.Preparation Stage for the

139、 Network Device to Transmit Feedback Information(t5):Based on the detection and decoding results of data packets sent to theterminals,the network device is required to transmit feedback information.Beforesending feedback information,it is necessary to generate,encode,modulate,andmap resources for th

140、e feedback information.Network Device Time Synchronization and Frame Alignment(t6):When thenetwork device sends feedback,it first undergoes time synchronization and framealignment processing.NetworkDeviceFeedbackTransmission(t7):Afterundergoingtimesynchronization and frame alignment processing,the n

141、etwork device transmitsfeedback over the air interface.Terminal Device Feedback Detection and Decoding(t8):Following thecompletion of data packet transmission(t3),the terminal device listens for feedbacksent by the network device,and detects and decodes the feedback received.Upon receiving feedback

142、from the network device,the terminal uses this information todetermine whether the packet was sent correctly.It then decides whether to stop the datapacket transmission or initiate a retransmission.The URAT transmission scheme simplifiesthe signaling interaction process for random access and data tr

143、ansmission of the terminal.Additionally,when integrated with the signaling interaction process illustrated in Figure 5-1,it simplifies the communication process in massive connection scenarios.As mobile communication evolves,the network sees a growing diversity of terminaltypes,encompassing smart te

144、rminals,NB-IoT terminals,RedCap terminals,and mMTCterminals.In scenarios with massive connections,addressing terminal power consumptionissues is crucial.6G can support ultra-low-power terminals through passive communicationmethods.Compared to existing communication terminals,these ultra-low-power te

145、rminalsmay face more stringent complexity constraints,requiring their communication processes toadhere to essential and simplified design principles.Regarding the types of communicationservices,ultra-low-power terminals can support both downlink-triggered and uplink-triggeredcommunication services.D

146、ownlink-triggered services involve terminals initiating datatransmission only upon receiving paging,polling,or other signals from the network.Theseservices are best suited for passive and semi-passive ultra-low-power terminals as they rely on22/84the network to provide radio frequency(RF)carriers fo

147、r uplink data transmission.Uplink-triggered services involve terminals initiating data transmission independently,fittingfor active ultra-low-power terminals capable of generating RF carriers on their own.Thesimplified communication process depicted in Figure 5-1 for massive connection scenarios isa

148、lso applicable to ultra-low-power terminal communication.Moreover,considering thecharacteristics of ultra-low-power terminals,besides data transmission,these terminals mightrely on network-provided RF carriers for energy.Therefore,aside from the communicationprocess,it is crucial to design specific

149、energy supply and harvesting processes for theseterminals.In particular,the access process for ultra-low-power terminals includes thefollowing steps:(1)Access Triggering:For downlink-triggered services,the terminal initiates uplinkaccess upon receiving paging,polling,or similar signaling.For uplink-

150、triggered services,theterminal starts uplink access after meeting specific internal conditions,such as periodictriggers,reaching a data buffer threshold,or energy levels falling below or rising above acertain threshold.(2)Obtaining the RF Carrier for Data Transmission:In the case of passive orsemi-p

151、assive terminals,initiating uplink access and data transmission relies on the networksending the RF carrier.Consequently,the terminal must acquire information related to the RFcarrier within the downlink-triggered signaling or search for the frequency of the RF carrierwithin a predetermined frequenc

152、y grid.(3)Obtaining the RF Carrier for Energy Harvesting:When both the network and theterminal support RF energy transfer,if the terminal is not connected or has not completed theauthentication process,the network may transmit low-power RF carriers on specific defaultfrequencies.In this case,the ter

153、minal searches within a predetermined frequency grid toreceive the energy supply signal and proceed with subsequent steps.Once the terminal isauthenticated and connected to the network,it can send a dedicated energy supply request,triggering the network to respond by transmitting a dedicated energy

154、supply RF signal to theterminal.(4)Synchronization:Comprises downlink synchronization and uplink synchronization.Ultra-low-power terminals can achieve downlink synchronization by either receiving publicbroadcast signals from the network side or being triggered by downlink signaling,includingfrequenc

155、y and timing synchronization.Moreover,these terminals can also transmit uplinksignals following network-side directives,allowing the network to measure timing advance(TA)and frequency offset,and then receive the signal at the proper frequency and timing.It isimportant to note that carrying out uplin

156、k synchronization will induce additional terminaloverheads and should be avoided whenever feasible.(5)Registration and Authentication:The ultra-low-power terminals initiate registrationrequests to the network side,after which the network side performs authentication andregistration.The registration

157、request may entail details regarding the device type,synchronization maintenance capability,energy harvesting capacity,and battery status.During initial network access,terminals can complete registration and authentication requestsin the first uplink access message or send them along with data to mi

158、nimize terminal23/84overhead.(6)Resource Configuration and Grants:This can be conducted in two ways:ConfiguredGrant/Grant-free or Dynamic Grant.The first method eliminates the need for terminals toobtain grants before initiating uplink access,thus reducing overhead.In particular,thenetwork side conf

159、igures a resource pool for the terminals during registration andauthentication.Terminals then competitively transmit uplink information using the allocatedresource pool.Conversely,the latter method requires the network to carry terminal-specificresources through downlink signaling for network access

160、 and data transmission.Alternatively,in scenarios with high terminal density,a resource pool for competitive access may beconfigured.(7)Connection release:Once the terminal has completed its service and the networkdetermines that no further communication with the terminal is required,the network wil

161、ldeactivate or disable the terminal,delete the network context,and release the correspondingresources or resource pool.Alternatively,if the terminal meets certain conditions(such asservice completion or low battery level),it can send a connection release request to thenetwork.Subsequently,the networ

162、k will deactivate or disable the terminal,delete the networkcontext,and release the resources or resource pool.It is important to note thatultra-low-power terminals may operate in a connectionless mode.In this case,neither theterminal nor the network needs to maintain a network context,making the co

163、nnection releaseoptional.24/846.Key Technologies for Enabling Massive AccessIn the context of 5G,NOMA technology has been studied,suggesting its potentialapplication in uplink grant-gree transmission.By implementing NOMA and uplink grant-freetransmission,the terminal simplifies the transmission proc

164、ess,resulting in reduced signalingoverhead,decreased power consumption,decreased latency,and increased system capacity.Inthe 5G standardization phase,a total of 17 NOMA schemes were studied,encompassingdesigns based on spreading,scrambling,interleaving,coding,and modulation.However,dueto issues like

