《Next G聯盟：2024數據中心與下一代核心網絡可持續性指標演進白皮書（英文版）（30頁）.pdf》由會員分享，可在線閱讀，更多相關《Next G聯盟：2024數據中心與下一代核心網絡可持續性指標演進白皮書（英文版）（30頁）.pdf（30頁珍藏版）》請在三個皮匠報告上搜索。

1、1A L L I A N C EAn ATIS InitiativeNext G Alliance Report:Evolution of Sustainability Indicators for Data Centers and Next Generation Core Networks 2As a leading technology and solutions development organization,the Alliance for Telecommunications Industry Solutions(ATIS)brings together the top globa

2、l Information Communications Technology(ICT)companies to advance the industrys business priorities.ATIS 150 member companies are currently working to address network reliability,5G,robocall mitigation,smart cities,artificial intelligence(AI)-enabled networks,distributed ledger/blockchain technology,

3、cybersecurity,IoT,emergency services,quality of service,billing support,operations and much more.These priorities follow a fast-track development lifecycle from design and innovation through standards,specifications,requirements,business use cases,software toolkits,open-source solutions,and interope

4、rability testing.ATIS is accredited by the American National Standards Institute(ANSI).ATIS is the North American Organizational Partner for the 3rd Generation Partnership Project(3GPP),a member of the International Telecommunication Union(ITU),as well as a member of the Inter-American Telecommunica

5、tion Commission(CITEL).For more information,visit www.atis.org.Follow ATIS on Twitter and on LinkedIn.The ATIS Next G Alliance is an initiative to advance North American wireless technology leadership over the next decade through private-sector-led efforts.With a strong emphasis on technology commer

6、cialization,the work will encompass the full lifecycle of research and development,manufacturing,standardization,and market readiness.FOREWORD3With the growing demand for 5G and the advent of 6G,the ICT Industry has an opportunity to achieve net-zero operations in the foreseeable future by implement

7、ing a comprehensive lifecy-cle approach.This includes the design,build,operations and maintenance and ultimately the decommissioning of mobile network resources.Core networks and other mobile Network Func-tions(NFs)deployed in data centers are replacing traditional dedicated hardware implementa-tion

8、s.This paper provides multiple perspectives on how observability,the use of circular econ-omy principals,and choice pave the way for organizations to embark on their sustainability journey,ultimately reaching net-zero operations.The increasing demand for digitization processes and functions has led

9、to a rise in energy re-quirements to power these applications.Considering sustainability,software applications must be optimized to efficiently use IT assets,which reside in data centers of various kinds.These data centers can be cloud and on-premise,including Central Offices,assets.Regardless of th

10、e data center asset class,it requires careful consideration of the design,building,and operation of the physical infrastructure supporting these data centers.The IT assets and the infrastruc-ture itself consume energy and other resources that need continuous improvement.The evolu-tion of sustainabil

11、ity indicators for data centers and next generation core networks becomes more relevant.It includes a comprehensive look at existing and new Key Performance Indica-tors(KPI)allowing data center and core network operators to measure and improve their cur-rent and future use of precious resources.Trad

12、itional KPIs,such as Power Utilization Effectiveness(PUE)measure a data centers overall energy consumption relative to the energy used for IT assets.While fundamentally effective,this simple metric fails to consider other important aspects beyond energy usage.Greenhouse Gas(GHG)emissions associated

13、with the production of components and buildings,water con-servation measures,waste reduction and elimination,and the overall impact on the local eco-system all need to be considered in the network operators sustainability journey.A fresh look on procuring energy,equipment,and materials from sustaina

14、ble and ethical sources is becom-ing essential.Sustainability has reached different levels of maturity throughout the ICT industry.This paper not only outlines the importance,but also serves as a guide to further educate,evaluate,and implement a holistic sustainability approach.Finally,practical rec

15、ommendations such as the development of an organizations sustainability strategy and identifying and implementing rel-evant and future KPIs are made.EXECUTIVE SUMMARY4.Foreword 2Executive Summary 31 Introduction 52 Enhanced Approach to Environmental Sustainability Indicators for 6G 6 2.1 Observabili

16、ty 6 2.2 Choice 6 2.3 Circular Economy 63 6G Applications Running on IT Assets 7 3.1 Workload Requirements Including Energy Consumption 7 3.2 Existing KPIs(Compute/Energy)8 3.3 Current Limitations and Relevance to the Data Center KPIs 8 4 IT Assets-Compute,Network,Storage 10 4.1 Guidelines for a Sus

17、tainable Compute Environment 10 4.2 Example of Scope 3 Evaluation 115 Data Centers 13 5.1 Key Trends in Data Centers 13 5.2 Types of Data Centers 13 5.2.1 Hyperscale Data Centers 13 5.2.2 Central Office and On-Premise 14 5.2.3 Edge Deployments 14 5.2.4 Application Server 15 5.3 Data Center Architect

18、ure,Redundancy,15 Availability and Impact on Sustainability 5.4 Current Data Center Sustainability KPIs and their Impact 16 5.4.1 Sustainability Attributes for Data Centers 16 5.4.2 Energy Efficiency 16 5.2.2.1 Power Usage Effectiveness(PUE)16 5.4.2.2 Renewable Energy Factor(REF)17 5.4.2.3 Energy Re

19、use Factor(ERF)17 5.4.2.4 Water Usage Effectiveness(WUE)17 5.4.3 Additional KPIs 196 Summary and Recommendations 207 Abbreviations 218 References 23Acknowledgments 25Next G Alliance Reports 28Copyright and Disclaimer 29TABLE OF CONTENTS5Addressing climate change is the biggest challenge faced by hum

20、anity and to avoid the worst environmental impacts of climate change,we need to limit long term global warming below 1.5C from pre-industrial levels.However,according to the United Nation(UN)humanity is currently not on track to achieve this objective 1.The Information Communication Technology(ICT)e

21、cosystem has established a collaborative framework aimed at achieving net zero emissions by 2050 at the latest.This pathway to net zero spans the entire ICT value chain,addressing emissions under direct operational control(Scope 1),indirect emissions from purchased electricity(Scope 2),and emissions

22、 associated with products during manufacturing and transportation(Scope 3).This whitepaper provides an overview of key data center sustainability strategies that are critical for achieving Net Zero emissions across the entire ICT value chain.Recent Global System for Mobile Communications Association

23、(GSMA)report indicates that the core network which handles centralized network functions and the data center on which it can operate consumes around 19%of total energy in the operation of mobile networks 2.As more core network services have moved into the cloud in 5G,data center power consumption ha

24、s increased.Additionally,according to the International Energy Agency electricity consumption from data centers,AI and the cryptocurrency sector could double by 2026 compared to 2022,reaching more than 1,000 TWh in 2026 3.Making sustainable operations of data centers critical for the ICT sectors goa

25、l to achieve Net Zero emissions.Focusing on the data center domain and covering cloud enterprise and edge data centers of all sizes,this whitepaper offers recommendation for benchmarking the carbon footprint and other key environmental sustainability Key Performance Indicators(KPIs).1INTRODUCTION6Th

26、e existing 5G network KPIs primarily emphasize observability to gauge performance and results with very coarse granularity.However,the advent of 6G necessitates a comprehensive paradigm shift.As Figure 1 illustrates,it encompasses three crucial dimensions in terms of implementing environmental susta

27、inability KPI metrics:Observability,Choice,and Circular Economy(OCC)4.2.1 ObservabilityObservability in the data center and core network enables near real-time recording and reporting of KPI metrics,which are crucial throughout the technological lifecycle.This is essential for assessing the environm

