《未來移動通信論壇：2025年通感一體賦能數字低空網絡白皮書（英文版）（128頁）.pdf》由會員分享，可在線閱讀，更多相關《未來移動通信論壇：2025年通感一體賦能數字低空網絡白皮書（英文版）（128頁）.pdf（128頁珍藏版）》請在三個皮匠報告上搜索。

1、ContentsAbstract.11ISAC Requirements for Digital Low-altitude Networks.21.1 Evolutionary Trends in Digital Low-altitude Development.21.1.1 Domestic and International Policies on Low-altitude Economy.21.1.2 Low-altitude Market Size and Expectations.41.1.3 Core Technical Domains in Low-altitude Techno

2、logy.51.2 ISAC Requirements for Different Scenarios.62Technological Evolution and Implementation Challenges in ISAC Technology.102.1 Development Status.102.2 Challenges of ISAC in Empowering Digital Low-altitude Applications.173ISAC Theory.213.1 Technical Performance Requirements.213.1.1 Communicati

3、on Performance.213.1.2 Sensing Performance.233.1.3 ISAC Performance.263.2 Sensing Channel Modeling.283.3 Multi-modal Integration Theory.303.3.1 Low-altitude Sensor and Environmental Modeling Technology.303.3.2 Environment Characterization Algorithm.314Low-altitude ISAC Architecture and Key Technolog

4、ies.334.1 Network Architecture.344.1.1 Overall Digital Low-altitude Architecture.344.1.2 ISAC Architecture.354.2 Coverage Enhancement Technology.394.2.1 Coverage Range Expansion.394.2.2 Continuous Coverage Technology.414.3 Sensing Accuracy Improvement Technology.464.3.1 Waveform Design.464.3.2 Sensi

5、ng Reception Algorithm.504.3.3 Integrated Sensing.604.3.4 Non-ideal Factors and Elimination Methods.634.4 Sensing Reliability Improvement Technology.664.4.1 NLoS Sensing Method.664.4.2 Clutter Suppression Method.694.5 Interference Suppression.754.5.1 Interference Analysis.754.5.2 Interference Coordi

6、nation and Power Control.764.5.3 Adaptive Beam Nulling Scheme in Collaborative Sensing.784.5.4 Reference Signal Silencing Algorithm.815Impact of ISAC Standardization.855.1 Reference Signal Design.855.2 TDD Frame Structure Configuration.885.3 Resource Preemption and Avoidance.895.3.1 Resource Preempt

7、ion Indication.915.3.2 Resource Avoidance.935.4 Signaling Interaction.935.4.1 Signaling Interaction between Base Stations.935.4.2 UE-Base Station Signaling Interface.966Field Trial of Low-altitude ISAC.1026.1 Indoor Trial.1026.2 Field Trial.1066.2.1 Low-Frequency Field Trial.1066.2.2 High-frequency

8、Field Trial.1107Summary and Outlook.113List of Abbreviations.115Members of the Editorial Team.119References.1201/124AbstractThe low-altitude economy constitutes a comprehensive ecosystem built upon digitallow-altitude networks,with unmanned aerial vehicles(UAVs)serving as the operationalcornerstone.

9、This framework integrates diverse sectors including express logistics,urbangovernance,agriculture,forestry and plant protection,and emergency response systems.Driven bysynergistic advancements in market demand,technological innovation,policy frameworks,andregulatory paradigms,the low-altitude econom

10、y is emerging as a transformative growth vectorwithin the global economy,catalyzing innovative operational models across multiple industries.The sectors pertinent to the low-altitude economy primarily focus on four technologicalapplications:communication,navigation,sensing,and management.In this fra

11、mework,theintegrated sensing and communications(ISAC)stands as a significant supportive technology forthe envisioned digital low-altitude network.The ISAC pertains to a cutting-edge informationprocessing framework that equips communication devices,such as base station(BS)or userequipment(UE)with dua

12、l capabilities of communication and sensing,achieved throughspecialized designs including shared hardware facilities,spectrum resource,and signal processing.Integrated sensing and communication(ISAC)technology offers cost-efficiency,ubiquitouscoverage,and high-precision,enabling telecom operators to

13、 pioneer novel sensing services.Thisinnovation drives service paradigm evolution and business model transformation,establishing agroundbreaking information service framework for 5G-Advanced and 6G network evolution.This white paper conducts an in-depth analysis of ISAC core technologies in digitallo

14、w-altitude scenarios.It systematically introduces the technical requirements for digitallow-altitude scenarios,current advancements,and existing challenges in ISAC development.Thenit addresses five critical barriers:lack of ISAC theory,insufficient low-altitude stereoscopiccoverage,limitedsensing ac

15、curacy,insufficient sensing reliability,and complexglobalinterference.Corresponding solutions are methodically presented through dedicated chapters onISAC theory and low-altitude ISAC architecture,and key technologies.To facilitate practicalimplementation,the paper evaluates the impact of related te

16、chnologies on standardization andintroduces the experimental validation of ISAC in low-altitude scenarios.This white paper is protected by copyright laws.Any entity or individual reproducing,excerpting,quoting,or otherwise utilizing the content or viewpoints contained herein mustproperly attribute t

17、he source.2/1241 ISAC Requirements for Digital Low-altitude NetworksThis chapter systematically outlines the current status and trends of low-altitude economicdevelopment.It begins with a comprehensive analysis of both domestic and international policyinitiatives in legislative assurance,infrastruct

18、ure development,and financial support from a policyperspective.Following this,the chapter delves into market size and development forecasts for thelow-altitude economy,drawing from market research data and emphasizing key technical areassuch as communications,navigation,sensing,and management.Finall

19、y,it examines the demandsof industry development,concentrating on typical applications like express logistics,agriculturalplant protection,emergency rescue,low-altitude security,UAV navigation obstacle avoidance,andinspection patrols,to provide a nuanced analysis of the core requirements for ISAC te

20、chnologiesin digital low-altitude advancement.1.1 Evolutionary Trends in Digital Low-altitude Development1.1.1 Domestic and International Policies on Low-altitude EconomyThe global low-altitude economy evolves through three distinct phases:applicationexploration,standardized development,and widespre

21、ad application.The application explorationphase spans from the late 18th century to 2006.This era began with the successful test of hot airballoon technology in Paris,France,marking the inception of low-altitude economy.In 2006,BPp.l.c.pioneered the use of UAVs for monitoring offshore oil platforms,

22、thereby significantlyadvancing the industrial application of UAVs.The standardized development phase,from 2006 to2019,was marked by significant advancements.In 2006,NASA(National Aeronautics and SpaceAdministration)and FAA(Federal Aviation Administration)collaborated to push forward thedevelopment o

23、f the UAV Traffic Management(UTM)system in the United States,establishing afoundation for safe UAV operations.In 2019,EASA(European Union Aviation Safety Agency)released two comprehensive sets of UAV regulations,rigorously defining the UAV standards andoperational requirements across Europe.The wide

24、spread application phase continues from 2019 topresent.Notably,in 2020,the U.S.President enacted the Low-altitude Flight Safety Act,streamlining the UAVs low-altitude flight authorization process,thereby accelerating the broad3/124application of UAV technology.In parallel,the U.S.government has impl

25、emented comprehensivesupport mechanisms for the electric Vertical Take-off and Landing(eVTOL)industry throughestablishing large-scale R&D initiatives,direct research funding allocations,and specializedfinancial instruments.Notably,NASA has spearheaded pivotal research on core eVTOL/Urban AirMobility

26、(UAM)technologies while providing substantial grants to academic institutions andresearch organizations.From 2021 to 2023,the European Union implemented remote IDregulations requiring real-time broadcast of UAV identification and location information duringoperations,significantly improving aviation

27、 safety.The“Horizon 2020”and“Horizon Europe”frameworks have successfully incubated numerous urban air mobility projects and enterprises.The EU demonstrated regulatory foresight by establishing the worlds first eVTOL-specificcertification basis under its airworthiness framework,issuing pioneering ope

28、rational approvals toindustry leaders such as Volocopter.The Chinese government places significant emphasis on advancing the low-altitude economy.Consequently,it has systematically issued relevant laws,regulations,policies,and industrialstandards from both national and local perspectives.In February

29、 2021,the concept of low-altitudeeconomy was written into national planning for the first time.In December 2023,the CentralEconomic Work Conference included the low-altitude economy as a strategic emerging industry.In March 2024,the low-altitude economy was included for the first time in the Report

30、on theWork of the Government presented at the State Councils Two Sessions,defining it as a“newgrowth engine”.Since 2024,27 provinces(municipalities and autonomous regions)havehighlighted the development of the low-altitude economy in their individual government workreports.This series of advancement

31、s illustrates that the low-altitude economy is progressivelybecoming a vital component of national economic development.The Implementation Plan forInnovative Application of General Aviation Equipment(2024-2030)aims to cultivate atrillion-yuan market scale for the low-altitude economy by 2030,facilit

32、ating infrastructuredevelopment,stimulating financial investment,and broadening application scenarios.To enhanceinfrastructuredevelopment,localgovernmentsareencouragedtointegratelow-altitudeinfrastructure into urban planning,supporting the exploration and advancement of pilot projectsfor establishin

33、g take-off and landing points on rooftops,land,and water scenarios.Concurrently,4/124at the governmental level,promoting supportive measures in terms of terrestrial infrastructuredevelopment and financial investments is key to advancing the low-altitude economy andexpanding its application scenarios

34、.Numerous provinces and cities have now included thelow-altitude economy within their government work reports,rolling out pertinent action plansaimed atthe high-quality advancementof the low-altitude economy.These highlightspredominantly pertain to infrastructure development and the expansion of dow

35、nstream applicationscenarios.In conclusion,both domestically and internationally,there is a pronounced attitude ofproactive encouragement towards low-altitude policies.The low-altitude economys growth isbeing fostered through various methods,including policy support,legislation,infrastructureestabli

36、shment,and financial investment.Simultaneously,regulatory measures are being fortifiedto ensure flight safety and the sustainable development of the industry.1.1.2 Low-altitude Market Size and ExpectationsCurrently,the global low-altitude economy is advancing through a stage of furtherapplication an

37、d popularization.The global low-altitude economy,particularly the UAV marketsegment,has exhibited remarkable growth dynamics over the past decade.The financialinvestment to the UAV industry has escalated sharply from USD 121 million in 2013 to USD4.806 billion in 2022,nearly a 40-fold increase.With

38、continuous technological innovation andsustained market demand drivers,this sector is projected to maintain its strong growth momentumin the future.Chinas low-altitude economy achieved a market size of RMB 505.95 billion in 2023,registering a year-on-year growth rate of 33.8%.By the end of 2024,the

39、market size of Chinaslow-altitude economy is expected to continue its expansion,potentially surpassing the RMB 700billion threshold.According to the estimations by the Civil Aviation Administration of China,by2025,the market size for the low-altitude economy is poised to exceed RMB 1,064.46 billion,

40、with a 3-year compounded annual growth rate(CAGR)of 28.1%.By 2026,its expected to reachRMB 1.5 trillion,with projections for 2035 aiming for RMB 3.5 trillion,reaching a ten-yearCAGR of 8.84%.In summary,the market size of the low-altitude economy has experienced substantial growthboth domestically an

41、d internationally over the past few years,and is projected to continue its5/124robust development in the upcoming years.Investment institutions widely recognize the growthpotential of the low-altitude economy sector,with market projections estimating its scale tosurpass the trillion RMB threshold.1.