165、 non-convergence in discussing candidate NOMA schemes,5G ultimately didnot standardize NOMAtechnology.In 5G,terminals are required to initially connect to the network.Upon successfulconnection,they engage in data transmission under the control of network.It supports thetwo-step and four-step RACH te

166、chniques.Compared to the four-step RACH,the two-stepRACH delivers lower access latency.Following the completion of access,to reduce datatransmission latency,it enables uplink grant-free transmission.This entails the terminalinitially connecting to the network and maintaining in connected state.The n

167、etworkpre-configures resources to the terminal,enabling direct data transmission on these resourceswithout the need for dynamic network scheduling when the terminal has data to send.In the context of 6G,it is essential to accommodate massive access scenarios while alsomanagingafurtherincreaseinconne

168、ctiondensity.Inmassiveaccessscenarios,accommodating a high volume of terminal devices is essential.Each terminal may only needto transmit a small data packet,ranging from 20 Bytes to 100 Bytes,and the transmissionfrequency and timing may lack strict regularity.If the current 5G approach of initial a

169、ccessfollowed by data transmission is employed,coordinating these terminals through the network,including initial access,connection maintenance,and resource scheduling/configuration,could lead to considerable overhead and latency.Furthermore,due to constrained datatransmission and network coordinati

170、on signaling resources,the number of terminals that canbe supported is limited.To accommodate massive access scenarios,6G must advance its access technology.Unliketheaccesstechnologyin5G,6Ghasthepotentialtotransformthenetwork-coordinated terminal access and data transmission process into uncoordinat

171、edprocess.Employing uncoordinated methods for terminal access and data transmission canlead to reduced average latency compared to the technology in 5G,where initial access isfollowed by network coordination before transmission.Simultaneously,it operates withoutdependency on finite data transmission

172、 and signaling coordination resources for schedulingand control.This saves network coordination signaling resources,eliminates limitations onthe number of terminals imposed by network coordination signaling resources,andsignificantly increases the number of supported terminal.To be more specific,in

173、the contextof 6G,using uncoordinated methods for terminal access and data transmission allows foradditional optimization of the communication process,transmitters,and receivers.This mayinvolve simplifying the procedures for terminal access and data transmission,employingconnectionless communication

174、tailored to massive access scenarios,as well as optimizing25/84pilot design and receiver design to suit such scenarios.These massive access technologies willbe introduced in detail in the following sections.6.1 UMASchemes6.1.1 URAT(CICT Mobile)In the 5G mMTC scenario,terminals typically transmit dat

175、a at a frequency ofapproximately one message per device every two hours or every day 15.In 6G,the growthin the number of terminals necessitates support for even higher data transmission frequencies.The increased connection density and frequency of terminal data transmission impose greaterdemands on

176、the access and transmission technologies in 6G.In the current 5G system,datatransmission follows the completion of access.However,persisting with this method inmassive connection scenarios of 6G could result in access network overload and signalingcongestion,leading to increased latency and signalin

177、g overhead.In massive connectionscenarios,limitations on resource pool size and excessively complex detection posechallenges in achieving the goal of massive access in 6G.Consequently,there is a pressingneed to develop novel massive access technologies to address this issue.The URAT scheme can incre

178、ase the number of terminal connections and mitigate theaccess network overload,signaling congestion,significant latency,and signaling overheadchallenges induced by the current multiple access technologies utilized by numerous terminaldevices.The URAT is designed with a dual focus on preambles and mu

179、lti-user encoding.Thisis accomplished by expanding the pool of candidate preambles to reduce collision risks andby applying control mechanisms such as scrambling,interleaving,and retransmission to thedata based on the preamble.Additionally,it leverages multi-user encoding to facilitate theconcurrent

180、 transmission of both the preamble and the data.The URAT scheme integrates therandom access and multiple access transmission processes to reduce signaling and resourceoverhead,thereby reducing access latency and accommodating a greater number of terminalconnections.In contrast to the traditional met

181、hod of initial access followed by data transmission inmultiple access scenarios,URAT technology integrates the initial access and multiple accesstransmission processes.This integration eliminates the need for network coordination andenables the simultaneous implementation of both random access and m

182、ultiple accesstransmission through a non-orthogonal method.-Uncoordination:Throughout the entire process from random access to multipleaccess transmission,the terminal operates without the need for UE-specific networkcoordination signaling,relying solely on the broadcast of basic cell configurationi

183、nformation.Based on the broadcasted basic cell configuration information,allterminals employ identical transmission methods,with potential differences amongthem primarily concerning the data.To enhance the reliability of air interfacetransmission,the network must provide terminal feedback confirmati

184、on messages forthe air interface transmission.This can be achieved using intrinsic terminal26/84identification to enable terminals to detect the confirmation message sent by thenetwork.-Non-orthogonal transmission:Terminals collectively utilize the time-frequencyresources indicated by the basic cell

185、 configuration information to achievenon-orthogonal multiple access transmission in the coding domain.In the codingdomain,non-orthogonal multiple access is achieved using finite block-length coding.Within the framework of compressed sensing and massive access technology,it ispossible to further incr

186、ease the distance between different terminal codewords.Inaddition to using various interleavers for coded bits,diverse scrambling codes canalso be employed for the coded bits.Moreover,unequal diversity can be applied torepeat code blocks,and the transmission can be distributed across the entiretime-

187、frequency domain.The interleaver,scrambler,diversity,and joint interleaverused by the terminal can be determined and indicated by metadata.-Integration of random access and multiple access transmission:can be achievedthrough a single transmission,eliminating the necessity for dedicated configuration

188、information transmission or terminal identity interaction,and solely requiringnetwork confirmation of terminal data.The entire process from random access tomultiple access transmission,is consistently carried out using URAT air interfacetechnology.Before transmission,the terminal sends a preamble si

189、gnal,which notonly contains metadata bits but also primarily serves to indicate to the network thepresence of concurrently transmitted multi-user coded signals.This action enhancesthe detection performance of multi-user coded signals,effectively serving as a formof random access.The signaling intera

190、ction that encompasses authentication,authorization,and security encryption between the terminal and the network can nowbe integrated into data transmission using URAT air interface technology.Thisintegration eliminates the necessity for distinct explicit processes for authentication,authorization,a