28、ental impact of 6G technologies,both during the design and operation of data center technologies.Near real-time reporting mechanisms in data centers allow for monitoring of the entire infrastructure,checking application performance and energy consumption,and increasing efficiency in future capacity

29、planning by analyzing resource utilization trends.Similarly,observability in the core network aids in traffic analysis and network optimization.Overall,observability contributes to advancing the entire ecosystem towards a more environmentally sustainable future.2.2 ChoiceThe concept of choice empowe

30、rs data center and core network operators,product developers,users as well as network architects,to navigate a landscape where sustainability goals align seamlessly with performance objectives.Near real-time dashboards provide updates about the environmental impact of various service configurations,

31、offering choices to make informed decisions based on sustainability preferences.This fosters collaboration in decision-making regarding architectural design,resource allocation,and operational practices.Below is an example of the KPI that helps to make those informed decisions.Example KPI:Sustainabi

32、lity Impact Index Definition:A composite index that dynamically quantifies the environmental impact of data center and core network operations,including energy consumption,carbon emissions,and resource utilization,allowing network architects to make informed decisions based on data gathered over a c

33、ertain period of time(e.g.,monthly).Example KPI:Green Service Performance Index Definition:A composite index that factors in both traditional performance metrics(e.g.,data speed,latency)and sustainability metrics(e.g.,energy efficiency,PUE,waste factor 5 to provide a holistic view of service quality

34、 in data center and core network operations.2.3 Circular EconomyCircular Economy is the departure from linear economic models as it is restorative and regenerative in nature and the aim is to keep components and materials at their highest utility and value at all times while reducing the waste strea

35、ms.The data center and the rest of the network infrastructure consist of equipment such as servers,storage units,networking devices,and cooling systems.As businesses and services transition toward higher-performing and more energy-efficient equipment,extending the lifetime of data center equipment b

36、ecomes crucial.Circular economy principles advocate for considering lifecycle factors,including durability,upgradability,repairability,reusability,refurbishability,and recyclability or recoverability at end-of-life,even during the design phase.Its essential to ensure that the production process itse

37、lf is sustainable,minimizing environmental impact through the use of renewable materials and energy sources.Figure 1:6G Enhance Environmental Sustainability KPIs2ENHANCED APPROACH TO ENVIRONMENTAL SUSTAINABILITY INDICATORS FOR 6GObservabilityCircularEconomyChoiceAllows to track and report environmen

38、tal indicators and enables choice by informing high levels of steerability and versatile actionsEmpowers the entire value chain,from standards and product developers to operators and users,to exercise preferences that impact sustainabilityEnsures near-zero waste through high degrees of reuse and rec

39、ycling of renewable or recovered components,raw and rare materials,and products73.1 Workload Requirements Including Energy Consumption6G operator networks,providing services including connectivity between a UE and applications residing in a data network(DN)are expected to continue the evolution star

40、ted in 5G from requiring network functions realized on bespoke,dedicated hardware with specialized platform resources to utilizing virtualized cloud native network functions that may be instantiated and managed as workloads running on data center resources.5G presented the first stage of this evolut

41、ion with the introduction of a service-based architecture and the explicit specification that 5GC user plane and control plane network functions(i.e.,AMF,SMF,PCF,NWDAF,NSSF,NEF,NRF,UPF,etc.)may be virtualized and instantiated on an appropriate platform,e.g.,on cloud infrastructure 6 rather than impl

42、emented as software running on a dedicated hardware.Supported virtualized deployments include:Network Function instances comprised of one or more services with constituent service operations,that may be deployed as distributed,redundant,stateless,and scalable network function instances.Services may

43、be provided from multiple locations with several execution instances in each location.Network Function instances deployed such that several network function instances are present within a network function set that have access to common,shared state information.Network functions instances within a ne

44、twork function set can be selected interchangeably to enable distributed,redundant,stateless and scalable functionality 6.Increasingly,a disaggregated virtualized Radio Access Network(RAN),especially centralized control plane aspects which do not require specialized processing capabilities(e.g.,the

45、5G CU-CP)will follow the core network in becoming data center application workloads that run on IT assets.The motivation for this is the well understood benefits of virtualization,which includes reliability,scalability,life-cycle management,interoperability,cost and the ability to dynamically instan

46、tiate functions according to user requirements and traffic demand.The evolution from dedicated hardware to edge and centralized data center hosted compute resources for 5G and 6G functions greatly facilitates selection of network function instances according to user requirements.Those requirements m

47、ay be satisfied in a number of ways.For example,a network slice or private network that comprises dedicated Virtualized Network Function(VNF)instances may be used.Slice or network specific KPIs associated with a standardized or operator specific Service Type characterize both performance and sustain

48、ability,where the Service Types may include those defined by 3GPP for network slices:Table 1:Standardized SST Values 6For sustainability,for example,a network slice to support MIoT may be comprised of network functions instantiated in a centralized data center where a prevailing efficiency metric su

49、ch as the energy consumption per registered MIoT user may be better optimized.Here the energy consumption is that consumed by the VNFs in the MIoT network slice.In contrast a network slice to support mobile broadband service(eMBB)may be composed of network functions instantiated in centralized or ed

50、ge data centers,and the efficiency metric of interest is energy consumed per kbps throughput.Sustainability may be used by operators to lower costs,or it may be offered as a service to subscribers and optionally monetized.In both examples above(MIoT and eMBB),sustainability may be a network slice at

51、tribute,and slices for a given service type may be created by a network operator to support sustainability trade-offs.For example,a network operator could offer consumers a choice of service with:an eMBB network slice,with high sustainability,but at a higher price or lower data rate or lower reliabi

52、lity an eMBB network slice,with lower sustainability,but at a lower cost or a higher data rate or higher reliability an MIoT network slice,with high sustainability,but at a higher price or lower data rate,etc.an MioT network slice,with lower sustainability at a lower cost or higher data rate,etc.In

53、6G,sustainability as a subscriber service is expected to become further integrated with the network.This may be realized by providing sustainability as an attribute of a network slice used by subscribers,or as a service to individual subscribers,similar to how Quality of Service(QoS)is currently pro

54、vided.Specifically,in the 5GS,subscriber service 36G APPLICATIONS RUNNING ON IT ASSETSSlice/Service typeDescriptioneMBB5G enhanced Mobile BroadbandURLLCUltra-reliable low latency communicationsMIoTMassive Internet of Things(IoT)V2XV2X servicesHMTCHigh-Performance Machine-Type CommunicationsHDLLCHigh

55、 Data rate and Low Latency Communications8allows differentiated QoS and charging.This is typically enabled via a Policy Control Function(PCF)that determines on a per-subscriber,per application basis the parameters of service(i.e.,charging rate and 5GS QoS attributes such as latency,guaranteed bit ra

56、te,etc.).Network function selection may be affected by the requirement to meet QoS attributes.For example,for ultra-low latency communication,a Power Distribution Unit(PDU)Session Anchor(PSA)UPF network function at an edge or enterprise data center may be selected rather than a PSA UPF at a centrall

57、y located data center.This shortens the distance and number of network hops for providing connectivity in a local environment such as a factory.Similar functionality may be extended in 6G to provide sustainability as a service.In 5G Advanced,3GPP is already studying 7 how subscription and policy con

58、trol can be enhanced to support energy efficiency and energy saving as a service.Furthermore,Internet Engineering Task Force(IETF)Computing-Aware Traffic Steering(CATS)working group and 3GPP Edge Computing 8 are determining how networking and compute resources may be coordinated.That effort may invo