42、1.3 Core Technical Domains in Low-altitude TechnologyThe development of low-altitude economy requires robust support from four foundationaltechnological pillars:communication,navigation,sensing,and management.Communicationstechnology serves as the cornerstone of the low-altitude economy,primarily fa

43、cilitating thetransmission of information between UAVs and ground control centers.The integration of 5G and5G-A technologies provides high-speed,low-latency communication networks,enhancing thesafety and efficiency of low-altitude flights.Low-altitude aerial vehicles must maintain real-timecommunica

44、tion with the ground control centers to receive instructions and transmit flight stateinformation.Currently,5G-A networks and low-earth orbit satellite communications are theprimarymeansforlow-altitudecommunication.Additionally,AutoDependentSurveillance-System(ADS-B)and secondary radar systems are a

45、lso utilized as means forlow-altitude communication.NavigationtechnologiesencompassBSpositioningtechnology,satellitenavigationtechnology,radar navigation technology,visual navigation technology,and inertial navigationtechnology.For low-altitude aerial vehicles,these technologies assist in determinin

46、g the exactlocation and trajectory of the aircraft.The“Implementation Plan for Innovative Application ofGeneral Aviation Equipment”mentions Beidou Navigation Satellite System,and the applicationof geographic information system(GIS)in cities is also very important.Sensing technology is deployed for t

47、he 3D spatial orientation and identification of variousUAVs,general aviation aircrafts,aerial bird,and airborne objects.It is imperative for low-altitudeair domain monitoring,essential territory intrusion detection,and enables real-time monitoring,tracking,safety alerts,and more concerning low-altit

48、ude sensed targets.Management technologies address the management and monitoring of low-altitude airdomains,encompassing UTM systems,air domain planning,coordination services,as well ascommunications and navigation services.The advancement of the low-altitude economy hasgenerated a critical need for

49、 management of low-altitude air domain.The UAV traffic6/124management,monitoring and management of low-altitude air domain,as well as the developmentand operation of associated communication and navigation facilities,are important directions forthe industry development.Moreover,the implementation of

50、 the low-altitude economy is heavilyreliant on sophisticated technical support,primarily encompassing air traffic control systems,communications,navigation,and surveillance.In summary,the advancement of low-altitude economy relies on synergistic developmentacross four core technological pillars:comm

51、unication,navigation,sensing,and management.Eachdomain serves an indispensable function in ensuring the safe and efficient operation of thisemerging economic sector.ISAC technology,in particular,is poised to substantially expand theservice boundaries of mobile communication networks,enabling the del

52、ivery of enhancedprecision,intelligent capabilities,efficient performance,and ubiquitous sensing services to UE.1.2 ISAC Requirements for Different ScenariosThe low-altitude economy has emerged as a vital component of the new quality productiveforces.With advancements in relevant technologies and th

53、e refinement of policies,low-altitudeaerial vehicles have become widely utilized across sectors such as express logistics,agriculturalplant protection,emergency rescue,low-altitude security,facilitating diverse economic activitiesand contributing to economic growth1.As a groundbreaking advancement i

54、n 5G-A and 6Gmobile communication systems,ISAC technology revolutionizes network capabilities by enablingboth ultra-high-speed data transmission and comprehensive sensing of low-altitude AVUs andtheir surroundings.This dual functionality achieves seamless coverage of sensing areas whileaddressing th

55、e inherent limitations of conventional radar systems2.Deployable across diverselow-altitudescenarios,thistechnologyfacilitatesreal-timedatatransmissionthroughcommunication network while enhancing low-altitude UAV mission efficiency via integratedsensing capabilities.Such advancements catalyze the de

56、velopment of intelligent low-altitudeeconomy networks incorporating intelligent resource allocation,secure data exchange protocols,and continuous operational monitoring systems3.Particularly critical for low-altitude securityand collision avoidance applications,ISAC technology enables effective dete

57、ction and tracking ofunauthorized UAVs to guarantee air domain security.Concurrently,its massive antenna-enabled7/124high-accuracy sensing assists authorized UAVs in precision navigation and collision avoidance,optimizing airspace utilization efficiency while mitigating risks of unauthorized flight

58、paths andoperational non-compliance.Six principal operational scenarios are presented below to delineatethe technical requirements of digital low-altitude systems for ISAC.1)Express Logistics ApplicationsRising living standards have driven growing urban demand for time-sensitive logisticssolutions w

59、ith enhanced operational efficiency.Conventional logistics systems face operationalconstraints from weather volatility and traffic congestion,exhibiting limitations in both efficiencyand adaptability.UAV-based delivery presents a viable alternative that mitigates terrestrial trafficburdens while ena

60、bling dynamic scheduling optimization.These advantages underpin its broadapplication potential in modern logistics4.ISAC technologies can play a crucial role across allstages of UAV delivery.The BS utilizes low-altitude ISAC technology to effectively plan UAVroutes and sense environmental conditions

61、,thereby ensuring flight safety and preventing collisions.By leveraging real-time data on weather,traffic,and environmental conditions,it can dynamicallyadjust transportation routes to mitigate issues like traffic congestion and adverse weather,enhancing transportation efficiency and enabling real-t

62、ime UAV status monitoring.Furthermore,the system architecture supports adaptive UAV flight scheduling aligned with fluctuating servicedemands,achievingmission-criticaldeliveryperformancethroughintelligentresourceallocation5.2)Agriculture,Forestry and Plant Protection ApplicationsIn agricultural oper

63、ations,there is a critical need for real-time crop state monitoring.Throughthe interpretation of monitoring data,one can perform essential actions such as irrigation,fertilization,ventilation,and pest control,enabling optimal crop growth.The ISAC UAV networkhas a wide range of applications in this f

64、ield.Multiple UAV devices are interconnected throughthe communication network,and carry high-resolution cameras and various sensors to collectaccurate remote sensing data in real time 6.The captured data are transmitted via ISAC-enablednetworks to centralized data processing facilities for comprehen

65、sive analytics,generatingactionable insights for agricultural optimization.3)Emergency Response Applications8/124Low-altitude ISAC implementations significantly improve emergency response metricsthrough enhanced disaster situational awareness,improved data acquisition accuracy,andoptimized resource

66、deployment efficiency.In the aftermath of natural disasters,such asearthquakes,floods,or fires,UAVs equipped with sensing equipment can swiftly reach disasterzones,capturing real-time images and temperature data through high-resolution cameras andthermal imagers.The low-altitude communication networ

67、k facilitates real-time feeds of situationaldata to command centers,establishing an evidentiary foundation for tactical decision-making.Insearch and rescue operations,UAVs deliver critical supplies to inaccessible zones whileintegrating sensor arrays for enhanced target detection,significantly accel

68、erating mission timelines.The ISAC network enables real-time monitoring and data interaction of UAVs.Leveraging thecommunication network,the sensed data are relayed back to dispatch the UAV rescue operationswhile ensuring a swift response to unforeseen events.4)Low-altitude Security ApplicationsIn t

69、he low-altitude scenarios,UAVs are extensively applied,leading to emerging UAV flightsafety concerns.Effective governance and operational control of low-altitude air domainsconstitute essential safeguards for low-altitude security.For authorized UAVs,real-timemonitoring can be conducted using ISAC t

70、echnologies to ensure no conflicts in equipment flightpaths,thereby guaranteeing safety;these UAVs can also be adjusted according to the scenariorequirements to enhance operational efficiency.This technology also enables detection andidentification of unauthorized UAV operations through spectrum sen

71、sing,effectively enforcing airdomain security protocols.5)UAV Navigation and Collision Avoidance ApplicationsIn the future,leveraging the capabilities of large-scale antenna arrays and other advancedtechnologies inherent in 6G networks,ISAC technologies will enable comprehensive networkcoverage in l

72、ow-altitude zones.This will facilitate precise UAV localization and real-timeenvironmental sensing around UAV positions using UAV UE,BSs,relays,and other nodes.Coupled with pre-sensed environmental data,this approach will construct more precise,comprehensive,and real-time navigation maps,thus aiding

73、 air traffic management department inexecuting navigational and obstacle avoidance strategies for authorized UAVs.9/1246)Patrol InspectionApplicationsLow-altitude ISAC technology facilitates structural monitoring of critical urban assetsincluding bridge decks,pavement systems,and high-rise buildings

74、.This real-time monitoringsystem is instrumental in promptly identifying and mitigating potential safety risks,therebyextending the useful life of the infrastructure.In densely populated regions that are challenging toaccess,low-altitude UAVs can conduct inspections of critical infrastructure such a

75、s power linesand pipelines,ensuring their uninterrupted and secure operation.Multidimensional data(includingpopulation mobility dynamics and thermal distribution patterns)collected through low-altitudeISAC-enabled devices provide scientific underpinnings for evidence-based urban planningframeworks.L

76、everagingreal-timeandhistoricaldataanalytics,municipalmanagementdepartment can optimize resource allocation,such as refining waste management procedures andscheduling for public transport vehicles.10/1242 Technological Evolution and Implementation Challenges inISAC TechnologyThis chapter systematica

77、lly provides the developmental milestones and fundamentaltechnical barriers in ISAC technology.First,it focuses on the important progress made instandardization,research and application of ISAC technology under the efforts of authoritativetechnical organizations and standardization institutions in C

78、hina and abroad,specifically coveringsensing technology system architecture,typical application scenarios,key performance indicatorsand core technology breakthroughs;then,an in-depth analysis of the five major challenges facedby ISAC technology in the large-scale application of digital low-altitude

79、fields:lack of ISACtheory,insufficient low-altitude stereoscopic coverage,limited sensing accuracy,insufficientsensing reliability,and complex global interference.2.1 Development StatusUnder the efforts of Chinese and foreign standardization organizations such as theInternational Telecommunication U

80、nion(ITU),the 3rd Generation Partner Project(3GPP),theInstitute of Electrical and Electronics Engineers(IEEE),IMT-2020/2030(International MobileTelecommunications for 2020/22030)Promotion Group,and the China CommunicationsStandards Association(CCSA),great progress has been made in the standardizatio

81、n and relatedresearch work of ISAC.1)ITUThe ITU-R Working Group commenced its comprehensive study on“IMT towards 2030 andBeyond”in February 2022,with the objective of formulating an ITU-R technical report addressing“Future Technology Evolution Trends”.In December 2022,the report M.25167 was formed,w

82、hich clearly listed the ISAC as an emerging technology trend and gave relevant enablingtechnologies,including coordinated sharing of resources such as spectrum and hardware,signalprocessing integration,joint waveforms and unified beamforming solutions,AI-enabled networkcollaboration and multi-node c

83、ollaborative sensing technologies.The draft proposal“Frameworkand Overall Objectives of the Future Development of IMT for 2030 and Beyond”was adopted in11/124June 2023,recognizing ISAC as one of the six major application scenarios of IMT-2030.InNovember 2023,ITU-R formally ratified Recommendation M.