191、nd security encryption.The network confirms the terminals URATtransmission by feedback.Based on the feedback,the terminal determines whether aretransmission is required.The schematic diagram of URAT technology is shown in Figure 6-1.Figure 6-1 Schematic Diagram of URAT TechnologyURAT enables the mul

192、tiplexing of random access preamble signals and multiple accesstransmission data signals for combined transmission,thereby establishing a correlationbetween the preamble and data signals.27/84-When generating data signals for multiple access transmission:The terminalencodes the information bits of t

193、he data to be transmitted(including application layeridentity information)at an exceptionally low code rate and multiplexes them usingmulti-user encoding.These are subsequently carried by a channel similar to thePhysical Uplink Shared Channel(PUSCH).-When generating preamble signals for random acces

194、s:Establish a connectionbetween the preamble signal and the data signal.Specifically,start by obtaining themetadata bits required to create the preamble signal from the information bitsintendedfortransmission.Subsequently,generatethepreamblesignalfortransmission based on the metadata bits and establ

195、ished rules.For example,following the index transformation of the metadata bits,the terminal identifies therelevant preamble signal from the candidate resource pool.Apart from its role inrandom access,the preamble signal can also carry metadata bit information.-When transmitting data signals:The dat

196、a signal should be associated with thepreamble signal.Specifically,the metadata bits used to generate the preamble signalshould be utilized to control the data signal.For example,these metadata bits areemployed to regulate multi-user encoding,encompassing scrambling,interleaving,repeating transmissi

197、on patterns,resource mapping,and power adjustments,all aimedat enhancing differentiation among users.In the URAT scheme,the joint transmission of terminal identity information and terminaldata information simplifies dynamic network coordination,effectively increasing the numberof terminals.This make

198、s it well-suited for handling the access and transmission of massiveterminals.Moreover,by integrating the initial access and data transmission processes,thereceiving end can simultaneously acquire terminal identity information and terminal datainformation,thereby reducing transmission latency,especi

199、ally for burst data transmissioninvolvingsmallpackets.Furthermore,theenhancedunequaldiversitytransmissiontechnology facilitates differentiated transmission for various terminal sets,thus improving thesuccess rate of access and transmission,particularly benefiting high-priority terminal sets.The URAT

200、 scheme enables the resources sharing of massive terminals through a unifiedencodingprocess,meetingindividualterminalerrorperformancerequirements.Itstransmission method operates without the need for network coordination,conserving networksignaling resources and eliminating restrictions on terminal q

201、uantity imposed by suchresources.Thus supports the access and transmission of massive terminals.6.1.2 Efficient Connectionless Transmission(ZTE)Traditionaltransmissionmethodsbasedonschedulingandpre-configuredSemi-Persistent Scheduling(SPS)are not suitable for massive connection scenarios in 6G.Tomax

202、imize system relief and reduce terminal power consumption,it is optimal for terminals topredominantly operate in a deep sleep state,known as RRC idle.When data transmission isnecessary,there is no need to establish a connection in advance.Instead,connectionlessterminals autonomously initiate immedia

203、te data transfer.Upon completion of the transmission,28/84the device promptly enters a deep sleep state almost akin to shutdown,as shown in Figure 6-2.This connectionless transmission method allows for extremely simplified data transfer,optimizing system spectral efficiency and minimizing terminal p

204、ower consumption.It ishighly suitable as an enabling technology for massive connection scenarios in 6G.Figure 6-2 Extremely Simplified Connectionless Transmission for Massive ConnectionsHowever,in massive connection scenarios,the challenge of establishing minimalconnectionless transmission is substa

205、ntial.Base stations may encounter a significant volumeof autonomously transmitted data packets from multiple terminals sharing the sametime-frequency resources,posing a major challenge in effectively demodulating these packets.First,signals carrying different user information need to have sufficient

206、 distinctiveness,whichforms the physical basis for distinguishing such user information.Second,these distinctionsmust be fully utilized by the base station,even during connectionless transmission.Non-orthogonaltechnologyenhancessignaldiversitybypermittingpartialnon-orthogonality among different user

207、 signals.This equalizes interference between users,aiming to minimize instances of severe and indistinguishable interference,thereby bolsteringmulti-user performance robustness.Moreover,advanced non-orthogonal multi-user detectiontechnology comes at a certain level of complexity cost but ensures rel

208、iable performance in thepresence of substantial multi-user interference.Therefore,connectionless transmissionrequires advanced non-orthogonal transmission and reception technologies.However,traditional non-orthogonal transmission and reception technologies all rely on pilots to obtainthe differences

209、 in data signals from different users and then utilize these differences toseparate user data.In the scenario of connectionless transmission,when pilots suffer severecollisions,it becomes challenging for base stations to estimate the differences in data signalsusing heavily collided pilots,thereby h

210、indering multi-user detection.Therefore,for efficientconnectionless transmission,it is necessary to further consider advanced non-orthogonaltechnologies that can minimize or eliminate reliance on pilots.At the transmitting end,it is essential to employ advanced non-orthogonal transmissiontechnologie

211、s to mitigate the inseparable severe user interference present in the data signal.Additionally,attention should be directed towards transmission schemes that minimize pilotcollisions.At the receiving end,it is imperative to explore advanced detection technologies thatminimize or eliminate reliance o

212、n pilots.This approach aims to effectively extracttransmission information from multiple users based on the distinctive characteristics of thepower domain,code domain,and space domain,thus ensuring optimal performance inconnectionless transmissions.(I)Efficient Non-orthogonal Connectionless Transmis

213、sion29/84Connectionless transmission based on the near-far effect in the power domainIn connectionless transmission,the absence of centralized control,like base stations,prevents precise power regulation,invariably resulting in the near-far effect.Whiletraditionally considered detrimental to wireles

214、s communication systems,the near-far effectenables the separation of multi-user information in the power domain.For example,when thebase station receives transmission signals from two users,one strong and the other weak,itcan initially demodulate and decode the information from the strong user.Subse

215、quently,byreconstructing this information into the transmission signal,the base station effectivelyeliminates the strong users signal from the received composite signal.As a result,there is nolonger any need to contend with interference from the strong user when demodulating theweak users signal.See