59、lve selecting a user application instance in coordination with mobility network anchor(i.e.,PSA UPF NF)on a data center compute node that is as close as possible to the UE and where compute resources are available.Doing so with sustainability metrics in mind is a logical extension of that capability

60、(i.e.,selecting both the user application instance and the PSA UPF network function considering data center sustainability as well as compute and network resources).The emerging ability to select compute resources according to sustainability requirements leads to an interrelationship between network

61、 KPIs used by network operators to assess sustainability and KPIs used to assess data center sustainability.3.2 Existing KPIs(Compute/Energy)5GS sustainability KPIs,for example those defined by 3GPP in“5G End to End KPIs”9 and“Management and Orchestration;Energy Efficiency of 5G”10 focus on energy c

62、onsumption and energy efficiency rather than overall sustainability.VNF residing as application workloads in data centers are an important component in the total energy and sustainability picture,though other components,particularly radio and transport networks are the biggest energy consumers and a

63、lso have the biggest impact on overall sustainability.Relevant 3GPP defined KPIs for discussion in this paper include:Network Slice Energy Efficiency=Network Slice Performance/Slice Energy Consumption,where the slice performance metric varies according to the type of network slice.For example,for eM

64、BB the performance may be based on data volume,for URLLC the performance may be based on a combination the experienced network latency and the data volume and for MIoT it may be based on the number of registered subscribers or number of active MIoT UEs.A portion of the network slice energy efficienc

65、y may be attributed to VNF workloads residing in data centers.For this component,a corresponding Network Slice Energy efficiency may be based on the Slice Energy Consumption of the VNF.Energy Consumption of VNFs can be approximated from the estimated energy consumption of its constituent Virtualized

66、 Network Function Components(VNFCs).The latter is determined from the estimated energy consumption of the virtual compute resource instances on which a VNFC runs.The energy consumption of a virtual compute at the server level can be estimated based on either its mean vCPU usage or its mean vMemory u

67、sage or its mean vDisk usage or its I/O traffic volume,as per 9.Energy Consumption of a Virtualized 5GC or 5GC Network Slice may be obtained by summing the energy consumption of each network function or the proportion of each network function that comprises the network slice.5GC Energy Efficiency,wh

68、ich is similar to the network slice energy efficiency without attribution of network function functionality to slices.The 5GC energy efficiency may be obtained for a private network.Similar metrics are expected to be applicable for 6G networks,and may include virtualized RAN components when applicab

69、le.It should be noticed that multiple network slices can use the same network element.In such a case,the energy consumption of the network element by a particular network slice cannot be directly measured but can only be estimated.Therefore,energy usage and energy efficiency metrics of a network sli

70、ce are determined approximately.If a network slice provider cannot get a full access of power consumption measurements with reasonable accuracy for the used virtualized infrastructure components hosted in a public and/or hybrid clouds,options to enhance energy efficiency of a specific network slice

71、are very limited.3.3 Current Limitations and Relevance to the Data Center KPIsThe KPIs described in Section 4.2 have limitations that may at least partially be addressed in the 6G time frame with the aid of a robust set of data center KPIs.This includes:1.Current sustainability KPIs are focused heav

72、ily on energy consumption and energy efficiency.Other sustainability metrics such as data center water usage,waste generation,land usage,building sustainability,etc.should also be considered.Note for a given data center,this includes static KPIs such as land usage and KPIs that may vary over time,su

73、ch as water usage,particularly for cooling.2.Energy consumption and efficiency KPIs based on compute resource utilization estimates do not consider 9other factors such as data center energy consumption for cooling/heating.These may vary by data center,season and time of day(night vs day).More comple

74、te energy consumption and efficiency KPIs should consider all aspects of energy consumption.3.The source of data center energy should be considered.This may include whether the energy generation is sustainable(e.g.,wind/solar/hydro power vs coal)and whether energy waste can be harvested and reused.T

75、hese may vary by data center,season and time of day(e.g.,as the generation plant mix on the grid changes).Data center KPI benchmarks considered in Section 6 provide the foundation to determine 6G KPIs that can factor in these considerations.104.1 Guidelines for a Sustainable Compute EnvironmentTo ac

76、hieve Net Zero GHG emissions targets,it is important to understand a products GHG emissions during the production stage in addition to the emissions while in its operation.Sustainable design of computing server is an integrated,holistic approach that helps to reduce emission of all phases of a data

77、center life cycle.While well-established procedure and metrics are available to assess GHG emissions from products in use,there are challenges including lack of consistent approach and metrics to accurately assess and control of GHG emissions released during the production.A good practice to develop

78、 a sustainable computing server is presented below.Material Selection Reduction of the embodied carbon starts with material selection considering the following factors:-Compliance with government regulations.-Adopt recyclable materials such as metal,plastics,fiber etc.,biodegradable,and synthetic ma

79、terials.-Responsibly manage conflict minerals(3TG such as tin,tungsten,tantalum,and gold).-Select low power,low carbon components.System Design Principle Simple Design of component type/quantity,share common design cross products,reduce system ingredients,decrease manufacture/service time,simplify a

80、ssembly/disassembly process.Circular design targeting no linear life cycle for product,reducing waste,and continuously adding value to the ecosystem.Server design needs to consider the maximum reuse,prioritizing for Reliability,Recyclability,Repairability,and Reusability(4R).Modularization:-Decouple

81、 server system components/functions from each other to reduce complexity,dependencies,and to ensure the module can be reused throughout its lifecycle-Reduce waste through standardized components and modular designs.A large-scale deployment of modules can also improve energy efficiency to achieve red

82、uce the carbon footprint of data center operations Design for easy manufacturing and service-Fewer parts-Standard components-Multi-use-Easy fabrication Design for telemetry and control-On-board telemetry in both board and system for power and thermal monitoring and control-Quick control mechanism fo

83、r power reduction,thermal cooling,and workload optimization Design for energy efficiency-Low power components and low power state management-Simple power reduction policy matching workload demand and achieve the lowest idle with single bios switch;Active Idle Efficiency Mode etc.-System power and pe

84、rformance management feature to utilize system power appropriate for certain performance demand.-Liquid cooling solutions:cold plate,immersion cooling,precision cooling etc.System package design Use recyclable materials Multiple pack design Eliminate heavy metals Manufacture carbon reduction Commit

85、to environmental sustainability and follow ISO 14064 standards 11 Low-Temperature Soldering(LTS)instead of wave soldering process Optimize Product Life Cycle Product longevity:Durable,repairable design Extended services Reuse:System module can be reuse on new product Recycle:All end-of-life products

86、 can be recycled4IT ASSETS-COMPUTE,NETWORK,STORAGE114.2 Example of Scope 3 Evaluation In the context of sustainability,IT assets within data centers,including compute,network,and storage components,play a crucial role.These assets have environmental impacts that can be measured in Scope 2 and Scope

87、3 emissions.Scope 2 is all the embodied carbon associated with the electricity production consumed by the IT assets.The more energy efficient the IT asset is,the lower the Scope 2 emissions would be for the same amount of compute,storage,or network traffic.The Scope 3 emissions are the embodied carb

88、on associated with the manufacturing and sourcing of the IT assets.As a result,there is a direct impact on Scope 3 emissions based on the refresh cycle of the IT assets.The longer we keep an asset,the less frequent it needs to be replaced over time.Consequently,the Scope 3 emissions will be lower ov