84、2160 concerning the IMT-2030framework8.Within the IMT-2030 framework,ISAC is formally defined as the integration ofsensing and communication capabilities to support new applications and services.This integrationharnesses IMT-2030s expansive multi-dimensional sensing capacities to generate spatialinf

85、ormation about unconnected objects,their dynamics,and surrounding environments.The usagescenarios encompass navigation,activity detection,and motion tracking(such as posture/gesturerecognition,fall detection,vehicle/pedestrian detection),environmental monitoring(such asrainwater/pollution detection)

86、,and other technologies that necessitate support for high-precisionpositioning and sensing capabilities,including range/speed/angle estimation,object detection,positioning,imaging,and mapping.Concurrently,ITU-T has initiated multiple standardization efforts pertaining to ISAC.InNovember 2023,the TR.

87、ISAC-fra9 research report led by China Telecom in SG13 ResearchGroup was officially defined.This report endeavors to investigate the functional requirements andarchitectural framework for ISAC implementation within IMT-2020 and next-generationcommunication networks.In August 2024,SG13 proposed the t

88、echnical report TR.qos-req-isac10,titled“Research on QoS Assurance for Integrated Sensing and Communication(ISAC)inIMT-2020 and Beyond”.This document systematically studies the QoS requirements andarchitectural frameworks for ISAC in IMT-2020 and future networks,aiming to improve theefficiency of mu

89、lti-domain resource coordination.In September 2024,SG16 proposed theH.ILE-ISAC-req framework11 defining technical specifications for Immersive Live Experience(ILE)services leveraging ISAC technologies.The document establishes application objectives forremote operation and assistance scenarios,while

90、formulating implementation requirements todeliver cost-efficient,highly interactive immersive experiences with operational practicality.2)3GPPFundamental research on ISAC was initiated in the 3GPP Release 19.In March 2022,thetechnical report TR 22.837“Study on Integrated Sensing and Communication”12

91、 was initiatedin 3GPP SA1.Subsequent iterative releases have progressively refined the definition of sensingservice use cases and potential requirements for 5G systems across various vertical domains and12/124application scenarios.The 32 use cases documented in TR 22.837 demonstrate 5G systemspotent

92、ial in ISAC,encompassing applications such as smart home intrusion detection,highwaypedestrian/animalintrusiondetection,precipitationmonitoring,transparentsensingimplementations,smart city flood detection,smart home perimeter intrusion detection,railwayintrusion detection,vehicular maneuver assistan

93、ce and navigation,Automated Guided Vehicle(AGV)detection/tracking in industrial environments,and Unmanned Aerial Vehicle(UAV)flighttrajectory tracking.TR 22.837 further proposes a series of novel requirements to support diverseuse cases,specifying both functional capabilities and performance metrics

94、 for 5G networks.Thesespecifications address critical aspects including authorization,network configuration,networkexposure,security protocols,and billing systems.A comprehensive set of KPIs has beenestablished for performance evaluation of sensing services,encompassing positioning estimationaccurac

95、y,velocity estimation precision,sensing resolution,maximum sensing service latency,refresh rate,missed detection probability,and false alarm ratio.Concurrently,RAN1 has initiated foundational research on ISAC channel modeling,including RP-234069 for ISAC channel modeling13 and RP-234018 targeting va

96、lidation of3GPP channel models within 7-24 GHz spectrum14.These research initiatives aim to advanceISAC technological maturity in 5G and 6G networks,facilitating efficient integration ofcommunication and sensing functionalities.RP-234069 is concentrated on establishing a universalchannel modeling fr

97、amework that supports the detection and tracking of multiple targets,whileRP-234018 is dedicated to validating the effectiveness of existing channel models within definedfrequency ranges,providing scientific grounding for the real-world deployment and application ofISAC technology.Through these stud

98、ies,RAN1 aspires to drive forward the development of ISACtechnology,ushering in its pivotal role across diverse application scenarios including unmanneddriving technology,smart transportation system,industrial automation,and other sectors.3)IEEEIn September 2020,IEEE formally launched the 802.11bf15

99、 Task Group,marking theinaugural phase of ISAC standardization initiatives.The establishment of the IEEE 802.11bfstandard group aims to promote the further development of wireless local area network(WLAN)in ISAC functions.The core goal of this standard is to use IEEE 802.11 PHY(Physical Layer)and13/

100、124MAC(Media Access Control)features to obtain measurement data,which can be used to estimatethe characteristics of objects in the area of interest,such as distance,speed,angle,motion.Objectsmay encompass biological entities such as humans and animals,with operational zones spanningresidential,enter

101、prise,vehicular,and industrial environments.The 802.11bf standard is projectedto catalyze ubiquitous deployment of radio access network(RAN)across emerging domainsincluding Internet of Things(IoT)and Vehicle-to-Everything.Wi-Fi sensing technology hasdemonstrated technical viability across multiple u

102、se cases,including proximity detection,gesturerecognition,object counting,and health monitoring.This technology utilizes in-situ Wi-Fiequipment for environmental sensing and activity identification eliminating the necessity for extrasensors or cameras,thereby reducing reliance on specific monitoring

103、 equipment.For instance,within the IoT sector,Wi-Fi Sensing can be leveraged for detecting user presence and monitoringthe environment in smart buildings,in addition to facilitating remote health monitoring.In theVehicle-to-Everything(V2X)sector,Wi-Fi Sensing technology facilitates communication amo

104、ngvehicles and between vehicles and infrastructure,thereby aiding in autonomous driving.To align with the trend of ISAC,the IEEE Communications Society(ComSoc)inauguratedthe ISAC-ETI(ISAC-Emerging Technology Initiative)in the same year.This initiative aims toadvance the research,standardization,and

105、application of ISAC technology,fostering collaborationbetween the academic and industrial sectors.ISAC-ETI enhances and promotes ISAC-relatedacademic and industrial activities by connecting researchers and experts across variouscommunities who share a vested interest in ISAC advancements.These field

106、s include,but are notlimited to,information theory,signal processing,mobile computing,aerospace and electronicsystems(including radar system),as well as vehicle technology and smart transportation system.ISAC-ETIs activities are diverse and specific16,including the organization of workshops,special

107、conferences and specialized conferences.ISAC-ETI successfully convened workshops atWi-Fi Sensing 2022 in Shenzhen,China,and at prominent conferences including ICC 2024 inDenver,USA,and ICASSP 2024 in Seoul,Republic of Korea.In the realm of special sessions,ISAC-ETI hosted a special session on Integr

108、ated Sensing and Communication(ISAC)at SPAWC2021(Lucca,Italy).Notable among specialized events is the IEEE JC&S 2024(Leuven,Belgium),an officially co-sponsored conference with ISAC-ETI.These initiatives have not only enhanced14/124academia-industry collaboration but also established a synergistic pl

109、atform for advancing ISACtechnology development.4)CCSAIn November 2020,the CCSA TC5 WG6 Frontier Wireless Technology Working Groupdefined the Research Report on Integrated Wireless Communication and Wireless SensingTechnology and Solutions 17,and output an internal research report in May 2023.CCSAla

110、unched a dedicated research initiative in August 2022 focusing on 5G ISAC technology,withemphasis on 5G-Anetwork architecture evolution and core wireless innovations18.5)IMT-2020/2030 Promotion GroupIn July 2021,the IMT-2020(5G)established the“ISAC Task Group,”conducting thoroughinvestigations into

111、ISAC application scenarios,network architecture,wireless air interfacetechnologies,and simulation assessments.In December 2021,comprehensive testing andvalidation were completed for the ISAC in the 5G-Advanced framework.On July 31,2022,WangZhiqin,Head of the IMT-2020(5G)Promotion Group and Vice Pres

112、ident of CAICT,released theResearch Report on Requirements for 5G-Advanced ISAC Scenarios 19 at the“InnovativeDevelopment of Integrated Sensing,Communication and Computing”parallel session during the“2022 China Computing Power Conference”.The report emphasizes the ISAC technologiesduring the 5G-A ph

113、ase,exploring application scenarios and technical requirements,ultimatelyproviding guidance for ensuing research and development,as well as standardization activities.Subsequently,a series of reports on ISAC technology were published.Notable among these werethe Research Report on 5G-Advanced ISAC Ne

114、twork Architecture in November 2022,theResearch Report on Evaluation Methods for 5G-Advanced ISAC Simulation in June 2023,and theResearch Report on Requirements for 5G-Advanced ISAC Scenarios V2.0 in September 2023,andthe Research Report on 5G-Advanced ISAC Air Interface Technology Solution in April

115、 2024.Table 1ASeries of Research Reports on ISAC Issued by IMT2020(5G)Promotion GroupTimeReport NameJuly 2022Research Report on Requirements for 5G-Advanced ISAC ScenariosV1.0November 2022Research Report on 5G-Advanced ISAC Network Architecture15/124June 2023Research Report on Evaluation Methods for

116、 5G-Advanced ISACSimulationSeptember 2023Research Report on Requirements for 5G-Advanced ISAC ScenariosV2.0March 2024Research Report on 5G-Advanced ISAC Network Architecture V2.0April 2024Research Report on 5G-Advanced ISAC Air Interface TechnologySolutionIMT-2030(6G)has also initiated comprehensive

117、 research initiatives in ISAC.In April 2021,the Radio Technology Group of the IMT-2030(6G)Promotion Group established an“IntegrationTechnology Subgroup,”comprising an AI Task Group and an ISAC Task Group.In September2021,the first edition of the Research Report on Integrated Sensing and Communicatio

118、n(ISAC)Technology was issued.This research offers a thorough exploration of 6G ISAC,investigating itscurrent situation and development trend,application scenarios,basic theories,and keytechnologies.In November 2022,the second edition of the Research Report on Integrated Sensingand Communication(ISAC

119、)Technology was issued,accompanied by synchronized full-bandISAC testing.On October 27,2023,at the third 6G ISAC Symposium in Xian,IMT-2030(6G)Promotion Group released a series of research reports on ISAC,including Research Report on 6GSensing Requirements and Application Scenarios,Research Report o