216、 Figure 6-3.It is evident that when the base station accurately acquiresand manages the users channel,leveraging the near-far effect can effectively enhance thesystems capacity to support a higher number of connections.However,in scenarios with massive connections,relying solely on the one-dimension

217、alpower domain,specifically the near-far effect,remains inadequate for achieving efficienttransmission.For example,when multiple users have the same received power,it becomesdifficult to differentiate between these users within the power domain.This situation is notuncommon in scenarios with massive

218、 connections,so additional support for multi-usertransmission is required through code domain extension and space domain extension.Figure 6-3 SIC-based Multi-User Detection TechnologyConnectionless transmission based on code domain extensionIn connectionless transmission using code domain extension,

219、the transmitter randomlyselects extended codewords,while the receiver utilizes these extended codewords to mitigateuser interference and enhance the symbol SNR.The properties of the extended sequencedirectly impact the performance and receiver complexity of connectionless transmission,making them cr

220、ucial to code domain extension schemes.When employing very longpseudo-random sequences(PN sequences)like traditional Direct-Sequence Code DivisionMultiple Access(DS-CDMA)like the IS-95 standard,ensuring low correlation between30/84sequences becomes achievable.This configuration provides a soft capac

221、ity for the system,allowing the number of users(i.e.,sequence count)to exceed the sequence length,effectivelyplacing the system into an overloaded state.While long PN sequences can provide a softcapacity to accommodate specific overload rates,the overload rate tends to be high in systemswith massive

222、 connection demands.Additionally,in scenarios with high overload rates,utilization of long PN sequences results in a complex and lengthy Successive InterferenceCancellation(SIC)process.Moreover,aconsiderabletime-frequencyextensionoftransmit-side symbols also escalates the complexity of terminal tran

223、smission.Conversely,if ashorter sequence can achieve a high overload rate comparable to that of a longer sequencethrough effective design optimization,then in terms of transmission/reception complexity andprocessing latency,using such a short sequence becomes more appropriate.However,thedecrease in

224、user overload rate when using shortened traditional PN sequences is quite rapid.This is because traditional PN sequences are binary real sequences,and shortening themmakes it challenging to maintain low cross-correlation within the randomly generatedsequence set.Conversely,effective code domain exte

225、nsion schemes improve the performanceof competitive non-orthogonal access in the code domain by optimizing symbol-extendedsequences.In particular,the enhanced Multi-User Shared Access(eMUSA)technologyutilizes multivariate code sequences over a complex field as the symbol extended sequence16,enrichin

226、g the combination of extended sequences.Despite its brevity,it canaccommodate numerous low cross-correlation sequences.For example,the real andimaginary parts of each element in a typical eMUSA sequence constitute a simple binary set1,1 or a ternary set 1,0,1,as shown in Figure 6-4.Even with very sh

227、ort lengths,like8 or even 4,these sequences can yield a substantial number of sequences with low correlation.For example,with a sequence length of 4,the PN sequence can yield a maximum of 8 distinctsequences.In contrast,the eMUSA sequence can generate as many as 156 sequences if thecross-correlation

228、 energy(which indicates interfering energy)is below 0.63,as listed in Table6-1.In this manner,even if numerous unconnected users independently opt to extend thesequence concurrently,collisions can be effectively minimized,thereby randomizing userinterference.Leveraging the efficient SIC receiver,eMU

229、SA enables a multiple-fold extension of nodecapacity to transmit in the same time-frequency resources by autonomously selectingextended sequences,thereby achieving minimal connectionless transmission for massivenodes.31/84Figure 6-4(a)Constellation Diagram of Binary Complex Sequence Elements,(b)Cons

230、tellation Diagram of Ternary Complex Sequence ElementsTable 6-1 Comparison between PN Sequences and eMUSASequencesSequence Length(4)PN Real CodeeMUSACodeTotal number166561Cross-correlation energy 0.25820Cross-correlation energy 0.5860Cross-correlation energy 0.638156Connectionless transmission based

231、 on multi-antenna space domain expansioncapabilitiesIn scenarios with massive connections,if users extended code sequences collide andtheir received powers are equivalent,neither the code domain nor the power domain candistinguish between these two users.What should we do?Even in cases where neither

232、 thecode nor the power domain is effective,efficient connectionless transmission schemes muststill leverage the space domain extension capability of multiple antennas to separate users.When M receive antennas are deployed at the base station,the M-dimensional spatialchannel vectors for different use

233、rs are usually not completely identical,which providesmulti-user multiplexing capability.In practical terms,a multi-antenna base station cangenerate distinct receiving beams for capturing signals from multiple users.The base stationcan form a yellow beam to receive signals from user 1,as shown in Fi

234、gure 6-5.This impliesthat despite users having identical received power and extended code,the base station canpotentially segregate them by exploiting differences in their spatial channels.The capabilityfor spatial multi-user multiplexing typically enhances with the increase in the number ofreceive

235、antennas at the base station.This is attributed to the larger antenna aperture anddecreased correlation of spatial channels resulting from a greater number of antennas under afixed antenna spacing,usually half-wavelength.For example,when two users are located neareach other,and the base station has

236、a limited number of receive antennas,the spatialcorrelation of the air interface between these users may be substantial,posing challenges fordifferentiation.Conversely,with an increased number of receive antennas at the base station,the aperture expands,leading to reduced spatial channel correlation

237、 between the users,thereby facilitating easier differentiation.32/84Figure 6-5 Spatial Multi-User Multiplexing Capability(II)Pilot Design in Efficient Connectionless TransmissionWith knowledge of variations in individual user signals such as strength,extended codes,and spatial channel vectors,the ba

238、se station can utilize advanced non-orthogonal detectiontechnology to achieve the separation and demodulation of multi-user data.However,pilots inconnectionless transmission are also independently selected by users,and different users canchoose the same pilot,leading to pilot collisions.In scenarios

239、 with massive connections,theprobability of pilot collisions is extremely high.If a pilot collision occurs,the base stationfaces difficulty in accurately estimating the differences among users,even when their spatialchannels are not highly correlated or exhibit noticeable near-far effects.Consequent

240、ly,separating user data based on signal strength or the low correlation of spatial channelsbecomes challenging.Additionally,the extended codes constitute a finite set and correspondto pilots.In most cases,if there is a pilot collision,there will also be a collision of extendedcodes.It can be seen th

