《易观梅森:2023额外中频频谱对5G移动网络碳足迹的影响分析报告:以6GHz频段为例(英文版)(70页).pdf》由会员分享,可在线阅读,更多相关《易观梅森:2023额外中频频谱对5G移动网络碳足迹的影响分析报告:以6GHz频段为例(英文版)(70页).pdf(70页珍藏版)》请在三个皮匠报告上搜索。
1、 Final report for Huawei Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band Janette Stewart,Sylvain Loizeau,Julia Allford 2 June 2023 Ref:728565284-225 Ref:728565284-225.Contents 1 Executive summary 1 2 Introduction 6 2.1 Background a
2、nd context 6 2.2 Objectives of the study 7 2.3 Structure of this document 8 3 5G mobile network modelling methodology,inputs and assumptions 9 3.1 Overview of approach 9 3.2 Deployment scenarios 12 3.3 User experience data rate and required capacity 13 3.4 Inputs for modelling MBB and FWA capacity s
3、upply 15 3.5 Inputs and assumptions for modelling the environmental impact 21 4 5G mobile network modelling carbon footprint results 23 4.1 Dense urban area results 23 4.2 Rural town or village results 32 4.3 Practical issues in building additional macro and small cells 37 4.4 Power efficiency measu
4、res in mobile network architectures 38 5 Relationship between Wi-Fi carbon footprint and the 6GHz band 40 5.1 Introduction 40 5.2 Fixed broadband connectivity targets and technology evolution in Europe 41 5.3 Additional spectrum and Wi-Fi throughput for typical premises 41 5.4 Implications for the e
5、nvironmental impact 44 6 Conclusions 45 Annex A 5G mobile network modelling methodology Annex B References to other published studies Annex C Bibliography Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band Ref:728565284-225.Copyright
6、2023.Analysys Mason has produced the information contained herein for Huawei.The ownership,use and disclosure of this information are subject to the Commercial Terms contained in the contract between Analysys Mason and Huawei.Analysys Mason Limited St Giles Court 24 Castle Street Cambridge CB3 0AJ U
7、K Tel:+44(0)1223 460600 Registered in England and Wales No.5177472 Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|1 Ref:728565284-225.1 Executive summary To better understand the environmental impacts of meeting future wireless co
8、nnectivity targets in Europe,and how these environmental impacts might be affected by spectrum availability,Huawei commissioned Analysys Mason to conduct this study to compare scenarios for future public mobile network deployments.Specifically,this study focuses on the carbon emissions of a Fifth Ge
9、neration(5G)mobile network in addressing future connectivity targets through the availability of additional spectrum,compared to the network meeting the same future connectivity targets without additional spectrum.Both with and without additional spectrum,we assume that the existing network grid wil
10、l be densified(in terms of the numbers of macro sites and/or outdoor small cells),but in the absence of additional spectrum,the required densification is greater due to less spectrum being available.The impact on carbon emissions over the 20222032 time period is considered,with modelling directed at
11、 achieving specific targets for future connectivity in Europe in 2030.Context for the study The spectrum we focus on in our study is mid-band spectrum,which refers to a type of spectrum that is widely used for 5G today,due to its ability to deliver both capacity and coverage.The GSMA assessed the fu
12、ture mid-band spectrum needs for 5G and beyond in a previous report,and in that context,it forecast a demand for further mid-band spectrum in the second half of the 2020s to meet connectivity needs 1.The intention of this report is to provide input to policy discussions on future 5G spectrum require
13、ments and,in Europe specifically,on the future use of mid-band spectrum,with particular reference to the upper 6GHz band(64257125MHz).While other studies have considered the impact of mid-band spectrum availability on 5G mobile networks in terms of delivering new services,meeting future traffic dema
14、nds,infrastructure requirements,or the cost of deployment,this study aims to consider the addition of mid-band spectrum from an environmental impact perspective.This consideration of impact on carbon emissions in relation to spectrum policy decisions aligns with the opinion published by the Radio Sp
15、ectrum Policy Group(RSPG)on the role of spectrum policy to combat climate change,which examines the aspects of spectrum management that relate to climate change,including considerations relating to future spectrum for 5G mobile networks 2.The methodology we developed for this study,which we describe
16、 in Section 3,is applicable to network modelling for 5G mid-band spectrum in general and is not band-specific.Results in Section 4 are also relevant to mid-band spectrum in general,although modelling assumptions regarding spectral efficiencies and coverage for the additional mid-band spectrum refer