120、n Evaluation Methods for 6GIntegrated Sensing and Communication 20,and Research Report on 6G ISAC System Design21.In November 2024,the Research Report on Key Air Interface Technologies for 6GIntegrated Sensing and Communication,the Frontier Report on Key Collaborative SensingTechnologies for 6G Inte

121、grated Sensing and Communication,and the Research Report onSimulation Evaluation and Methods for 6G Integrated Sensing and Communication(Edition 2)were issued.Table 2ASeries of Research Reports on ISAC Issued by IMT2030(6G)Promotion GroupTimeReport NameSeptember 2021Research Report on Integrated Sen

122、sing and Communication TechnologyNovember 2022Research Report on Integrated Sensing and Communication TechnologyV2.016/124October 2023Research Report on 6G Sensing Requirements and ApplicationScenariosOctober 2023Research Report on Evaluation Methods for 6G Integrated Sensing andCommunicationOctober

123、 2023Research Report on 6G ISAC System DesignNovember 2024Research Report on Key Air Interface Technologies for 6GIntegrated Sensing and CommunicationNovember 2024Frontier Report on Key Collaborative Sensing Technologies for 6GIntegrated Sensing and CommunicationNovember 2024Research Report on Simul

124、ation Evaluation and Methods for 6GIntegrated Sensing and Communication(Edition 2)6)FuTURE Mobile Communication Forum(FuTURE)The FuTURE has convened multiple conferences focused on digital low-altitude networks,engaging in extensive discussions on overall planning of low-altitude information infrast

125、ructureand standardization,scenarios and requirements concerning low-altitude economy,and thetechnologies and standards related to ISAC.These deliberations also cover advancements intechnology,networks,applications,and service innovation,injecting renewed dynamism andmomentum into the advancement of

126、 ISAC technology and the future of the industry.On April 16,2024,under the theme“Better Together,Better Future”,the FuTURE MobileCommunication Forum hosted the Global 6G Conference.Mr.Bi Qi,the Chief Scientist of ChinaTelecom,put forward during the conference that as we look towards 6G,the domains o

127、f AI,ubiquitous connectivity,and ISAC will become the three most promising development directions.The ISAC technology-based“low-altitude economy”holds the potential to achieve double-digitrevenue growth for operators and is expected to unlock greater value in scenarios such as UAVinspection,delivery

128、 services,and lightweight logistics solutions.On August 2,2024,the FuTURE Mobile Communication Forum convened the DigitalLow-Altitude Conference.Under the theme“Building a Robust Digital Infrastructure to DriveLow-altitude Development,”the conference delved into the development of low-altitude netwo

129、rks.17/124Fang Xinping,Deputy Director of the Information Development Bureau of the Office of theCentral Cyberspace Affairs Commission,pointed out that the development of low-altitudeeconomy requires strengthening top-level design and layout,and improving related policies,technologies,standards,indu

130、stries and regulatory systems of low-altitude economy.Promoting theinnovative integration of digital technologies(such as 5G,6G,integrated sensing andcommunication,and high-precision navigation)with the low-altitude flight industry,proactivelyplanning low-altitude information infrastructure for the

131、space-air-ground coordination and theintegration of communication,sensing,navigation,and control,such efforts will help build a soliddigital foundation for the development of the low-altitude economy.In 2024,the FuTURE MobileCommunication Forum initiated the formation of the Digital Low-Altitude Wor

132、king Group.At thisconference,the Working Group was formally launched and its vision was released.On August 11,2024,the kick-off meeting of the Digital Low-Altitude Working Group underthe FuTURE Mobile Communication Forum was convened.Mr.Bi Qi,the Chairman of theWorking Group and the Chief Scientist

133、of China Telecom,highlighted that this DigitalLow-Altitude Working Group under the FuTURE Mobile Communication Forum will prioritizethe development of the industry chain,steer advancements in key technologies,including ISACtechnology,and will merge academic research with industrial development.2.2 C

134、hallengesofISACinEmpoweringDigitalLow-altitudeApplicationsWhile facilitating digital low-altitude applications,the ISAC confronts several key issues andchallenges,such as resource integration,coverage,accuracy,reliability,and interference.Theseissues not only impede the progressive development of IS

135、AC technology but also have directimplications for the safety and reliability of low-altitude networks.To guarantee stable operationof digital low-altitude services with superior performance,premium quality,and optimizedefficiency,this subsection focuses on five pivotal challenges:the lack of ISAC t

136、heory,insufficientlow-altitude stereoscopic coverage,limited sensing accuracy,insufficient sensing reliability,andcomplex global interference.1)Lack of ISAC Theory18/124Currently,the performance indexes of communication systems primarily encompasseffectiveness and reliability,exemplified by paramete

137、rs like system capacity,bit error rate,andframe error rate.The main performance indexes of sensing system consist of accuracy,resolution,detection probability,and false alarm probability.The indicators for communication and sensingare completely different.Nevertheless,in physical ISAC networks,both

138、communication andsensing applications are concurrently operational,synergistically enhancing network performanceand user experience.Therefore,it is imperative to adopt standardized performance indexes toassess the overall performance of ISAC system and thereby improve the overall system efficiency.F

139、urthermore,the absence of ISAC theory also extends to the lack of sensing channel modelingmethod and the lack of multi-modal integration theory.2)Insufficient Low-altitude Stereoscopic CoverageLow-altitudestereoscopiccoverageformsthecornerstoneofimplementingdigitallow-altitude applications.Conventio

140、nal BSs,however,are principally optimized for ground UEcoverage.With vertical beam widths typically constrained to 20 and downward-tilted antennaconfigurations for interference mitigation,their effective service altitude rarely exceeds 100meters-substantially inadequate to meet the 300 m+coverage re

141、quirements inherent to digitallow-altitude operational scenarios.Moreover,sensing performance exhibits heightened sensitivityto echo signal quality,incurring more severe path loss compared with communication systems.Current detection ranges remain limited to several hundred meters,making long-distan

142、ce coverageextension a critical technical challenge demanding immediate resolution.Moreover,future digitallow-altitude applications must address the dual demands of“land+air”and“communication+sensing”functionality.Resolving how to maintain the continuity and stability of low-altitudesensing applicat

143、ions while achieving high transmission rates is a critical challenge that the ISACnetwork must address.3)Limited Sensing AccuracyHigh-accuracysensingconstitutesafundamentalrequirementforenablingdigitallow-altitude services.In traditional radar systems,high transmission power and extremelylarge-scale

144、 antenna arrays are employed to guarantee system accuracy.The ISAC system isconstrained by the transmission power,windward side,and propagation losses in low-altitude19/124environments inherent to civilian systems,resulting in a significant gap in sensing accuracycompared to the traditional radar sy

145、stem.In ongoing trials of 4.9 GHz 5G-Advanced technology,the sensing accuracy in low-altitude scenarios ranges from approximately 0 to 20 m.While thismay suffice for basic low-altitude detection applications,it falls short of the sensing accuracyrequired for tasks like trajectory tracking and auxili

146、ary obstacle avoidance.Moreover,the sensing capacity for individual nodes is constrained by their inherentprocessing capabilities,posing challenges to enhancing accuracy.Taking the Cramr-Rao LowerBound(CRLB)-a crucial sensing performance metric-as an example,multi-node CRLB exhibitslinearenhancement

147、characteristicscomparedwithsingle-nodeCRLB.However,theimplementation of multi-node collaborative ISAC technology fundamentally depends oninter-node synchronization accuracy.Therefore,resolving synchronization error mitigation,maximizing the advantages of multi-node collaborative ISAC,enhancing sensi

148、ng accuracythrough multi-node data fusion,and overcoming inherent limitations of single-node sensingconstitute critical challenges requiring immediate solutions.4)Insufficient Sensing reliabilityHigh-reliabilityISAC serves as the fundamental safeguard for secure operations in digitallow-altitude ser

149、vices.In practice,clutter interference and Non-Line of Sight(NLoS)propagationfrom non-target objects directly degrade sensing accuracy,potentially elevating false alarm rateand missed detection rate,thereby severely undermining sensing reliability.Practical systemsreceive echo signals containing not

150、 only target information but also significant static cluttercomponents from fixed objects like ground,buildings,and mountains and dynamic cluttersinduced by moving objects such as birds and clouds.The complexity and unpredictability ofclutter strength pose challenges to clutter elimination and imped

151、e the processes of target detectionand identification.Furthermore,in NLoS propagation scenarios obstructed by buildings,there ispotential for detecting“false targets”,which may significantly deviate from the estimatedlocations of true targets,thus severely undermining the sensing reliability.Therefo

152、re,maximizingthe advantages of collaborative network sensing to achieve intelligent multi-dimensional clutterelimination and NLoS identification and utilization is essential for enhancing the reliabilityoftarget sensing in real-world scenarios and represents a critical issue that demands urgent20/12

153、4resolution.5)Complex Global InterferenceControlling global interference is a prerequisite for the large-scale deployment of digitallow-altitude networks.However,the integration of sensing capabilities introduces interferencechallenges both between sensing domains and between sensing and communicati

154、on.Specifically,the ISAC signals sent to space will inevitably cause long-range interference.Specifically,sensingecho signals are subject to simultaneous interference from uplink UE,adjacent cells,and remoteBSs.Such complex and severe interference not only jeopardizes network stability and reliabili

155、ty,but also directly degrades the accuracy of UE-centric sensing services and quality of experience(QoE)in communication systems.The heterogeneous interference scenarios coupled withdynamic environmental variations present real-time complexities that critically impair bothnetwork communication quali

156、ty and sensing performance.Furthermore,interference signalintensity is not only linked to ISAC resources but also exhibits characteristics of temporalvariation and global reach.A pressing challenge lies in developing a dynamic interferencerecognition and control approach that is both globally applic

157、able and adaptable,enabling theefficient management and intelligent control of multiple interferences.21/1243 ISAC TheoryThis chapter systematically addresses the core issue of the lack of ISAC theory.As discussedin Section 2.2,the lack of ISAC theory manifests primarily in three domains:the incompl

158、ete ISACoverall performance index system,the insufficient sensing channel modeling methods,and theabsence of a multi-modal integration theory.Addressing this systemic challenge,this chapterinitially delves into the basic theory level,outlining the performance evaluation indicator systemsfor both com

159、munication and sensing.It then introduces integrated performance indexes for ISAC.Subsequently,it meticulously details the segmented cascade sensing channel modeling method,tailored to the unique characteristics of sensing channels.Finally,an in-depth exploration of themulti-modal integration theory

160、 framework is conducted,emphasizing core technologies such asenvironmental modeling methods and environment characterization algorithms,thereby laying arobust groundwork for the development of a comprehensive sensing fusion theoretical system.3.1 Technical Performance RequirementsA comprehensive per

161、formance characterization system is essential for systematicallyevaluating the performance boundaries and theoretical limits of communication system,sensingsystem,and the ISAC system.Given the dual-performance requirements of ISAC system,thissectionsystematicallyelaboratesonperformanceevaluationfram