89、er the same time period.By exploring these aspects,we aim to gain a comprehensive understanding of the sustainability implications associated with IT assets in data centers and identify strategies to enhance their environmental performance.In order to help quantify the impact,Schneider Electric has

90、developed a Data Center Lifecycle CO2e Calculator 12 which allows users to analyze a data centers footprint over its entire lifecycle.For this paper,we only focus on the Scope 3 emissions of the IT assets.To calculate the Scope 3 emissions of the IT assets,one needs to consider the average embodied

91、carbon of the data centers servers or networking equipment.It is expressed as per unit of power they consume(tonnes CO2e/kW).The power is generally estimated as a fraction of the rated power supply capacity,typically below 50%of the rated capacity.The embodied carbon data may be available from the I

92、T equipment suppliers and is typically based on the Product Attribute Impact Algorithm(PAIA)model 13.The calculator by default assumes a hypothetical rack with a mix of servers from different vendors.The user has the options to adjust the refresh interval and the embodied CO2e intensity based on the

93、 PAIA model,or the data obtained from the particular vendors.As an example,we define the refresh cycle for server and storage every 3 years and the network equipment every 5 years.Using the PAIA industry validated assumption of 9.0 t/kW.We assume a 1MW data center capacity and a mix of 95%for server

94、 and storage and 5%for networking equipment.The overall data center IT capacity accounts for 70%of the 1MW overall capacity with a rack density of 10 kW/rack.As indicated by the calculator,the CO2e embodied carbon over a 20-year period equals 59,752 t of CO2e.If,for example,the IT server/storage is

95、prolonged to be refreshed every 5 years,the 20-year CO2e is 41,918 t.The following screenshot depicts the first scenario where the IT server and storage equipment is going to be replaced every 3 years.The value of the calculator is to allow users to input data center specific information based on th

96、eir operations.Figure 2:Data Center Lifecycle CO2 Calculator 1212Data center operators need to ask themselves:What does the evolution of IT equipment mean for the overall impact and sustainability of data center operations?How do they correspond with our broader environmental objectives?Where should

97、 I focus my data center carbon-reduction efforts?What does the refresh cycle of IT equipment mean for Scope 2 and Scope 3 emissions?13The purpose of data centers in their various forms is to offer the physical infrastructure to host IT assets.It is imperative that data centers are designed,built,and

98、 operated in sustainable fashion so to minimize the impact on the environment.As the global data center capacity,by power consumption,is projected to go from 460TWh in 2022 to over 1,000TWh by 2026 3,this becomes a critical endeavor.This section will discuss the various types of data centers,their d

99、esign,build,maintain,and operate attributes.KPIs described in these sections can enable the evaluation of sustainability characteristics.5.1 Key Trends in Data CentersWith the increase in the digital infrastructure,hyperscalers,colocation service providers and edge applications are experiencing rapi

100、d expansion in the data center industry.Hyperscale data centers are designed with an enhanced capacity to handle higher data volume than the enterprise data centers and possess a range of distinctive characteristics that set them apart.These are designed to operate at scale and can support thousands

101、 of servers.Per the Hyperscale Data Centers Global Market Report 2023 report published by the business research company 14,the global hyperscale data center market is predicted to grow from$35.72 billion in 2022 to$76.73 billion in 2027 at Compound Annual Growth Rate(CAGR)of 16.5%.Additional growth

102、is expected from the colocation providers,enterprise and edge applications.The following figures show projected power consumption by data center type,while the colocation segment appears to grow the fastest.Hyperscale facilities also help in decreasing the carbon footprint.By implementing methodolog

103、ies that include renewable energy sources,free cooling,consolidation,and waste recycling,hyperscalers are becoming more energy efficient.Data center consolidation is another factor driving hyperscalers demand.However,challenges remain,such as the environmental impact of these massive facilities,a sh

104、ortage of skilled professionals to manage them,and the ever-present threat of cyberattacks.5.2 Types of Data Centers5.2.1 Hyperscale Data CentersHyperscale data centers are becoming a crucial part of the infrastructure due to the growing need for cloud services and applications that require massive

105、data processing.These data centers provide significant processing power,the ability to scale up effortlessly,and superior efficiency.Hence,they are an optimum selection for entities that deal with large-scale,crucial tasks.There are several dominating factors that are shaping the evolution of the Hy

106、perscale Data Center market.In addition to building out more capacity,the current trend is focusing on sustainability,with massive scale data centers adopting technologies that are energy efficient and the use of renewable energy sources to minimize their carbon footprint.With the continuous evoluti

107、on of environmental,social,and governance(ESG)regulations,enterprises are making efforts to collaborate with cloud service providers.These providers not only optimize the advantages of cloud migration but also help reduce their environmental footprint.Prefabricated data center modules enable rapid d

108、eployment and scalability,allowing for swift capacity increases as demanded.These modules can house high-density servers,packing more processing power into less space,maximizing efficiency and performance per square foot.Custom cooling solutions,ranging from liquid to free air cooling,optimize energ

109、y consumption and thermal management 5DATA CENTERSFigure 3:Data Center Power Consumption Growth Predictions 1414within these densely packed facilities.Automation permeates every aspect,from provisioning new servers to managing workloads,mitigating human error and maximizing operational efficiency.Hy

110、perscale data centers are designed to handle extensive amounts of data,traffic,and computation processing and are characterized by its scalability,both horizontally and vertically.These are becoming a preferred choice for organizations which deal with large-scale data processing and storage requirem

111、ents.Hyperscale data centers are designed to operate optimally,with numerous redundancy systems to ensure consistent availability and dependability.Given that they typically host mission-critical workloads,their infrastructure is required to have the adaptability and flexibility to distribute comput

112、ing resources as per demand.The exponential growth of cloud computing necessitates ever-growing hyperscale data center capacity.Additionally,data-hungry technologies like AI and IoT require scalable and efficient data centers.Hyperscalers,heavily invested in these technologies,design their facilitie

113、s accordingly.To serve a global user base,geographically distributed data centers are essential.Hyperscalers strategically select locations based on optimal connectivity,energy availability,and cost efficiencies.Hyperscale computing simplifies networking by integrating the compute,storage,and virtua

114、lization layers into a unified computing ecosystem.The architecture of a hyperscale data center is engineered to be extremely streamlined and flexible,facilitating the distribution of processing tasks across the infrastructure and the swift inclusion or exclusion of servers or other resources in res

115、ponse to fluctuations in capacity requirements.From a 6Gs perspective,the infrastructure service provider may expose power usages of the virtualized infrastructure components with reasonable accuracy,improvement options of energy efficiency of a specific VNF/CNF may be limited as a mobile operator h

116、as no direct control of hardware resources.Looking ahead,collaboration between industry stakeholders on standardization,sustainability,and talent development will be crucial for responsible growth.Continued innovation in modularization,cooling technologies,and AI will further shape the future of hyp

117、erscale data centers.Ultimately,ensuring responsible and sustainable practices remains paramount as these digital titans continue to power our ever-evolving world.5.2.2 Central Office and On-Premise While there are different applications,such as telecommunication specific application vs enterprise a

118、pplications,on-premise,including Central Office and other Enterprise Data Centers are owned by the companies and are typically designed to be located on the same business premises.Businesses have more control of on-premises IT assets.These on-premises data centers are an ideal option to meet single-

119、tenancy requirements.These let enterprises handle their data and security directly.While the clouds meteoric rise has challenged the role of on-premises data centers,these are undergoing a metamorphosis,adapting to new trends and finding their niche in the modern IT landscape.They are modernizing wi