241、at in traditional non-orthogonal multiple access technologies whenpilot collisions occur,the ability to distinguish between multiple users in the space,power,and code domains is essentially compromised.To significantly reduce instances of pilot collisions,it is common practice to substantiallyincrea

242、se the quantity of pilots used.However,traditional detection technologies require pilotsnot only to estimate the radio channel but also to determine time and frequency offsets.Thisnecessitates a substantial distribution of known symbols for each pilot across the entiretime-frequency resource during

243、transmission.This will result in a notable rise in pilotoverhead,potentially leading to situations where the pilots utilize more time-frequencyresources than the data packets.Moreover,during multi-user detection,the base station needsto detect a large number of long pilots,leading to a significant i

244、ncrease in complexity.To minimize pilot overhead and complexity,it is essential to explore advanced detectiontechnologies that rely minimally or not at all on pilots.This will ensure optimal utilization ofspace,code,and power domain multi-user multiplexing capabilities to support a vast numberof con

245、nections,even in connectionless transmissions.Taking eMUSA connectionless data transmission technology as an example,this whitepaper presents detection technologies that minimize or eliminate the reliance on pilots:33/84data-onlytechnology1718192021andultra-low-collisionpilottechnology222324.These t

246、wo technologies share a common guiding principle:traditionalmulti-user detection methods typically begin with channel estimation using pilot symbols.This is followed by leveraging the estimated channel for user interference suppression,datasymbol equalization,and subsequent demodulation and decoding

247、,as shown in the upper halfof Figure 6-6.However,the high collision of connectionless pilots leads to inaccurate channelestimation,thereby restricting subsequent multi-user detection.The eMUSA detectiontechnology reverses the traditional detection sequence by initially employing space,code,andpower

248、domains for interference suppression.Subsequently,it estimates the channel using thedata symbols post-interference suppression,followed by equalization,demodulation,anddecoding.See the lower half of Figure 6-6.Even in heavily loaded high-connection scenarios,the SNR of the data symbols,following int

249、erference suppression in the space,code,andpower domains,can remain sufficiently high to ensure effective channel estimationperformance.Figure 6-6 Traditional Multi-User Detection Technology vs.eMUSAConnectionlessDetection TechnologyData-only technology without the need for traditional pilotsData-on

250、ly technology enables efficient blind detection by analyzing the statistical andgeometric properties of received data symbols from different users.The primary focus is toestimate the channel and time-frequency offset across the entire transmission resource byleveraging the geometric characteristics

251、of the modulation symbol constellation diagram.Low-order modulation symbols,such as Binary Phase Shift Keying(BPSK)andQuadrature Phase Shift Keying(QPSK),exhibit simple geometric shapes in their respectiveconstellation diagrams.Even when the received modulation symbols are distorted by thechannel,th

252、e constellation diagrams of these symbols undergo only rotation and scaling,thuspreserving their relatively simple geometric shapes.The BPSK constellation followingchannel rotation and scaling is shown in Figure 6-7.34/84Figure 6-7 Schematic Diagram of the Partition Matching(PM)MethodTherefore,the r

253、eceiver can utilize the geometric shape of the constellation diagram,asshown in Figure 6-7,to estimate the amount of rotation and scaling experienced by theconstellation diagram.The following outlines a specific method known as the PM method:First,divide the two-dimensional plane or signal plane int

254、o four partitions.For example,two common methods for partitioning the two-dimensional signal plane into four sections areas follows:The first method involves dividing the plane into four quadrants,each representing adistinct partition,with the x-axis and y-axis acting as partition lines,as shown inF

255、igure 6-8(a).The quadrant filled with diagonal lines corresponds to partition 1.The quadrant filled with dots corresponds to partition 2.The quadrant filled with vertical lines corresponds to partition 3.The quadrant filled with brick-like shapes corresponds to partition 4.The second method involves

256、 rotating the four partitions from the first method by 45degrees to create the required partitions,as shown in Figure 6-8(b).That is:Partition 1 is defined by filling the area between the 45 and 135 partition lineswith diagonal lines.Partition 2 is defined by filling the area between the 135 and 225

257、 partition lineswith dots.Partition 3 is defined by filling the area between the 225 and 315 partition lineswith vertical lines.Partition 4 is defined by filling the area between the 315 and 45 partition lineswith brick-like shapes.In addition to the two partition methods illustrated in Figure 6-8,i

258、t is also feasible to35/84divide the two-dimensional plane into four partitions.Nonetheless,employing the partitionmethods depicted in Figure 6-8 allows for straightforward determination of a constellationpoints placement within a partition through basic addition and subtraction,eliminating theneed

259、for complex multiplication operations and thus streamlining the process.Figure 6-8 Example of the PM MethodOnce the receiver divides the two-dimensional signal plane into four partitions,it adds upthe constellation points within each partition separately and then divides the total by thenumber of co

260、nstellation points within that partition.The resulting point represents the centerof the constellation points within that partition.Subsequently,an estimate of the rotation andscaling amount for the entire constellation diagram can be derived by using the center of allconstellation points within eac

261、h partition.In the presence of Additive White Gaussian Noise(AWGN),certain modulation symbolsmay experience overshooting when subjected to high levels of AWGN.For a more preciseestimation of the amount of rotation and scaling,it is typically necessary to employ the twoaforementioned partitioning met

262、hods.Subsequently,the rotation and scaling amount for theconstellation diagram is calculated separately for each partition using the methods describedabove.The final rotation and scaling amount for the constellation diagram is then determinedby selecting the larger modulus value from the two calcula

263、ted values.Figure 6-9 shows the performance of connectionless transmission using data-onlytechnology.At the transmitting end,an eMUSA extended sequence of length 4 16 isemployed without pilot insertion.The transmission exclusively allocates two PhysicalResource Blocks(PRBs)for transmitting data symb

264、ols and incorporates 84 coded bits for theCyclic Redundancy Check(CRC).The base station is equipped with two receive antennas.Asshown in Figure 6-9,data-only transmission without the need for pilots can closely approachthe performance of MMSE-SIC under ideal channel estimation.The latter represents

265、theoptimal receiver capable of achieving capacity.36/84Figure 6-9 Connectionless Transmission Performance of Data-only TechnologyUltra-low-collision pilot technology based on ultra-sparse pilot and independentmulti-pilot technologiesTo mitigate pilot collisions,the core concept of eMUSA ultra-low-co