162、eworksencompassingconventionalcommunication,sensingandISACperformancerequirementsfrommulti-dimensional perspectives.Among them,traditional communication performance indexesmainly cover two dimensions:system effectiveness(such as spectral efficiency and channelcapacity)and reliability(such as bit err

163、or rate and bit error ratio);sensing performance indicatorsinclude measurement accuracy(such as mean squared error),resolution,target detectionprobability and false alarm probability,sensing capacity;for the ISAC system,a comprehensiveevaluation indicator represented by equivalent minimum square err

164、or is proposed to realize unifiedquantitative evaluation of communication and sensing performance.3.1.1 Communication PerformanceThe function of communication system is to realize information transmission,and its22/124advantages and disadvantages are generally evaluated from two aspects:effectivenes

165、s andreliability.1)ValidityEffectiveness quantifies the information transmission efficiency of a communication system.Various performance requirements characterizing the systems effectiveness include channelcapacity,transmission rate,frequency band utilization,energy efficiency,and latency,amongothe

166、rs.Channel capacity denotes the maximum error-free information transmission rate achievableunder specified channel conditions.Transmission rate is defined as the quantity of characters,symbols,or bits transmitted per unit time.Spectral efficiency refers to the data transmission rateper unit frequenc

167、y band.Energy efficiency refers to the maximum number of bits that can betransmitted per unit energy consumed.Latency represents the time required for data transmissionfrom the transmitter to the receiver.Among these indicators,channel capacity serves as the mostcommonlyused communicationrepresentat

168、ion,settingthefundamentallimitationofacommunication system.The following sections respectively elaborate on channel capacitycharacteristics in both time-invariant and time-varying channel conditions.For single-user time-invariant channels,the Claude Shannon channel capacity is defined asthe maximum

169、information quantity of mutual information I(;),between channel inputs andoutputs typically expressed in C=maxp(x)I(;)bit/s or bit/symbol.The applications definedin the above capacity can be extended to multi-user time-invariant channel scenarios,includingmultiple access channels,interference channe

170、ls,broadcast channels,and relay channels22.Wireless fading time-varying channels can be classified into two categories:fast fading and slowfading.In fast fading channels,the channel capacity is characterized by ergodic capacity-thelong-term average of the maximum achievable transmission rate under v

171、arying channel conditions.The ergodic capacity can be expressed as the expectation of the capacity function of the channelstate Cerg=HC(),where C()represents the instantaneous capacity under a specific channelstate and represents the expectation of the channel state.In slow fading channels,thechanne

172、l may remain in a disadvantageous state for an extended period of time,causing theinstantaneous capacity to fall below the required transmission rate,leading to communicationoutages.In this case,the channel capacity is defined as the outage capacity,i.e.,the maximum23/124transmission rate that can b

173、e achieved by the channel under a given outage probability23.Outage capacity addresses the scenarios of communication outage and is utilized to characterizethe continuity and stability of communications in a fluctuating channel environment.2)ReliabilityReliability means the accuracy of information t

174、ransmitted by a communication system.Mainperformance requirements that describe system reliability include signal-to-interference-noiseratio,bit error ratio,and network coverage.The signal-to-interference-noise ratio refers to the ratiobetween signal power and noise power,serving as an important ind

175、icator for evaluating channelquality.A higher signal-to-interference-noise ratio signifies improved reliability of informationtransmission.Bit error ratio/bit error rate represents the ratio of the number of erroneous bits orsymbols during transmission to the total number of bits or symbols.A lower

176、bit error ratio/bit errorrate indicates a higher reliability of the information transmission.Network coverage is defined asthe proportion of an area where UE can receive signal power above a specified minimumthreshold in a zone.Higher network coverage reflects higher reliability of informationtransm

177、ission within that zone.3.1.2 Sensing Performance1)Mean Squared Error andAssociated Lower BoundsThe sensing system is required to complete target localization by estimating basic parameterssuch as angle,distance,and Doppler frequency shift.The main indicators for assessing theaccuracy of parameter e

178、stimation are the mean squared error(MSE),its associated lower bound,and the Equivalent Fishers Information Matrix(EFIM).MSE criterion:The MSE is a commonly used indicator for evaluating the performance ofan estimator,indicated as MSE=?2.Where,is the real parameter vector and?is theestimated paramet

179、er vector.The optimal estimator is typically determined by minimizing the valueof MSE.However,estimations derived through this approach are typically complex to construct,and the characterization of the minimum value of MSE is equally challenging.To facilitateanalysis,relevant lower bounds of MSE ar

180、e employed as indicators for estimation,including theCramr-Rao Bound(CRB)and Bayesian Lower Bounds.The Cramr-Rao lower bound is the24/124most commonly used indicator,providing the theoretical minimum variance for an unbiasedestimator,which is defined as CRB=I1.Where,I is the Fisher Information,which

181、measures how much information quantity of parameter is contained in the observation data Xand is defined as I =2lnf;2.Where,f;is the probability density function of theobservation data X24.The Cramr-Rao lower bound can be extended to the case where theparameters are random variables and the prior di

182、stribution is known,that is,the posterior CRB,which provides a lower limit on variance for the posterior estimator25.Due to the absence ofglobal information in the log-likelihood function of the Cramr-Rao Bound,it predominantlyaddresses local error,achieving a higher accuracy under high signal-to-in

183、terference-noise ratioconditions while performing less effectively under low signal-to-interference-noise ratioconditions.To enhance accuracy,Bayesian lower bounds are introduced for scenarios in whichparameters are random variables with a known prior distribution.The two most representativebounds i

184、n the Bayesian lower bounds are the Weiss-Weinstein Bound(WWB)and the Ziv-ZakaiBound(ZZB).WWB extends the CRB framework by relaxing regularization conditions andintroducingfreeparameters,therebyexpandingitsapplicabilitytoglobalsignal-to-interference-noise ratio while accommodating both biased and un

185、biased estimation26.Derived from the quadratic form of the MSE matrix,ZZB similarly extends the applicablesignal-to-interference-noise ratio range for CRB27.While these two lower bounds demonstratesuperior accuracy,their computational complexity limits practical application.Consequently,CRBremains t

186、he preferred choice except for specific conditions or scenarios demanding extremeaccuracy.Equivalent Fisher information matrix:EFIM is a simplified form of Fisher informationmatrix.FIM is an important indicator to describe the accuracy of parameter estimation,which isdefined as I =logp;logp;T.where,

187、is the parameter estimation vector andp;is the likelihood function of observed data x.EFIM only focuses on the estimationaccuracy of specific parameters or sub-parameters,which is obtained through subspace projectionof FIM onto parameters of interest,thereby simplifying the calculation process.2)Res

188、olutionRadar resolution stands as an important indicator in radar systems for evaluating target25/124recognition performance.Enhanced resolution directly correlates with improved discriminationaccuracy between close targets in radar detection.One of the most commonly used theoreticalestimation model

189、s for radar resolution is R=c2B,where,c represents the speed of light and Brepresents the bandwidth.This model is used to characterize the radars range resolution,which isthe minimum distance between two close targets that the radar can distinguish.In addition,theresolutions for speed measurement an

190、d angle measurement are:=2T 0.886DWhere,represents the wavelength of electromagnetic waves,T represents the radar signalinterval,and Drepresents the antenna panel aperture.Another commonly used estimation model is the ambiguity function,defined as A,=+s?s ej2d,where,is the time delay and is the Dopp

191、ler shift.This modelis used to characterize the correlation of signals under different delays and frequency offsets.Onthis basis,the normalized cross-ambiguity function28 and multi-dimensional ambiguityfunction29 have been extended to introduce more parameters to characterize the tradeoffbetween dis

192、tance,angle,and Doppler resolution.3)Probability of detection and false alarmTwo important performance indicators for target detection are the probability of detection andthe probability of false alarm.Detection probability refers to the probability of successfullydetecting a target when it exists.F

193、alse alarm probability refers to the probability of incorrectlyjudging that a target exists when it does not exist30.Increasing the probability of detection oftenleads to an increase in the probability of false alarms,so a trade-off between the two needs to beconsidered when designing systems.4)Sens

194、ing capacity indicatorThe above sensing indicators,such as sensing resolution and sensing accuracy,more describethe sensing ability of a single point.Sensing indicators for low-altitude networks should reflect thesystem performance of networked sensing,and their QoS is crucial.Therefore,a systematic

195、indicator system for sensing capacity is proposed,which is defined as the maximum number of26/124targets that can be sensed per square kilometer under the sensing QoS.Sensing QoS can include avariety of requirements,including sensing delay,positioning accuracy,speed measurementaccuracy,false detecti

196、on rate,and false alarm rate.Sensing QoS can be dynamically defined fordifferent scenarios and services.For example,for illegal flight intrusion monitoring,the sensingservice quality can focus on false detection rate,false alarm rate,etc.;for UAV logistics scenarios,the sensing service quality can p

197、ay more attention to sensing time delay,positioning accuracy,speed measurement accuracy,etc.Interference from communication will be included in the evaluation method of sensingcapacity to reflect the comprehensive capability of ISAC system.We provide an evaluationmethod for the sensing capacity indi

198、cator,which can be used to obtain the KPI requirements of thesensing capacity:Step 1:Scatter points on the sensed target in system simulation,with a total number ofscattered points being N,and configure simulation parameters such as air interface delay.Step 2:Generate the transmission signal and obt

199、ain the received signal,performsensing processing and calculate the sensing positioning error r corresponding to each target,as well as the proportion of targets that meet the sensing positioning accuracy requirements.Step 3:If necessary,change the total number of scattered points N and repeat steps

200、 1-2 until at least 90%of the target total number of scattered points meets the sensing qualityrequirements.Step 4:Based on the total number of scattered points N,calculate the sensingcapacity/sensing density C=N/A.Area A is the total coverage area of TRxP used insimulation.Sensing capacity will ser

201、ve as an important indicator of digital low-altitude networks andprovide guidance for technical research,system design and standard formulation of ISAC.3.1.3 ISAC PerformanceTo characterize the overall performance of ISAC,a capacity-distortion performance indicator27/124is proposed to quantify the t

202、rade-off between communication capacity and sensing distortion.Thefollowing are three ways to express capacity-distortion performance indicator:Equivalent communication capacity:Based on the idea of converting sensing indicators intocommunication indicators,the concept of sensing-estimated informati

203、on rate is proposed.Thismethod establishes a definition of approximate mutual information between the observed value Yand the ground-truth parameter value.Assuming that follows a Gaussian distribution withvariance P,the estimated value is?,and the mean squared error is D,then the followinginequality

204、 holds:I;I;?12logPD.Therefore,the lower bound of parameterestimation is transformed from the mean squared error to the sensing estimation informationrate31.The trade-off between communication information rate and estimation information ratecan be considered under the integrated sensing and communica