120、th energy-efficient upgrades and software-defined networking for agility and cost-effectiveness.The“cloud vs.on-premises”debate has yielded to a more nuanced reality:hybrid deployments.Enterprises are increasingly adopting a hybrid approach,strategically allocating workloads between public clouds an

121、d on-premises data centers based on factors like security,compliance,latency,bandwidth,and cost.This trend ensures optimal balance between flexibility,scalability,and control.While hyperscalers handle mass workloads,on-premises data centers can provide value in specialized areas.High-performance com

122、puting(HPC)for scientific research,mission-critical applications demanding stringent security,and data processing requiring low latency benefit greatly from the dedicated resources and control offered by on-premises facilities.Recognizing the need for agility and efficiency,on-premises data centers

123、are embracing modernization initiatives.Infrastructure upgrades with energy-efficient cooling systems,software-defined networking for automation,and containerization for flexible resource allocation are becoming commonplace.This modernization wave helps them compete with the agility and scalability

124、often associated with the cloud.The rise of edge computing presents a unique opportunity for on-premises data centers.Strategically located at the network edge,these facilities can be leveraged to host micro data centers,supporting edge computing deployments and reducing latency for geographically d

125、ispersed users.This synergy unlocks new possibilities for IoT applications,remote healthcare,and real-time analytics.Concerns around data privacy and regulatory compliance are driving continued demand for on-premises data centers.5.2.3 Edge DeploymentsWith the explosion of internet-connected devices

126、(IoT),low-latency applications,and the rise of AI,processing power needs to move closer to the action.Those edge data centers typically host a smaller scale of physical computing and network infrastructures like those found in traditional datacenters.They are strategically positioned at the network

127、edge to provide faster,more efficient data processing.Unlike traditional data centers,edge facilities are distributed closer to the users and machines and data sources where they manage data storage,processing,and assessment nearby the location of the end users.This minimizes latency,crucial for rea

128、l time applications like autonomous transportation,remote surgery,quality control and assurance,Augmented and Virtual Reality(AR and VR)experiencesEdge data centers are frequently designed to be compact and modular,easily deployed and scaled based on specific needs.Micro data centers,housed in shipp

129、ing containers or other prefabricated units,are becoming increasingly popular,offering rapid deployment.Edge data centers often work hand 15in hand with cloud providers.Data can be preprocessed at the edge for faster analysis,then sent to the cloud for long term storage or deeper processing.This hyb

130、rid approach leverages the strength of both for optimal performance and efficiency.High performance applications,such as AI are finding a home at the edge from analyzing sensor data and factories to powering intelligent traffic management systems,an application benefit greatly from lower latency and

131、 real time insights offered by the edge computing.This trend is expected to accelerate as AI applications become more diverse and ubiquitous.Options to improve energy efficiency and suitability may be limited due to constraints of edge datacenter locations.At the same time,lower overall power consum

132、ption opens the possibility to leverage sustainable and renewable onsite power generation more easily.As the edge data center handles sensitive information closer to the users,security becomes paramount.Implementing robust security measures like encryption,access control,and threat detection is cruc

133、ial to safeguard data and ensure reliable operations.Edge data centers can also solve data sovereignty requirements as user specific data is stored within their jurisdictions.5.2.4 Application ServerAn application service provider provides special applications to a mobile operator,e.g.,AI applicatio

134、n,video stream service.This application server can be hosted in a separate data center from the mobile operators datacenter.The physical infrastructures and the virtualization of the physical hardware(servers,storage,and other networking hardware)are entirely managed by the application service provi

135、der.While the application service provider may expose power usages of the virtualized infrastructure components with reasonable accuracy,improvement options of energy efficiency of a specific VNF/CNF may be limited as a mobile operator has no direct control of hardware resources.5.3 Data Center Arch

136、itecture,Redundancy,Availability and Impact on Sustainability The high demand of various cloud applications and computing resources increases the need for deployment of large-scale datacenter facilities and/or edge computing nodes.Providers considering establishing a data center facility need to mak

137、e a pre-analysis of the requirements related to selection of location,server/hyper scalers capabilities and capacity,cooling demand of the facility,and power infrastructure capabilities.This consideration needs to be taken for the efficiency and availability that shall be provided by the data center

138、.Considering the current 3GPP service requirements 15 for various applications and service requirement for various domains 16,availability of service is one of the key performance requirements for applications stated.For 6G,3GPP are currently looking onto the 6G requirement,that will be in significa

139、ntly higher demands compared to 5G.Users of data center infrastructure typically define service availability as part of Service Level Agreements(SLA).As a result,data center provider design,build,operate,and maintain their infrastructure with various redundancies in mind.An increase in redundancy an

140、d availability of services results in more infrastructure equipment,particularly in the power and cooling infrastructure.Consequently,there is an associated impact on sustainability.The greater the redundancy,the more equipment,the greater the impact on the environment.Additionally,as redundancy inc

141、reases,the operational efficiency decreases as the typical optimal equipment performance is achieved at near maximum load.As an example,from an architecture and system developments perspective,an analysis needs to consider the single point of failure in the data center architecture,determining the c

142、onstraints of the system,related from the perspective of Hardware(HW)selection,Software(SW),cooling to server HW,enclosures,power infrastructure,power grid considerations and last selection of operation and functionality orchestrating the servers,or e.g.,virtual machines(VM).Figure 4:One example of

143、redundancy consideration in a datacenter 17Figure 4 shows some components of a redundant data center infrastructure that consists of UPS(Uninterrupted Power Supply units+batteries,PDUs),servers and switch boards,but it also includes is cooling,local controller and central management system,though no

144、t visible in Figure 4.Availability is not only determined by a high-end datacenter.The communication service availability and positioning service availability on network also defines it.Furthermore,it is also related to the power availability and total system availability in the end to end(e2e)netwo

145、rk.UPS 1.000kWSwitch boardUPS 1.000kWSwitch boardserverserverserverserver165.4 Current Data Center Sustainability KPIs and their ImpactVarious industry organizations in conjunction with typically large data center operators are pushing sustainability efforts.The iMasons Climate Accord 18 is a data c

146、enter industry coalition united on carbon reduction.Their mandate is to achieve global carbon accounting of ICT infrastructure to influence market-based decisions and drive the industry to achieve carbon neutrality.The Uptime Institute 19 has evolved their Tier Certifications to offer industry-wide

147、recognized sustainability services that help to identify,document,track and report on sustainability programs across the entire IT estate.As large portions of embodied carbon can be attributed to the actual building,the U.S.Green Building Councils 20 mission is to transform buildings and communities

148、 to advance human and environmental wellbeing.5.4.1 Sustainability Attributes for Data CentersThere are multiple sustainability attributes that need to be considered when implementing a sustainability strategy to reduce the impact on the environment.The following is a list of multiple sustainability

149、 recommendations to explore and implement.Energy efficiency:Data centers should optimize energy use to reduce carbon footprint and operational costs.Renewable energy use:Incorporating renewable energy sources such as solar,wind,or hydro power to power data center operations.Water conservation:Implem

150、enting water-efficient cooling systems and minimizing water usage for cooling purposes.Waste reduction:Implementing strategies to reduce,reuse,and recycle waste generated by data center build and operations.Carbon footprint:Minimizing GHG emissions associated with data center build and operations.Gr

151、een building design:Constructing data centers with environmentally friendly materials and energy-efficient designs.Sustainable sourcing:Procuring equipment and materials from sustainable and ethical sources.Environmental impact monitoring:Tracking and reporting environmental metrics to identify area