266、llision pilottechnology revolves around employing Data-only technology and leveraging data modulationsymbols to complete the majority of channel estimation tasks.This approach minimizes theburden of pilot estimation while maximizing quantity and minimizing collisions.Inultra-low-collision pilot tech

267、nology,the pilot is solely responsible for estimating the spatialdomain combined weights,without the need to estimate multipath selective fading ortime-frequency offset.Figure 6-10 shows the specific eMUSA ultra-low-collision pilot technology.Intraditional practice,pilots necessitate the estimation

268、of the channel across the entiretime-frequency resource.As a result,pilot symbols must cover the entire time-frequencyresource,incurringsubstantialoverheadandexhibitingdensedistribution.Uponimplementing Data-only technology,each pilot may require channel estimation at just onetime-frequency position

269、.This estimation is typically utilized for computing combinedweighting across multi-antenna space domains,while the remaining channel estimation tasksare fulfilled through data modulation symbols.Therefore,each pilot may exclusively occupya single time-frequency position,minimizing overhead and achi

270、eving utmost sparsity.Whenoperating with identical pilot overhead,an ultra-sparse pilot scheme enables maximum pilotquantity while minimizing instances of pilot collisions.Moreover,typically only one pilot exists within traditional data transmission frames.Inthe event of a collision,precise detectio

271、n becomes unattainable.Another major part ofeMUSA ultra-low-collision pilot technology is independent multi-pilot technology:multiplepilots are included in a single transmission,and the pilots are independent and unrelated toeach other.The probability of multiple independent pilots from different us

272、ers colliding37/84simultaneously is much lower than that in traditional single-pilot scenarios,as shown inFigure 6-10.By employing iterative multi-user receivers,the base station can successivelydecode user data associated with collision-free pilots in each iteration.Subsequently,thesesignals are re

273、constructed and eliminated from the receiving end,continuing this iterativeprocess until all decodable user data is resolved.Figure 6-10 Schematic Diagram of Ultra-Low-Collision-Rate Pilot TechnologyBecause of the effectiveness of data-only technology and ultra-low-collision pilottechnology in mitig

274、ating or completely avoiding pilot collision issues,eMUSA connectionlesstransmission can efficiently leverage multi-user multiplexing capabilities in time,code,andpower domains to support a high user load,even under minimal connectionless transmissionconditions.Figure 6-11 shows the performance of e

275、MUSA connectionless transmission,withsimulation parameters detailed in Table 6-2.eMUSA enables simultaneous data transmissionfor up to 100 users without requiring individual connections,even with 32 receive antennas atthe base station.Moreover,the average spectral efficiency can surpass 1 bps per an

276、tenna.Thisrepresents a significantly high connection density and spectral efficiency,particularlyachieved in connectionless transmissions.Table 6-2 Simulation Parameters for Evaluating eMUSAConnectionless TransmissionPerformanceParametersAssumptionsCarrier frequency5GHzWaveformCP-OFDMNumerology15kHz

277、 subcarrier spacing,1 mssubframe with 14 OFDM symbolsResource unit6RBs(1.08 MHz),1 msRS overhead2/7TBS per UE40/60/80 Bytes38/84Modulation andcoding schemeBPSK,LDPC encoding(coderate:0.46670.9111)UE antennaconfiguration1 TxBS antennaconfiguration32 RxPropagation channel&UE velocityCDL-A 30ns;CDL-D 3

278、0ns;UE velocity:3km/h.Timing offset04.7 us,uniformly distributedFrequency offset200 Hz 200 Hz,uniformlydistributedDistribution ofavg.SNREqualSpatial combiningMMSE+ICFigure 6-11 Performance Evaluation Results of eMUSAConnectionless TransmissionIn scenarios demanding high connection density,low latenc

279、y,and high reliability,eMUSA connectionless transmission makes use of additional independent multi-pilottechnologies,for example,employing four to six independent ultra-sparse pilots.Thisapproach aims to achieve ultra-low collision rates,ensuring highly reliable connectionlesstransmission.The simula

280、tion parameters are shown in Table 6-3.Upon analyzing thesimulation performance illustrated in Figure 6-12,it is evident that eMUSA can facilitatesimultaneous connectionless transmission of data for up to 12 users,utilizing 32 receiveantennas while maintaining a reliability level lower than 10-6.No

281、connection naturally implieslow-latency transmission,and for this evaluation,we employed a low-latency receiverwithout interference cancellation.Therefore,achieving low-latency and high-reliabilityperformance in connectionless transmission for a large number of terminals is feasible.Table 6-3 Simula

282、tion Parameters for Evaluating Independent Multi-Pilot PerformanceParameterValue39/84Carrier frequency5 GHzSubcarrier spacing30 kHzTransmission resource6 RB,1 slot(0.5 ms)Number of UE8,12Modulation and codingschemeBPSK,LDPCSize of transport block32 bytesLength of CRC16 bitsNumber of independentpilot

283、s4,5,6Antenna configuration32 RxChannel model and delayspreadTDL-D 30 ns,3 km/hTime offsetUniform within 0,1CPFrequency offsetUniform within 250 HzReceiver algorithMMSE(withoutinterference cancellation)Figure 6-12 Performance Evaluation Results of Independent Multi-Pilot Technology(III)Efficient Con

284、nectionless Transmission in V2V ScenariosFurthermore,the application of eMUSA connectionless transmission extends beyondcellular-based IoT scenarios to encompass a wider range of applications,including V2Vcommunication scenarios 25.High-density IoV information transmission exhibits bothmassiveness a

285、nd burstiness,thus necessitating low latency and high reliability.Meeting therequirements for low latency and high reliability in scenarios of massive and burstyinformation transmission poses a considerable challenge.Moreover,the rapid movement ofvehicle nodes results in swift changes in the network

286、 topology of the IoV,making it evenmore challenging to simultaneously meet the requirements for massive,bursty,low-latency,40/84and high-reliability communication.The primary reason for the decline in eV2X performance during large-scale vehicleinformation exchange in high-density vehicle areas is th

287、e conflict arising from resourcecompetition.This challenge is being encountered by both LTE-V and IEEE802.11p/IEEEl609.eMUSA connectionless transmission,coupled with efficient full-duplex technology,holds thepotential to address this challenge without requiring pre-established connections.Specifical