205、tion system.Notably,this methodassumes Gaussian-distributed sensing parameters with known mean squared error of estimator,imposing practical constraints in real-world implementations.Equivalent MSE:Based on the idea of converting communication indicators into sensingindicators,the concept of equival

206、ent MSE is proposed,that is,the communication information rateis equivalent to the MSE indicator.Assume a Gaussian channel Y=+,where X and Zfollow a complex normal distribution with a mean of 0 and a covariance of 1.The minimum meansquare error(MMSE)between input X and output Y is estimated to be:D

207、=11+.Therefore,for a given communication capacity C =log 1+,it can be converted into an MSEindicator by the equation=232.The trade-off between communication equivalentMSE and estimation MSE can be considered under the ISAC system.However,this method isonly applicable to simple linear Gaussian channe

208、l models.Capacity-distortion function:Without converting communication indicators and sensingindicators,a capacity-distortion function C is proposed to represent the trade-off betweencommunication capacity and sensing distortion.Considering a point-to-point ISAC channel,thetransmitter expects to tra

209、nsmit information to the receiver while estimating the channel statethrough echo signals.The capacity-distortion function can be expressed as:C =p maxI;|,and meets the condition of ,?D.Where X and Y represent the input and output symbolsrespectively,?is the estimated sensing state,and ,?is the mean

210、error of the estimator.28/1243.2 Sensing Channel ModelingThe ISAC channel can be modeled as the sum of the target channel and thebackground channel,expressed by the following formula:=+Among them,the target channel refers to all(multi-path)channel components affected bythe target(the so-called chann

211、el components refer to the paths,clusters,etc.,that constitute thechannel),while the background channel refers to all other(multi-path)channelcomponents not included in the target channel.The background channel can be modeled as a random clutter/cluster using methodologiesspecified in existing 3GPP

212、TR 38.901 protocol.The target channel can be effectively modeledthrough piecewise convolution.Segmented convolution modeling is to divide the target channelinto two segments,namely the incident channel from the sending node to the target,and thebackscatter channel from the target to the receiving no

213、de,as shown in Figure 1 below.In theincident channel and the backscatter channel,there are often multiple paths forming severalclusters.In segmented channel modeling,two approaches exist for convolution processing:eithercombining individual paths selected from two channel segments or combining indiv

214、idual clustersfrom two segments.The following takes the cluster-level segmented convolution as an example tomodel a collaborative sensing channel.Figure 1 Schematic Diagram of Segmented Convolution ModelingAssuming that there are P and Q clusters in the incident channel and the backscatter channel,r

215、espectively,then the total number of clusters in the entire target channel is N=PQ,where the thcluster is composed of the pthcluster in the incident channel and the thcluster in the backscatterchannel,with =P 1+.Then,the channel parameters of the thcluster are obtained29/124from the pthcluster in th

216、e incident channel and the thcluster in the backscatter channel in thefollowing manner.Upon acquiring the channel parameters of thethcluster,they can besubstituted into the channel model specified in the 3GPP TR 38.901 protocol to obtain the finalcollaborative sensing channel model.The cluster power

217、 of the target channel is equal to the product of the cluster powers of theincident channel and the backscatter channel:P=PPThe Doppler frequency shift phase of the target channel is the convolution of the Dopplerfrequency shift phases of the incident channel and the backscatter channel:exp 2d()0=ex

218、p 2d()0exp 2d()0The departure angle of the target channel is equal to the departure angle of the incidentchannel:,ZOD=,ZOD,AOD=,AODThe arrival angle of the target channel is equal to the arrival angle of the backscatter channel:,ZOA=,ZOA,AOA=,AOAThe time delay of the target channel is equal to the s

219、um of time delays of incident channeland backscatter channel:=+If the scattering matrix of the target can be stripped from the cross-polarization ratio matrix,then the cross-polarization ratio matrix for the thcluster can be expressed as:exp(),1exp(),1exp()exp()=exp()1exp()q1exp()exp()exp()1exp()1ex

220、p()exp()30/1243.3 Multi-modal Integration Theory3.3.1 Low-altitude Sensor and Environmental Modeling TechnologyThe two most important types of sensing in the process of UAV interaction with theenvironment are recognition of targets of interest and modeling of static background environment.Target rec

221、ognition based on visual sensors extracts targets from visual information of key framesextracted from a single frame or continuous video.In recent years,with the popularization of deeplearning technology,target recognition and tracking technologies have become increasinglymature.It is different from

222、 recognition of targets of interest.In the modeling of a static backgroundenvironment,it is often necessary to find interrelated environmental information in an imagesequence.Feature points in the vision camera data are generally used when searching forassociated information.These feature points are

223、 usually places in the image where the grayscalechanges drastically,such as corner points and edge points.These algorithms can extract featuresunder different illumination,scale and rotation conditions.Lidar completes sparse scanning of theenvironment in a short time(usually 0.1 s)through ranging,ex

224、tracts features based on thegeometric characteristics formed by adjacent points,and then models the entire environment bymatching features within adjacent scanning time periods,as shown in Figure 2.In addition tousing feature point extraction methods to match visual information,end-to-end deep learn

225、ingtechnology will also be used to directly complete the modeling of environmental information.Figure 2 Lidar Environment Modeling Effect31/1243.3.2 Environment Characterization AlgorithmKalman filtering is a powerful optimal estimation theory and algorithm widely used in manyfields such as signal p

226、rocessing,control theory,and navigation system.The core idea is tocombine the prior knowledge(state equation)and observation data(observation equation)of thesystem,and obtain the optimal estimation of the state of the system through continuous predictionand update.It predicts the state at the next m

227、oment based on the system model in the predictionstage,and corrects the prediction results based on new observation data in the update stage toobtain a more accurate state estimate.Under the condition that the linear system and Gaussiannoise assumptions are met,Kalman filtering can provide a minimum

228、 mean square error estimateof the system state.At the same time,it is also a recursive algorithm.It can be calculated only bystate estimation at the previous moment and current observation data,without storing a largeamount of historical data.It has high calculation efficiency and is suitable for re

229、al-time processingsystems.Graph-based optimization is a technology widely used in many fields such as robotics,computer vision,and machine learning.It uses a graph as a data structure to represent the problem.The node in the graph usually represents the state,variable,or entity of the system,while t

230、he edgerepresents the constraint relationship or cost function between nodes.Data for constructing thesenodes and edges may originate from sensors(e.g.,cameras,lidar,inertial measurement units)oranother prior knowledge.The goal of graph-based optimization is to find the state values of a setof nodes

231、(for example,the optimal estimate of all the poses and map features of the aircraft),so asto minimize the total cost of the whole graph.The task of corresponding sensors in modeling theenvironment is to make the whole model smoother and more accurate.Deep learning has achieved remarkable advancement

232、s in the field of environmental modelingand is now widely used in end-to-end methods for equipment such as cameras and lidar.In theprocess of environmental modeling,neural radiance fields(NeRF)and 3D Gaussian splatting(3DGS)technologies have shown powerful capabilities.They can accurately build mode

233、ls ofcomplex environments by learning from large amounts of data.The widespread application ofthese technologies has not only improved the accuracy and efficiency of environmental modeling,but also brought more possibilities to related fields such as autonomous driving and virtual reality.Similarly,

234、in the field of environmental modeling required at low altitudes,these technologies canalso help with high-quality services for UAV.32/124With the development of disciplines,researchers have gradually changed their granularity inobtaining environmental information from rough to fine.Decades ago,remo

235、te sensing satelliteswere able to acquire low-resolution ground object information on a large scale.In recent years,with the advancement of photogrammetry and SLAM technology,environmental models havebecome more and more sophisticated.Zhongyu Liu et al.33 proposed a propagation prediction model base

236、d on three-dimensionalray tracing to obtain environmental information through digital maps,including geometric andelectrical parameters of buildings,terrain,green belts,etc.The ray path classification technologyis used to divide the propagation path into five categories,and different methods are use

237、d todetermine different categories of paths,effectively handling complex situations such as multiplereflections and diffraction,and improving the efficiency of ray tracing.The environmentalinformation is preprocessed,such as ignoring geometric factors that have little influence on theprediction resu

238、lts according to Tx and Rx positions,evaluating the visible relationship betweenvertical walls and simulation points,and determining the direction surface set of vertical walls,soas to improve the calculation efficiency.By establishing and using environmental models,the propagation characteristics o

239、f radiosignals in different environments can be predicted more accurately,thus providing importantassistance and support for network planning,coverage optimization,communication performanceimprovement,etc.For example,it helps determine the best location for BSs,the direction andheight of antennas,op

240、timizes communication resource allocation,improves the reliability andquality of communications,etc.Jakob Thrane et al.34 proposed a signal strength predictionmethod based on deep learning,combining geographic images with expert knowledge.Using openstreet map(OSM)to generate environmental maps,predi

241、cting reference signal received power(RSRP)through deep learning models,and using real data sets for evaluation,the average rootmean squared error(RMSE)is reduced by 53%compared with ray tracing technology.In thecross-scenario evaluation,the optimal RMSE is 6.3 dB.The study found that the image dist

242、anceperformance is better at 250-300 meters,and the RMSE of conventional images is lower thanthat of full-size images.33/1244 Low-altitude ISAC Architecture and Key TechnologiesThis chapter provides a comprehensive overview of the technical system architectureand key technological breakthroughs of l

243、ow-altitude ISAC,focusing on four corechallenges:insufficientlow-altitudethree-dimensionalcoveragecapability,limitedimprovement in sensing accuracy,confidence level of sensing results to be promoted,andcomplex global interference environment.The Section I focuses on the design of overalldigital low-

244、altitude architecture,and deeply analyzes the functional architecture of ISACsubsystem;the Section II elaborates on the key technologies of coverage enhancement inresponse to the problem of insufficient low-altitude three-dimensional coverage capabilities,including key solutions such as dual wavefor

245、m optimization,beam management and nodeswitching;the Section III focuses on the challenges of limited sensing accuracyimprovement and conducts in-depth analysis from three dimensions:beam design,sensingalgorithm,and integrated sensing;Section IV addresses the issue of improving thereliabilityof sens

246、ing results from three aspects:NLoS sensing,clutter suppression methodsand elimination of non-ideal factors;Section V comprehensively introduces the interferencesuppression technology system in response to the complex problem of global interferenceenvironment,focusing on key technologies such as int

247、erference feature analysis,collaborativeinterference coordination andpower control,adaptivebeamnullingoptimization,and reference signal nulling algorithm,supplemented by simulationvalidation results in typical scenarios.34/1244.1 NetworkArchitecture4.1.1 Overall Digital Low-altitudeArchitectureFigur