152、s for improvement and measure progress over time.Mainly driven by the hyperscalers,the data center industry has recognized that it has to transform itself into incorporating sustainable operations.The adoption of various sustainability practices,as outlined above,depends on the maturity of each data

153、 center operators,and pressure asserted by their customers and investors.There are several KPIs that are more commonly used today.In March 2024,the European Commission released a new delegated regulation concerning the energy efficiency of data centers 21.This regulation establishes a reporting fram

154、ework to evaluate the sustainability of data centers within the European Union(EU).It mandates the reporting of KPIs,including Power Usage Effectiveness(PUE),Water Usage Effectiveness(WUE),Energy Reuse Factor(ERF),and Renewable Energy Factor(REF).This marks a pivotal regulatory advancement in the re

155、porting of data centers sustainability indicators.5.4.1.1 Existing Metrics:A holistic,yet simple framework to measure data center sustainability performance should include(at least)the following performance metrics:Table 2:Data Center Sustainability Performance Metrics The metrics above are already

156、defined as ISO standards.5.4.2 Energy Efficiency5.4.2.1 Power Usage Effectiveness(PUE)Data centers use electrical energy to power its equipment to produce digital services to its users.From a sustainability perspective,this energy should be used as efficiently as possible.This means that a data cent

157、er should use as little energy as possible for other purposes than useful IT Work,and that IT Work per unit of energy should be maximized.To track this performance,metrics are needed to cover both facility and IT efficiency.Data center facility efficiency PUEPUE is a well-known and well-established

158、metric that is widely used across the global data center industry.It is an industry metric that measures how efficiently a datacenter consumes and uses the energy that powers the datacenter,including the operation of systems like powering,cooling,and operating the servers,data networks and lights.A

159、simplified way to think of PUE Is:Energy Efficiency Facility Power Usage Effectiveness(PUE)Zero Carbon Renewable Energy Factor(REF)Energy Re-use Energy Reuse Factor(ERF)Water Usage Water Usage Effectiveness(WUE)PUE=Total Facility EnergyIT Equipment Energy17The closer the PUE number is to“1,”the more

160、 efficient the use of energy.A PUE of 1.0 would mean that all energy is used for useful work in the IT Equipment,and nothing is wasted on cooling and other facility-related systems.PUE is defined in ISO standard ISO/IEC 30134-2:2016 22.As beneficial as establishing,measuring and improving PUE has be

161、en,from a sustainabilitys perspective,one can improve PUE,by reducing energy needed for cooling.This can be achieved by the use of evaporating water(swamp cooler method).While improving PUE,water is being used,which in itself is a precious resource.5.4.2.2 Renewable Energy Factor(REF)As our society

162、transitions to using more and more digital services,it is important to ensure that the energy that is being used to produce these digital services has minimal impact on carbon emissions.This can be achieved through sourcing of renewable energy to power the data center.To what extent a data center is

163、 using renewable energy can be measured and tracked in a metric called Renewable Energy Factor(REF).A simplified way to think about REF is:A higher number is better,and a maximum value of 1.0 indicates that all energy used in a data center comes from renewable energy sources.REF is defined in ISO st

164、andard ISO/IEC 30134-3:2016 23.5.4.2.3 Energy Reuse Factor(ERF)Most of the electrical energy that is used in a data center is converted to thermal energy.As our society is looking to move away from fossil fuels,thermal energy from data centers can be reused for heating applications such as district

165、heating,food production,etc.How much energy that is being reused for other purposes can be measured through ERF.It compares the amount of reused energy to the total amount of energy used in a data center.A simplified way to think about ERF is:5.4.2.4 Water Usage Effectiveness(WUE)Water is increasing

166、ly becoming a strained resource around the world.Some types of data centers rely on access to water for cooling purposes.To minimize the impact on water systems from data center operations,a data center should use as little water as possible.How much water a data center uses can be measured by compa

167、ring the amount of water used in relation to the energy used for IT equipment.This metric is described as Water Usage Effectiveness(WUE)A simplified way to think about WUE is:WUE is defined in ISO standard ISO/IEC 30134-9:2022 24.Again,as stated before,not all data centers are measuring these existi

168、ng already standardized KPIs.While these KPIs provide a good starting point,additional sustainability factors can be recommended.In one of Schneider Electrics white papers,a Guide to Environmental Sustainability Metrics for Data Centers proposes 28 key metrics to help operators on their sustainabili

169、ty journey,which include the beforementioned ones 25.Those metrics are divided across five core sustainability segments:Energy,GHG Emissions,Water,Waste,and Local Ecosystem.All listed KPIs focus on the ability to measure and ultimately aid in the improvement of efficiency,reductions of GHG emissions

170、,water usage,waste and the impact on the local ecosystem.REF=Renewable EnergyTotal Facility EnergyERF=Reused EnergyTotal Facility EnergyWUE=Water UsageIT Equipment Energy18Table 3:Schneider Electric:Environmental Sustainability Metrics for Data Centers 25Metric categoriesKey metricsUnitsEnergy(6)Tot

171、al energy consumption Power usage effectiveness(PUE)Total renewable energy consumption Renewable energy factor(REF)Energy Reuse Factor(ERF)Server utilization(ITEUsv)kWhRatio kWhRatio Ratio%GHG emissions(7)Scope 1 GHG emissions Scope 2 Location-based GHG emissions Market-based GHG emissions Scope 3 G

172、HG emissions Carbon usage effectiveness(CUE)Total carbon offsets Hourly renewable supply&consumption matchingmtCO2emtCO2emtCO2emtCO2ekg CO2e/kWh mtCO2e%Water(5)Total site water usage Total source energy water usage Water usage effectiveness(WUE)Water replenishment Total water use in supply chainm3 m

173、3m3/MWh m3 m3Waste(6)Waste generated Total waste E-waste Battery Waste diversion rate Total waste E-waste Battery Metric tonMetric ton Metric tonRatio Ratio RatioLocal ecosystem(4)Land Total land use Land-use intensity Outdoor noise Mean species abundance(MSA)m2kW/m2 dB(A)MSA/km2 195.4.3 Additional

174、KPIsIn addition to the data center infrastructure and energy usage KPIs,one of the important impacts of the above mentioned KPIs can be on energy consumption and/or carbon emission estimation of VNF.As mentioned in Section 3.2,energy consumption of a VNF can be approximated according to estimated en

175、ergy consumption of the virtual compute resource instances on which its VNFCs run.The energy consumption of a virtual compute in server level can be estimated based on its resource usage portion on the server that the virtual compute runs(e.g.,its mean vCPU usage).However,we know that a VNF not only

176、 uses data center IT equipment resources but also impacts on non-IT equipment usage of a data center such as cooling systems.The above mentioned KPIs can help to estimate energy consumption of virtual computes in data center level which is more accurate and comprehensive than the server level estima

177、tion.For example,energy consumption of a VNF can be estimated by multiplying the estimated energy consumption of a virtual compute in sever level with the energy efficiency ratio(e.g.,PUE)of the data center that the virtual compute runs.In this case,estimated energy consumption of a VNF on a datacen

178、ter(e.g.,with PUE X)can be different from another data center(e.g.,with PUE Y).Furthermore,the Power Waste Factor can be considered and is outlined below:The power waste factor(W)is a new metric 5 proposed to evaluate the power efficiency in data centers.The power consumption for data processing can