288、ly:In scenarios with frequent information exchange among numerous vehicles,the situationmay arise where multiple vehicles broadcast their messages simultaneously.Withoutcollaboration between vehicles,each vehicle autonomously determines its signal generationand transmission resource selection.Conseq

289、uently,instances arise where different vehiclesopt for the same transmission resource and utilize identical signal generation methods,resulting in transmission collisions.This will pose significant challenges for other vehicles inreceiving these broadcast messages.The decentralized eMUSA connectionl

290、ess transmissionenables easier demodulation of superimposed multi-user signals in V2V.Moreover,traditionalV2VinformationtransmissionschemeslikeLTE-VandIEEE802.11p/l609 are fundamentally rooted in a half-duplex transmission mechanism.Thismeans that when a vehicle transmits data on a specific channel,

291、it is unable to simultaneouslyreceive data on the same channel.Consequently,this leads to the issue of missed reception:Ifmultiple vehicles broadcast their data packets on the same channel,they will be unable toreceive each others transmissions,which significantly compromises high reliability.Whilee

292、nhancements in the sensing mechanism and increased time-frequency resources can decreasethe missed reception rate,ensuring zero missed receptions remains a challenge.Thefull-duplex mechanism effectively eliminates the issue of missed receptions by enabling eachvehicle to transmit and receive simulta

293、neously on a single channel.It resolves the issue ofmissed reception using significantly less spectrum.Nevertheless,traditional full-duplexcommunication encounters significant self-interference.Even with the application of intricateself-interference cancellation methods,achieving complete cancellati

294、on remains challenging,thus substantially compromisingthe reliabilityof V2V transmission.TheeMUSAconnectionless transmission scheme leverages the unique characteristics of V2V transmissionto achieve a highly efficient V2V communication system,capitalizing on the following twoaspects:1)The vehicle ha

295、s a significantly larger volume compared to standard communicationterminals.2)Information becomes increasingly crucial as the vehicle approaches.To minimize self-interference,it is recommended to position the transmit and receiveantennas as distantly as possible,as shown in Figure 6-13.This approach

296、 ensures that even inthe absence of self-interference cancellation,the full-duplex self-interference signal does notsignificantly surpass the intended signal.Moreover,eMUSA connectionless transmissionmaximizes the multi-user multiplexing capacity in the space,code,and power domainswithout direct con

297、nection.It treats full-duplex self-interference as the target signal,ensuringhigh reliability of connectionless transmission in full-duplex scenarios.Furthermore,this41/84approach eliminates the need for extra self-interference cancellation modules,furtherstreamlining full-duplex V2V communication.F

298、igure 6-13 eMUSAConnectionless Transmission for V2VThrough the integration of eMUSA connectionless transmission technology and efficientfull-duplex technology,numerous vehicle terminals can directly exchange informationwithout the need to listen before transmission.This mitigates the issues of misse

299、d receptionand hidden nodes present in traditional technologies.Ultimately,this integration facilitatesV2V direct communication with ultra-low latency and ultra-high reliability.Simulation resultsindicate that in a high-density IoV scenario,as shown in Figure 6-14,this method reducestransmission lat

300、ency to 1/5 to 1/10 compared to traditional methods while enhancingreliability by 1 to 3 orders of magnitude.Figure 6-14 Performance of the eMUSAConnectionless V2V Transmission Scheme42/846.1.3 Multi-User Encoding and Decoding Schemes Based on ODMA(XDU)Current Code Division Multiple Access(CDMA)meth

301、ods like DS-CDMA,Sparse CodeMultiple Access(SCMA),and IDMA,all employ a technique known as repetition,alsoreferred to as spread spectrum.In SCMA,certain repetitions are substituted with 0(representing idle transmission),forming an extended sequence that includes both periods ofidle transmission and

302、partial repetitions.In IDMA systems,repetition codes are commonlyemployed as internal codes to improve iterative decoding processes.Compared to usingrepetitions,we have found that incorporating a significant amount of idle time can lead toimproved decoding performance.Therefore,references 26 and 27

303、introduced a schemeknown as ODMA.Each user employs an identical channel code with a length of m.Following BPSK modulation,the encoded bits are transmitted using random hopping.The mencoded bits are randomly scheduled and transmitted using n time slots,that is,m time slotsfor signal transmission,whil

304、e the remaining n m time slots remain idle.Each users slotselection rule,whether for transmission or idle,is referred to as on-off mode,with each userhaving a unique on-off mode.This approach ensures that only a small number of users canaccess the channel concurrently within each time segment,leadin

305、g to a significant reductionin multi-user overlapping burden and decoding complexity.Thanks to its ultra-sparse accesscharacteristics,the iterative multi-user decoding can essentially be seen as decoding on atree-shaped factor graph,forming the foundation of this approach.Unlike other CDMA schemes,O

306、DMA operates independently of repetition(spreadspectrum)and user interleaving.Substituting vacancies for repetitions causes the decodingfactor graph to become ultra-sparse within a very short period,thereby enhancing theperformance of iterative decoding.Therefore,the ODMA scheme enables a notableenh

307、ancement in multi-user decoding performance,while also significantly reducing decodingcomplexity.This section presents an unidentified multiple access method designed for mMTC 27.For an unidentified multiple access system with Ktotusers and Ka,Ka Ktotactive users,the number of active users is known

308、at the receiving end.In n transmissions,each active usercollectively transmits B bits of information.Active user information is segmented into twoparts,with lengths Bsand B Bsrespectively.The first part serves as index information,influencing the choice of transmission channels for the coded data.Fo

309、llowing channel coding,the latter part is transmitted over the selected channels,as shown in Figure 6-15.Specifically,the process begins by generating a binary regular sparse matrix A 0,12Bsnas the on-off mode codebook,with defined Hamming weights as ncand nc2Bsnfor its rows and columns.The first pa

310、rt of information is transformed into decimal valuesM,and row M of A selected through indexing modulation becomes the on-off mode codebookaMfor the active user.Subsequently,wk 1,1ncis obtained from the second part of theinformation following channel coding and BPSK modulation.ncsymbols are thentrans