248、e 3 Digital Low-altitude Basic ArchitectureThe overall digital low-altitude architecture can be composed of supporting infrastructure,network infrastructure,and applications and UE.Among them,the supporting infrastructureincludes operational supervision,data storage and ground control centers.The op

249、erationalsupervision is responsible for the service,supervision and operation of the entire low-altitudenetwork,including flight data management,UE authority management,security and privacymanagement,application expansion,and flight path planning.Data storage mainly refers to thelow-altitude flight

250、related data collected from sensors and other equipment,which assists inmission planning,flight monitoring,and data analysis.The ground control center is mainlyresponsible for the real-time monitoring and management of network access,take-off and landing,flight trajectory,yaw warning and other aspec

251、ts of intelligent UE such as UAV.The network infrastructure includes ISAC networks,satellite networks and edge computingpower networks.The ISAC network relies on ground network infrastructure,such as the existing5G-A public network and the future 6G network,and optimizes the existing public network

252、orbuilds a low-altitude private network to provide communication or sensing services forlow-altitude intelligent connected UE at a certain altitude.Satellite networks can serve as aneffective supplement to the ISAC network.By combining technologies such as Beidou anddifferential positioning,they pro

253、vide three-dimensional high-precision positioning services forlow-altitude flying chess.The edge computing power network can realize low-altitude dataprocessing and optimization,as well as provide intelligent algorithm capabilities such as35/124navigation,sensing,and positioning.The supporting infra

254、structure and network infrastructure jointly provide services forintelligent UE and provide external interfaces for innovative applications.4.1.2 ISACArchitecture4.1.2.1OverallArchitecture of ISACFigure 4 Overall Architecture of ISACIn the 6G ISAC scenario,ISAC is further subdivided into NET for sen

255、sing and sensing forNET.The most typical example of NET for sensing is the combination of ISAC services andlow-altitude economy.According to the report from the IMT-2020 ISAC task group,it is estimatedthat the sensing data transmission rate for scenarios related to flight path management and obstacl

256、eavoidance is about 10 Mbps.The sensing for NET includes assisting mobility management andbeam management based on the channel environment information obtained by sensing.ISACenables 6G networks to not only transmit traditional user data but also generate,process and use alarge amount of intra-netwo

257、rk data.However,since the intra-network data in 5G-A networks ismainly signaling,it is characterized by small data volume,high priority,and real-time data that iseffective in a short period of time.For sensing data within 6G networks,due to its large datavolume,flexible data transmission endpoints,a

258、nd flexible priority,it is necessary to provide aunified and efficient solution from the network architecture level.In order to meet the needs of collecting,transmitting and processing a large number of sensed36/124data,and considering that there are problems such as repeated UE(UE)/network function

259、 data inthe data collection scheme related to the single use case in the 5G-A standard and repeatedprotocol functions for air interface data collection based on the control plane,the data plane can beintroduced into the 6G network architecture to avoid the control plane from carrying a largeamount o

260、f non-signaling data,and ensure that user-plane transmission is terminated inside themobile network.The functional design and parameter configuration of the data plane are intendedto provide a unified and optimized solution,thereby avoiding the fragmentation of individual usecases.In addition,the da

261、ta transmission efficiency is improved by designing and optimizing dataplane protocols.Therefore,the introduction of data plane can not only improve the efficiency ofdata transmission but also realize data reuse,enhance cross-domain data collaboration andfacilitate expansion to meet new demands.As s

262、hown in the figure above,the end-to-end data plane functions include data privacy andsecurity function,data control function,data transmission function,data processing function,anddata storage function.The details are as follows:Data source:For ISAC application scenarios,the data source mainly refer

263、s to the sensingdata collected from BSs and UEs.Sensing function:In ISAC scenarios,the sensing function is a data consumer.The sensingfunction obtains the sensing measurement data required for different sensing requests andresponds to use case requests such as target detection and environmental moni

264、toring.Data privacy and security function:used to support the authentication and authorizationmechanisms for data access.In addition,the privacy and security functions of data need to beconsidered to support privacy and security mechanisms such as authentication,authorization,andaccess control initi

265、ated by data functions.Data control function:It is used to support data collection coordination,data serviceconfiguration(such as network packet size and interval),data transmission configuration(such asestablishing/modifying/releasingdataplanetransmissionchannels)anddataprocessingconfiguration(such

266、 as data preprocessing or data analysis configuration).The core network orradio access network collects and coordinates data according to the data requirements of differentsensing use cases to avoid repeated data collection by BSs or UEs.37/124Data transmission function:It is used to forward and tra

267、nsmit data of data plane according tothe configuration of the data control function.According to the different peers of data transmission,data plane transmission includes UE and RAN,RAN and CN,UE and CN,RAN internalfunctions or CN internal functions and other multiple data transmission protocols.Dat

268、a processing function:ISAC-related data processing includes sensing data analysis,andmulti-modal data fusion.Pre-processing and aggregation of data through the data analysis functioncan reduce the collection of raw data across network nodes.For example,raw data can becollected from data producers ac

269、cording to the needs of data consumers,summarized,sorted outand correlated,and data with specified granularity and meeting time period requirements will beoutput.Through the data fusion function,the BS can fuse the sensing measurement data ofmultiple UEs and the sensing measurement data of the BS to

270、 generate multi-dimensional spectruminformation.Data storage function:Many data in the 5G network are real-time data used for a short timeand will not be stored persistently,such as measurement information reported by UE assistingscheduling.However,large amounts of data may need to be stored and pro

271、cessed in ISAC.Forexample,for sensing needs such as UAV intrusion monitoring,environmental sensing,traffic flowand weather conditions,the BS or UE will measure and generate a large amount of sensing data.In order to support the reuse of sensing measurement data for different sensing use cases and AI

272、model training based on AI sensing processing algorithms,6G needs to provide a storagemechanism for persistent data.4.1.2.2Sensing Function ArchitectureThe main functional entities to be included in the system for sensing service are shown in thefollowing figure,which summarizes the possible interac

273、tion between sensing functions in the formof a bus.The sensing function is a logical function,and there may be situations where multiplefunctions are implemented by one device.38/124Figure 5 Sensing functionThe functional entity for sensing service management is responsible for interacting with thes

274、ensing client to enable it to obtain sensing services.The functional entity for sensing task control controls the execution of sensing tasks toachieve basic sensing functions.The functional entity for sensing task control can determine thesensing method,resources required for sensing,sensing signal

275、transmission/receiving nodes andfunctional nodes for sensing data processing according to the sensing tasks.The target oflow-altitudesensingismainlylow-altitudeUAVs.Appropriatesensingsignaltransmission/receiving nodes can be selected for sensing according to the electronic fence area andUAV track.Th

276、e sensing data processing function processes the sensing data,including processing theoriginal sensing data(including the received data of the sensing signal and the sensing dataobtained by third-party sensing)to obtain intermediate result sensing data,and processing theintermediate result sensing d

277、ata to obtain final result sensing data.Since low-altitude UAVs areusually mobile,they may span the sensing coverage of different sensing signal transmissionreception points,so that the network may obtain multiple sensing data for the same target,andmultiple sensing data need to be fused.For network

278、ed UAVs,their own 3rdparty sensingtechnology can also be used to obtain sensing data to assist the network in sensing.Sensing signal transmission requires the use of wireless sensing resources.The functionalentity for wireless sensing resource scheduling determines the wireless sensing resources use

279、d forsensing signal transmission according to the demand for wireless sensing resources.The following table shows an example of sensing function deployment in the network.39/124Table 3Deployment of Sensing Function in the NetworkSensing FunctionCorenetworkAccessNetworkUESensing service managementfun

280、ctionSensing task control functionSensing data processingfunctionWireless sensing resourcescheduling functionSensing signaltransmission/receptionThird-party sensingfunction4.2 Coverage Enhancement Technology4.2.1 Coverage Range ExpansionIn the mono-static sensing mode of BS,the use of pulse waveform

281、s with time-sharingtransmission and reception can avoid the requirement for full-duplex capability,thereby greatlyreducing the hardware cost of the equipment.When using a pulse waveform,the echo signalreturned to the BS before the transmission of the sensing signal can not be received,making itimpos

282、sible to detect close-distance sensing targets,that is,there are close-distance blind areas.Ifthe pulse width is,all sensing targets with a distance from the BS less than c 2 will be in theclose-distance blind area.However,reducing the pulse width will reduce the power of the sensingsignal,making it

283、 impossible to effectively detect long-distance sensing targets.In order to take into account both long-distance and short-distance sensing coverage,twopulse waveform designs can be used to improve the sensing coverage capability.As shown inFigure 6,waveform I is used for long-distance coverage and

284、waveform II is used forshort-distance coverage.The modulation methods of the two waveforms can be the same or40/124different;for example,waveform I is an LFM(liner frequency modulation)waveform,waveformII is an OFDM(orthogonal frequency division multiplexing)waveform,or both waveforms areLFM wavefor

285、ms.It is assumed that the pulse width and pulse period of waveform I are1pulseand 1period,respectively,and the pulse width and pulse period of waveform II are2pulseand 2period,respectively.Then the close-distance blind area boundary and maximumunambiguous detection range of waveform I are c1pulse2 a

286、nd c1period2,respectively,and theclose-distance blind area boundary and maximum unambiguous detection range of waveform IIare c2pulse2 and c2period2,respectively.It is assumed here that the modulation sequences ofeach pulse in waveform I and waveform II are the same,which is also a design commonly a

287、doptedin radar technology.If individual pulses within the same waveform can be distinguished throughmodulation sequences,the maximum unambiguous distance can be increased.In order to achievecontinuous coverage over a certain distance range,the maximum unambiguous measurementrange of waveform II need

288、s to be less than or equal to the close-distance blind area boundary ofwaveform I,that is,2period 1pulseis required.Meanwhile,in order to maximize the use of thetime domain resources of a pulse waveform,it is necessary to satisfy Mperiod MOFDMTOFDM,=1 or 2,that is,continuous Mpulses need to be time-

289、aligned with continuous MOFDMOFDM symbols,where TOFDMrepresents the OFDM symbol duration.As shown in Figure 6,2consecutive pulses of waveform I are aligned with 1 OFDM symbol,and 8 consecutive pulses ofwaveform II are aligned with 1 OFDM symbol.Figure 6 Schematic Diagram of Double-pulse Waveform Sen

290、sing SignalIn summary,according to the above-mentioned dual waveform design method,by designinga smaller period of waveform II to compensate for the problem of insufficient blind area coverageof waveform I,a sensing signal waveform with small blind area and high coverage capability canbe designed wi

291、th fewer parameters.41/1244.2.2 Continuous Coverage Technology4.2.2.1Beam ManagementIn the field of radar,the beamforming technology used in phased array radar has maturehardware implementation and signal processing solutions.Currently,5G BSs deployed on a largescale have 32 antenna ports,while LTE