179、 be derived considering the data center as a single component.W=Paux/Pinfo(PUE 1),where W represents the Waste Factor for the data center.Where Pinfo is the sum of all powers of each component(e.g.,routers,switches,processors,and other network equipments that carry or store information)that has been

180、 used for carrying information in the system,Pnoninfo is the power used by the other components but not directly involved in data transmission(e.g.,servers,storage devices,firewalls),and Paux is the power used by the cooling systems,PDUs,and other auxiliary equipment apart from the data transmission

181、 and access/storage.Two data centers(WA and WB)can be compared based on their waste factors even though they have same PUE one may offer a lower waste factor,allowing the identification of the more energy-efficient one.To implement this new power waste factor,further refinement on how to practically

182、 measure Pinfo,Paux and particularly Pnon-info is required.20Exploring key data center sustainability strategies is pivotal for achieving Net Zero emissions in data centers and the ICT sector.Rising energy consumption of core networks and data centers is driven by the shift to cloud based 5G service

183、s,the introduction of 6G,and the overall growing energy consumption of data centers.This emphasizes the need for sustainable operations to reach Net Zero emissions.Recommendations:1.Develop a data center sustainability strategy for your organization,including objectives and potential outcomes,and co

184、mmunicate that position.2.Identify relevant sustainability KPIs applicable to data center operators and their tenants.Start with a gradual approach,select pertinent KPIs,implement them and determine their effectiveness.Continue to integrate additional KPIs to achieve a truly sustainable data center.

185、By using these recommendations as a guide,companies can effectively apply the insights from the whitepaper to drive sustainable practices and make well informed decisions regarding data center operations and design.6SUMMARY AND RECOMMENDATIONS214R.Reliability,Recyclability,Repairability,and Reusabil

186、ity AI .Artificial Intelligence AR .Augmented RealityCAGR.Compound Annual Growth RateCATS.Computing-Aware Traffic SteeringCUE.Carbon Usage Effectiveness DN.Data Networke2e.End to end eMBB.Enhanced Mobile Broadband ERF.Energy Reuse Factor ESG.Environmental,Social,and Governance EU .European Union GHG

187、.Greenhouse Gas GSMA.Global System for Mobile Communications AssociationHDLLC.High Data Rate and Low Latency Communications HMTC.High-Performance Machine-Type CommunicationsHPC.High-performance ComputingHW .HardwareICT .Information Communication TechnologyIETF.Internet Engineering Task ForceIoT.Inte

188、rnet of ThingsISO.International Organization for Standardization KPI.Key Performance Indicator LTS.Low-Temperature Soldering MIoT .Massive Internet of ThingsMSA.Mean Species Abundance NF .Network Function OCC.Observability,Choice,and Circular Economy P-A-I-A.Product Attribute Impact Algorithm PCF.Po

189、licy Control Function PDU.Power Distribution Unit PSA.PDU Session AnchorPUE.Power Usage Effectiveness QoS .Quality of ServiceRAN .Radio Access Network7ABBREVIATIONS AND ACRONYMS2122REF.Renewable Energy FactorSLA.Service Level AgreementSW.SoftwareUE .User EquipmentUN.United NationsUPS .Uninterrupted

190、Power SupplyURLLC.Ultra-Reliable Low Latency Communications V2X.Vehicle-to-everythingVM.Virtual MachinesVNF.Virtualized Network Function VNFC.Virtualized Network Function Components VR.Virtual Reality WUE.Water Usage Effectiveness 22238REFERENCES1.United Nations.World heading towards new temperature

191、 records,UN weather watchdog warns.UN News.https:/news.un.org/en/story/2024/06/1150656 2.GSMA Intelligence.“Going green:measuring the energy efficiency of mobile networks”https:/ 3.International Energy Agency(IEA).Electricity 2024 Analysis and Forecast to 2026.https:/ G Alliance Report:Evolution of

192、Sustainability Indicators for Next Generation Radio Network Technologies https:/nextgalliance.org/white_papers/evolution-of-sustainability-indicators-for-next-generation-radio-network-technologies/5.Ying,M.,Shakya,D.,Poddar,H.,&Rappaport,T.S.(2023).Waste Factor:A New Metric for Evaluating Power Effi

193、ciency in any Cascade.In GLOBECOM 2023-2023 IEEE Global Communications Conference,Kuala Lumpur,Malaysia,2023,pp.1-6,doi:10.48550/arXiv.2309.010186.3GPP TS.23.501.“System Architecture for the 5G System(5GS)”.7.3GPP TR.23.700-66.“Study on Energy Efficiency and Energy Saving”.8.3GPP TS.23.548.“5G Syste

194、m Enhancements for Edge Computing;Stage 2”.9.3GPP TS.28.554.“Management and orchestration;5G end to end Key Performance Indicators(KPIs)”.10.3GPP TS.28.310.“Management and Orchestration:Energy Efficiency of 5G”.11.ISO 14064-1:2018.Greenhouses gases Part 1:Specification with guidance at the organizat

195、ion level for quantification and reporting of greenhouse gas emissions and removals,https:/www.iso.org/standard/66453.html12.Schneider Electric,Data Center Lifecycle CO2e Calculator,https:/ Systems Laboratory.Product Attribute to Impact Algorithm(PAIA).https:/msl.mit.edu/projects/paia/main.html14.Mc

196、Kinsey&Company.(2023).“Investing in the rising data center economy,”https:/ 15.3GPP TS.22.261.“Service Requirements for the 5G system”.16.3GPP TS.22.104.“Service requirements for cyber-physical control applications in vertical domains”.17.Data Center Redundancy:N+1,2N,2(N+1)or 3N2(distributed),https

197、:/ Climate Accord,Achieving carbon neutrality in digital infrastructure,https:/climateaccord.org/2419.Uptime Institute,“Uptime Institutes Global Data Center Survey Results 2024,”https:/ 20.U.S.Green Building Council,https:/www.usgbc.org/21.European Commission,Energy Efficiency Directive,https:/energ

198、y.ec.europa.eu/topics/energy-efficiency/energy-efficiency-targets-directive-and-rules/energy-efficiency-directive_en 22.ISO/IEC 30134-2:2016.Information TechnologyData CentresKey Performance IndicatorsPart 2:Power Usage Effectiveness(PUE),https:/www.iso.org/standard/63451.html 23.ISO/IEC 30134-3:201

199、6.Information technology Data centres Key performance indicators Part 3:Renewable energy factor(REF),https:/www.iso.org/standard/66127.html 24.ISO/IEC 30134-9:2022.Information technology Data centres key performance indicators Part 9:Water usage effectiveness(WUE),https:/www.iso.org/standard/77692.h

200、tml 25.Schneider Electric.“Guide to Environmental Sustainability Metrics for Data Centers.”https:/ Report Leader:Carsten Baumann,Schneider Electric Industry Contributors:Rohit Abhisek,AT&T Lackis Eleftheriadis,Ericsson Bhushan Joshi,Ericsson Micaela Giuhat,Microsoft Corporation Ehsan Ahvar,Nokia Sho

201、hreh Ahvar,Nokia Gagandeep Bhatti,Nokia Colin Kahn,Nokia Margaret Mauch,Schneider Electric Academic Contributors:Ali Mehrizi-Sani,Virginia Tech 26GREEN G WORKING GROUPGREEN GThe mission of the Green G WG is to position North America as the global leader in environmental sustainability by creating a

202、sustainable 6G ecosystem and enabling other industries to reduce greenhouse gases and energy consumption,limit land and water use,and move towards circular economy.Green G Working Group LeadershipChair,Bhushan Joshi,Ericsson Vice Chair,Ralf Bendlin,AT&TVice Chair,Gagandeep Bhatti,Nokia Ian Deakin,AT