311、mitted on the designated channels as per the on-off mode codebook aM,while theremaining n ncchannels remain inactive.In particular,aM=aM,1,aM,2,aM,n,aM,j0,1,when aM,j=1 indicates channel jfor information transmission,transmission signal xk43/84is finally produced for dissemination.Figure 6-15 Encodi

312、ng Process ofActive User kRegarding the aforementioned encoding scheme,this section presents a two-stageiterative algorithm for the collective restoration of the encoded data 27,as shown in Figure6-16.The initial phase involves detecting the selected on-off mode to retrieve the first part ofuser inf

313、ormation.The fundamental concept involves treating the encoded data as a randomvariable based on the received signal and exhaustively exploring all potential on-offmodes.Employing the Bayesian rule,the on-off mode with the highest Log-Likelihood Ratio(LLR)is chosen.In the subsequent phase,a collabor

314、ative iterative multi-user decodingscheme is developed using the on-off modes identified in the initial phase to retrieve thesecond part of user information.The decoding output,is then fed back to the initialstage for further iteration.Through the joint iterative process,the output,from thesecond st

315、age gradually updates the on-off modes obtained in the first stage,thus facilitating atwo-stage iterative process.The iteration stops once all user data is fully restored or when themaximum iteration limit is reached.Figure 6-16 Two-Stage Iterative On-off Mode Detection and Data Decoding44/84Figure

316、6-17 Performance Comparison Between ODMAand IDMA/SCMAFigure 6-17 compares the decoding bit error rate(BER)for uplink multiple accesssystems with eight users using ODMA and IDMA/SCMA transmissions,respectively.Toensure a fair comparison,all users have a fixed information length of 4096 and employiden

317、tical code rates,specifically utilizing a non-systematic repeat accumulate(RA)code witha 1/3 rate.In the context of IDMA,a repetition code(spread spectrum)of length 4 is appliedas the concatenated code following the RA code.In the case of ODMA,it inserts idle slotsthat are three times the code lengt

318、h.In the context of SCMA,it repeats twice and inserts idleslots that are twice the original length.Therefore,all three transmission schemes haveidentical total transmission lengths and rates.The decoding process involves a joint iterativedecoding technique incorporating multi-user parallel soft inte

319、rference cancellation(PSIC)based on elementary signal estimation(ESE).ODMA transmission yields a performance gainof approximately 0.65 dB over IDMA.In the low BER region,IDMA demonstratessuperiority over SCMA,attributed to the utilization of user interleaving.Figure 6-18 Performance Comparison Chart

320、 of Unidentified RandomAccess Scenarios45/84Figure 6-18 shows the Per-User Probability of Error(PUPE)curves for the proposedscheme,the existing recovery scheme with separate modes and data requiring preambletransmission,and the mode-known scheme in the scenario with unidentified random accessand the

321、 use of the same channel coding.Observations indicate that with a PUPE of 0.01,thisscheme delivers performance gains of 0.5 dB,0.7 dB,and 0.9 dB for 50,150,and 250 activeusers,respectively.Additionally,the proposed performance closely aligns with that of themode-known scheme.Therefore,the proposed s

322、chemes demonstrate superior performanceacross varying numbers of active users compared to the recovery scheme with separate modesand data.Figure 6-19/0Value Comparison Chart of Different Schemes When PUPE=0.05Figure 6-19 shows the values of/0required in the existing scheme and theproposed scheme whe

323、n the PUPE is 0.05.It can be seen that,when 200,the existing scheme differed from the achievable boundary by more than 2 dB.Furthermore,due to ODMAs ultra-sparse access characteristics and the benefits of joint iterative sparsegraph multi-user decoding,this scheme has significantly lower decoding co

324、mplexitycompared to the LDPC-MMSE scheme and is comparable to the BiG-AMP scheme.6.1.4 Sparse IDMAfor UMA(China Mobile)IMT 2030 outlines six key scenarios,with three being natural extensions of the majorscenarios in the 5G era,as shown in Figure 6-20 4.To meet the requirements of future 6Gimmersivec

325、ommunication,ultra-large-scaleconnectivity,URLLC,andubiquitousconnections,it is necessary to research new multiple access technologies.Among theserequirements,ultra-large-scale connectivity is the most challenging and urgent;therefore,itwill be the focus of this white paper.In contrast to NOMA in th

326、e 5G era,the novel multipleaccess feature of 6G is evident in its ability to accommodate a larger number of users,such as300 users.Enabling random access of user equipment,the majority of traditional 5G schemespresume knowledge of user identities,whereas in 6G,this information is unknown,necessitati

327、ng signal detection and estimation.By employing novel pilot design and signal46/84detection technologies,the traditional pilot designs of 5G prove inadequate for fulfilling theactivity detection and channel estimation requirements for as many as 300 users in 6G.Figure 6-20 Six Scenarios and Four Pri

328、nciples of IMT-2030Sparse IDMA combined with compressed sensing represents a crucial technical schemefor the URA of a large number of users 28.The fundamental concept behind the sparseIDMA scheme involves the integration of two technologies.The first technology employspilot encoding based on compres

329、sed sensing.Enabling transmission without physical layeridentification requires the indication of unique user information,which involves compiling acomprehensive codebook containing interleaving patterns,bit repetition counts,and zero fillnumbers.The codebooks codeword sequence number is utilized to

330、 distinguish the uniqueinformation of various users.The codebook number is mapped into a shorter pilot aftercompressed sensing.This pilot is then appended to the data to form a compound packet fortransmission.The second technology is sparse IDMA superimposed coding,which involvesrepetition zero fill

331、ing interleaving superimposition.It enhances resistance tomulti-user interference through bit repetition,effectively diminishes multi-user interferenceby incorporating a large number of zero elements,distinguishes users via diverse interleavers,and randomizes multi-user interference.In this scenario

332、,we assume that the interleavingpattern differentiates between users,even though the receiving end is already known,whichcontradicts the characteristic of having no physical layer identification.Therefore,sparseIDMA uses IDMA as the basic coding paradigm to repeat,zero-fill,and interleave each user,

333、forming a sparse superimposed signal of multiple users to reduce user interference.47/84Figure 6-21 SystemArchitecture of Sparse IDMA28Figure 6-21 shows the system architecture of sparse IDMA,with the first part performingpilot encoding.A method for pilot encoding involves utilizing the Fast Fourier transform(FFT)matrix.This approach entails randomly interleaving and puncturing the rows of anortho