292、BSs also have 8 antenna ports.Each antenna port isconnected to multiple antenna arrays,laying a solid physical foundation for multi-antenna sensingbased on beamforming.Through digital or analog beamforming,sensing nodes equipped withmultiple antennas can form high-gain narrow beams,so that most of t

293、he energy of the sensingsignal is concentrated on the sensing area or sensing target.On the one hand,it improves thesignal-to-interference-noise ratio of the reflected signal,and on the other hand,it can also form abetter suppression effect on clutter interference from other directions.Figure 7 Sche

294、matic Diagram of Signal Relationship between Antenna Ports in BeamformingBased on beamforming sensing technology,the signals of each transmitting antenna port arecorrelated and differ from each other only by a phase difference related to the antenna spacing andbeam pointing angle,as shown in Figure

295、7.This solution is simple to implement,but it also hascertain limitations.On the one hand,when the array is beamformed,the angle sensing accuracy ofthe system,that is,the angle estimation resolution of the system array,is related to the beam width.When the angle difference between two sensing target

296、s is less than one beam width,the sensingbased on beamforming cannot distinguish these two targets in the angle domain.At this time,it isnecessary to distinguish targets in other domains(such as delay domain and Doppler domain);onthe other hand,when the sensing area is large or there are multiple ar

297、eas to be sensed in anenvironment,multi-antenna sensing based on beamforming may need to be completed by means42/124of beam scanning.Beam scanning takes longer than a single sensing,which can easily lead toreduced sensing performance in time-varying environments(e.g.,high-speed UAVs).Although wecan

298、use a wider beam to cover the sensing area or sensing target,it sacrifices sensing accuracy orsensing signal-to-interference-noise ratio to a certain extent when the total transmitted power isthe same.Figure 8 is a schematic diagram of the above two situations assuming that the BSsensing mode is mon

299、o-static,and the sensing target is UAV in a certain area.For multi-antennasensing based on beamforming,the configuration of sensing beams(such as beam width,numberof beams and beam scanning time)may need to be determined based on some prior informationabout sensing targets/sensing areas,such as appr

300、oximate distribution range of sensing targets,density of sensing targets,and approximate size/azimuth of sensing areas.Figure 8 Schematic Diagram of Beamforming-Based Multi-antenna Sensing(a)The sensing target spacing is less than the beam width,resulting in failure to distinguish 2targets(b)The sen

301、sing area is larger than the beam width and needs to be combined with beamscanningIt should be noted that for ISAC scenarios,since the sensing target and the communicationtarget are not necessarily the same target,the sensing beam and the communication beam may notbe the same beam.For the communicat

302、ion function,the communication beam needs to be alignedwith the communication receiver to obtain a reliable and stable communication link;for thesensing function,the sensing beam needs to be configured according to the specific location of thesensing area or sensing target,as shown in Figure 8.Howev

303、er,in the early stages of sensing,theprecise location of the sensing area or sensing target is often unknown.To address this problem,43/124one strategy that can be adopted is for the ISAC system to first use wide beams for coarse-grainedsensing and then use narrow beams for fine-grained sensing afte

304、r determining the approximatelocation of the sensing area or target.The other strategy that can be adopted is for the ISACsystem to perform sensing beam scanning and sensing beam measurement processes anddetermine the sensing beam based on the sensing beam measurement results.ISAC systems canuse mul

305、tiple beams,some of which are used for communication and some for sensing,or both.In terms of beam management process,the communication and sensing beam managementprocesses may be performed independently.For the sensing mode of mono-static by BSs or UE,sensing beam management can be directly impleme

306、nted based on its own algorithm withoutinteracting with the other end of the communication;for the sensing mode of A-transmitting andB-receiving by BSs or UE,the BS and UE determine the best transmission/reception beamthrough communication beam scanning;at the same time,through sensing beam scanning

307、,the BSor UE determines the optimal transmission/reception beam of the BS and UE based on themeasured values of sensing performance evaluation indicators such as sensing signal interferencenoise ratio(defined as the ratio of the power of the reflected signal of the sensing target to the sumof the cl

308、utter and noise power).In addition,communication beam management and sensing beammanagement can also be a joint processing flow,that is,through the same set of beamconfigurations,in the beam scanning step,the BS or UE simultaneously obtains thecommunication measurement value,the sensing measurement

309、value,or the measurement value ofthe joint evaluation metrics of ISAC through beam measurement.Based on the abovemeasurements,the optimal communication beam and the optimal sensing beam are determined.4.2.2.2Sensing Node SwitchingSensing nodes in the mobile communication networks,such as BSs and UEs

310、,can performsensing/ISAC application through two different sensing modes:sending or receiving sensingsignals between nodes,or mono-static sensing signals by nodes.Changes in the state of thesensing target(including spatial position,spatial orientation,and movement speed)or the sensingenvironment,as

311、well as changes in the sensing node(such as UE movement),may lead to changesin sensing performance.In order to ensure the application continuity of sensing/ISAC andcontinuously guarantee the QoS of sensing/ISAC,the network may need multi-node collaborative44/124sensing.However,multi-node collaborati

312、on requires multiple BSs or UEs to serve the samesensing target,resulting in large physical hardware and time-frequency resource overhead.Therefore,another alternative is for the network to switch sensing nodes.Considering thedifference in sensing capabilities of different sensor nodes,uneven distri

313、bution of sensing nodes,and uncertainty of the sensing environment,the switching process of sensing nodes may also beaccompanied by switching of different sensing modes:The difference in sensing capabilities is reflected in the capability differences betweendifferent BSs and between BSs and UE.For e

314、xample,the difference in antenna array aperturebetween different BSs,or the difference in the number of antenna ports between a BS and a UE.The uneven distribution of sensing nodes mainly refers to the strong randomness andvariability in the distribution of UEs.The uncertainty of the sensing environ

315、ment refers to the differences in the environmentnear different sensing nodes.For example,different buildings near different sensing nodes havedifferent degrees of radio wave propagation phenomena such as occlusion and multipath.Figure 9 Schematic Diagram of Sensing Node Switching Scenario in A-Tran

316、smitting andB-Receiving Sensing Mode45/124The following briefly discusses several typical sensing node switching scenarios.Figures 9(a)and 9(b)show the scenarios where a BS acts as a sensing node and is switched,including theswitching of either the transmitter or the receiver in A-transmitting and B

317、-receiving mode,and thesimultaneous switching of both.Figure 9(c)-(e)shows the scenario where at least one of the UEor BS is switched when the UE participates in A-transmitting and B-receiving sensing.Figure 10Schematic Diagram of Sensing Node Switching Scenario in Mono-static SensingModeFor the mon

318、o-static sensing mode of sensing nodes,in order to ensure the continuity ofsensing/ISAC services,sensing node switching may also be required.Figures 10(a)and(b)respectively show the scenarios of mono-static,and transmission switching by BS and UE.The above sensing node switching scenarios all use UA

319、Vs as sensing targets for illustration.The UAV has a large flight range,generally ranging from several hundred meters to kilometers.Therefore,when the target UAV moves from BS 1 to BS 2,the quality of the sensing signalreflected by the target UAV may be significantly reduced due to the increase in t

320、he distancebetween BS 1 and the target UAV or the change in the radar cross section(RCS)of the target UAV,which will cause the reduced performance of the sensing/ISAC application,and even interruptedapplication.Base station 1 or the sensing network element can schedule BS 2 to perform sensingas a sw

321、itched sensing node based on the information of other BSs in the area.46/124Figure 11Schematic Diagram of Sensing Mode SwitchThe switching of sensing nodes may also be accompanied by the switching of sensing modes.Still taking the positioning/trajectory tracking of flying UAV as an example,Figure 11

322、 shows aschematic diagram of sensing mode switching.It is assumed that before the switching,the BSrealizes sensing of the target UAV by mono-static sensing signals.When the target UAV graduallymoves away from the BS,due to the increase of sensing signal echo path or the change of targetUAV RCS,or th

323、e significant decrease of target UAV RCS change,the application performance ofsensing/ISAC will be reduced or even the application will be interrupted.At this time,the BS orsensing network element can dispatch UE and BSs near the target UAV area based on the otherBSs and UE information in the area t

324、hat it has mastered,and continuously track the target UAV bysending sensing signals from the BS and receiving them from the UE(or the UE sends sensingsignals and the BS receives them).Notably,in this scenario,the BS participating in sensing afterswitching and the source BS may be different BSs.In so

325、me cases,the sensing mode in which theBS sends a sensing signal and the UE receives it(or the UE sends a sensing signal and the BSreceives it)can also be switched to a sensing mode in which the BS(or UE)spontaneouslytransmits and receives sensing signals.4.3 SensingAccuracy Improvement Technology4.3

326、.1 Waveform DesignISAC waveform:Waveform design is an important foundation of the ISAC system.Forcomplex application scenarios with small targets and slow movement in digital low-altitudenetworks,the demand for optimization of waveform design with high sensing accuracy is more47/124prominent.A lot o

327、f research has been devoted to the optimization and design of waveforms inISAC systems,which can be roughly divided into radar-centric,communication-centric and jointdesign.Radar-centric waveform design:The sensing waveforms commonly used in radar systemsinclude pulse waveform(PW),continuous wavefor

328、m(CW),frequency modulated continuouswaveform(FMCW)and phase coded waveform(PCW).The pulse waveform has a high peakpower and enables the detection of long-distance targets.The time delay resolution of the pulsewaveform is related to the pulse width.Combined with pulse compression,a higher time delayr

329、esolution can be achieved and the distance of the target can be accurately measured.Thedisadvantages of pulse radar are that there is a blind area for close-range sensing,and the pulserepetition frequency(PRF)limits the ability to measure speed.For fast-moving targets,Dopplerambiguity may occur.The

330、continuous waveform can provide continuous speed and distance measurement,and hashigh detection sensitivity for low-speed targets.The disadvantage is that the distance cannot bemeasured directly,which usually needs to be used in combination with modulation technology.Liner frequency modulation(LFM)w

331、aveforms are commonly used for FMCW radar.Three typesof waveforms are commonly used:linear sawtooth frequency modulation,linear triangular wavefrequency modulation and segmented linear frequency modulation,as shown in Figure 12.Thefrequency modulated continuous wave has high waveform spectrum effici

332、ency and can achievehigher distance resolution within a smaller bandwidth.However,for fast-moving targets,there isdistance and Doppler coupling,which will lead to distance estimation errors.(a)Linear sawtooth frequency modulation(b)Linear triangular wave frequency modulation48/124(c)Segmented linear

333、 frequency modulationFigure 12Three Commonly Used LFM WaveformsThe phase coded waveform has a strong anti-interference ability,can improve thesignal-to-interference-noise ratio,is suitable for multi-target detection,and can detect multipletargets at the same frequency at the same time.The phase-encoded signal is sensitive to Doppler,and the Doppler frequency shift in the echo signal will seriously