203、IS27GREEN G WORKING GROUPGreen G Working Group Membership Analog Devices,Inc.AnritsuApple Inc.AT&TATISBell CanadaCableLabsCase Western Reserve UniversityCharter CommunicationsCiena CorporationCiscoCohere TechnologiesDeepSigDOCOMO Innovations Inc.EricssonFlorida Atlantic UniversityFutureweiGeorge Mas

204、on UniversityGeorgetown UniversityGoogleHewlett Packard EnterpriseIMECIntel CorporationInterDigital Communications CorporationIowa State UniversityITRIJohns Hopkins University Applied Physics LaboratoryKeysight Technologies,IncLG ElectronicsLockheed MartinMavenirMediatek IncMicrosoft CorporationMITR

205、E CorporationMotorola Mobility LLC (A Lenovo Company)Murata Manufacturing Co.,Ltd.NISTNokiaNorth Carolina State UniversityNorthwestern UniversityNTIAOfinnoOld Dominion UniversityPurdue UniversityQualcomm IncorporatedSamsung Research AmericaSchneider ElectricSharp Labs of AmericaSpreadtrum Communicat

206、ion USA,Inc.T-Mobile USATDSTELUSUniversity of ManitobaUniversity of Notre DameVerizonVerizon WirelessViavi SolutionsVirginia TechVT-ARCWIN SEMICONDUCTORS CORP.Wireless Internet of Things at Northeastern University28NEXT G ALLIANCE REPORTSNext G Alliance Report:6G Distributed Cloud and Communications

207、 SystemNext G Alliance Report:Trust,Security,and Resilience for 6G SystemsNext G Alliance Report:Digital World ExperiencesNext G Alliance Report:Cost-Efficient SolutionsNext G Alliance Report:Sustainable 6G Connectivity A Powerful Means of Doing GoodNext G Alliance Report:AI-Native Wireless Networks

208、Next G Alliance Report:6G Radio Technology Part II:Basic Radio TechnologiesNext G Alliance Report:6G Roadmap for Vertical Industries6G Market Management and OrchestrationA L L I A N C EAn ATIS InitiativeNext G Alliance Report:Beyond Speed:Promoting Social and Economic Opportunities through 6G and Be

209、yondNext G Alliance Report:Beyond Speed:Promoting Social and Economic Opportunities through 6G and BeyondNext G Alliance Report:6G Technologies for Wide-Area Cloud EvolutionNext G Alliance Report:6G Radio Technology Part I:Basic Radio TechnologiesNext G Alliance Report:6G Spectrum ConsiderationsNext

210、 G Alliance Report:Network-Enabled Robotic and Autonomous SystemsNext G Alliance Report:Terminology for Frequency RangesNext G Alliance Report:Shaping Tomorrow:The Evolution of Personalized Digital Experiences Through 6G TechnologiesNext G Alliance Report:Multi-Sensory Extended Reality(XR)in 6GNext

211、G Alliance Report:Distributed Sensing and CommunicationsA L L I A N C EAn ATIS InitiativeNext G Alliance Report:6G TechnologiesComponent TechnologiesRadio TechnologiesSystem and Network ArchitectureNetwork OA&M and Service EnablementTrustworthiness Security,Reliability,Privacy,&ResilienceJune 2022Ne

212、xt G Alliance Report:6G TechnologiesA L L I A N C EAn ATIS Initiative EVERYDAY LIVINGEXPERIENCECRITICAL ROLESSOCIETAL GOALSNext G Alliance Report:6G Applications and Use CasesNext G Alliance Report:6G Applications and Use CasesNext G Alliance Report:Roadmap to 6GFebruary 2022Next G Alliance Report:R

213、oadmap to 6GGreen G:The Path Toward Sustainable 6G6G Market Development A North American PerspectiveThe Next G Alliance(NGA)is a bold new initiative to advance North American mobile technology leadership over the next decade through private sector-led efforts.With a strong emphasis on technology com

214、mercialization,NGAs scope of activities encompasses the full lifecycle of research and development,manufacturing,standardization,and market readiness.NGAs growing membership reflects support from stakeholders in academia,government,and industry.This Perspective outlines the strategic importance of 6

215、G and the imperatives that will shape North Americas competitiveness,economy,and global leadership.A L L I A N C EAn ATIS Initiative6G Market Development:A North American PerspectiveNext G Alliance Report:6G Sustainability KPI Assessment Introduction and Gap Analysis1A L L I A N C EAn ATIS Initiativ

216、eNext G Alliance Report:North American 6G Roadmap Priorities Next G Alliance Report:6G Sustainability KPI North American 6G Roadmap PrioritiesNext G Alliance Report:Evolution of Sustainability Indicators for Next-Generation Radio Network TechnologiesNext G Alliance Report:Channel Measurements and Mo

217、deling for Joint/Integrated Communication and Sensing,as well as 7-24 GHz CommunicationNext G Alliance Report:Spectrum Needs for 6GNext G Alliance Report:Spectrum Working Group:Spectrum Access Mechanisms29Published September 2024Copyright 2024 by Alliance for Telecommunications Industry Solutions Al

218、l rights reserved.Alliance for Telecommunications Industry Solutions 1200 G Street,NW,Suite 500 Washington,DC 20005No part of this publication may be reproduced in any form,in an electronic retrieval system or otherwise,without the prior written permission of the publisher.For information,contact AT

219、IS at(202)628-6380.ATIS is online at http:/www.atis.org.The information provided in this document is directed solely to professionals who have the appropriate degree of experience to understand and interpret its contents in accordance with generally accepted engineering or other professional standar

220、ds and applicable regulations.No recommendation as to products or vendors is made or should be implied.NO REPRESENTATION OR WARRANTY IS MADE THAT THE INFORMATION IS TECHNICALLY ACCURATE OR SUFFICIENT OR CONFORMS TO ANY STATUTE,GOVERNMENTAL RULE OR REGULATION,AND FURTHER,NO REPRESENTATION OR WARRANTY

221、 IS MADE OF MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OR AGAINST INFRINGEMENT OF INTELLECTUAL PROPERTY RIGHTS.ATIS SHALL NOT BE LIABLE,BEYOND THE AMOUNT OF ANY SUM RECEIVED IN PAYMENT BY ATIS FOR THIS DOCUMENT,AND IN NO EVENT SHALL ATIS BE LIABLE FOR LOST PROFITS OR OTHER INCIDENTAL OR C

222、ONSEQUENTIAL DAMAGES.ATIS EXPRESSLY ADVISES THAT ANY AND ALL USE OF OR RELIANCE UPON THE INFORMATION PROVIDED IN THIS DOCUMENT IS AT THE RISK OF THE USER.NOTE-The users attention is called to the possibility that compliance with this document may require use of an invention covered by patent rights.

223、By publication of this document,no position is taken with respect to whether use of an invention covered by patent rights will be required,and if any such use is required no position is taken regarding the validity of this claim or any patent rights in connection therewith.Please refer to www.atis.o

224、rg/legal/patentinfo.asp to determine if any statement has been filed by a patent holder indicating a willingness to grant a license either without compensation or on reasonable and non-discriminatory terms and conditions to applicants desiring to obtain a license.COPYRIGHT AND DISCLAIMER30Building the foundation for North American leadership in 6G and beyond.nextgalliance.orgA L L I A N C EAn ATIS Initiative