17、to the upper 6GHz band specifically.This is because the upper 6GHz band has attracted significant interest as a candidate band for 5G evolution.Hence this report provides some specific analysis for this band.The analysis of the upper 6GHz band provides an assessment of the impact on carbon emissions
18、 for a scenario in which the band would be used for 5G,and for a scenario in which the upper 6GHz band would be used for wireless local area networks(WLANs).We also consider whether there Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz
19、band|2 Ref:728565284-225.would be an impact on the carbon footprint of Wi-Fi if the upper 6GHz band were used in addition to the existing frequency bands available for Wi-Fi in Europe.This study has considered the density of macro sites and outdoor small cells needed to meet future connectivity targ
20、ets within a representative model of site deployment in a market where three mobile operators run 5G networks(with spectrum deployments and spectral efficiencies in line with the frequency bands used for 5G in Europe today),based on two typical scenarios:5G mobile broadband(MBB)in a highly populated
21、 European city with a population density of 15 000/km(i.e.dense urban)5G MBB as well as fixed broadband through 5G-based fixed-wireless access(FWA)in a rural town or village.The rural town or village is modelled assuming a population density of 300/km and assuming the population lives outside of the
22、 reach of fibre.Modelled connectivity targets in Europe In Europe,the European Commissions Digital Decade policy programme has established objectives and targets for digital transformation,and lays out a vision to be achieved by 2030.Of most relevance to this study are targets for all end users at f
23、ixed locations to have gigabit connectivity at least equivalent in speed to that of 5G,and for all populous areas to be covered by high-capacity 5G mobile networks 3.For 5G MBB,the radiocommunications sector of the International Telecommunications Union(ITU-R,which is responsible for defining global
24、 telecommunications standards and for international frequency allocations)has defined minimum performance requirements for IMT-2020(being ITU-R terminology for 5G)as including a downlink speed of 100Mbit/s and an uplink speed of 50Mbit/s 4.In accordance with the targets described above,our study con
25、siders the infrastructure required to meet the following targets:In the dense urban area we model a 5G mobile network designed to target1 delivery of MBB services with a downlink speed of 100Mbit/s and an uplink speed of 50Mbit/s to users in the busy hour(with assumptions on the number of active use
26、rs likely to be accessing the service)In the rural town or village we model a 5G mobile network designed to target delivery of MBB and FWA services,with the MBB targets as above,and FWA supporting 1Gbit/s downlink and 200Mbit/s uplink to users in the busy hour(with the same assumptions on the number
27、 of active users as above).1 In a wireless network,the data rate experienced by a user can vary depending on the users location in a cell,the loading of the cell and propagation effects.Hence when we model a target delivery,we are modelling a network designed to achieve the stated downlink and uplin
28、k speeds to users in a busy hour,in a typical urban(or rural)cell.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|3 Ref:728565284-225.Summary of our analysis We have built a 5G mobile network model in Microsoft Excel to simulate th
29、e size of deployment needed(in terms of the densities of macro sites and outdoor small cells)in the dense urban area and in the rural town or village,with and without the additional mid-band spectrum.The model has several key assumptions and inputs defined by Analysys Mason,validated by Huawei and,w
30、here possible,aligned with published studies.For example,we have taken note of a previous study conducted on behalf of the GSMA,entitled“Estimating the mid-band spectrum needs in the 20252030 timeframe”5.Whilst that study had a different objective(to estimate the amount of additional mid-band spectr
31、um needed),there are some similarities with the approach and assumptions we have used in our modelling.In our modelling,the key output is in terms of the number of sites per km,and then we use environmental inputs per site multiplied by the number of sites to arrive at the carbon footprint.When calc
32、ulating the number of sites required,the following should be noted:We calculate the capacity of macro cells and outdoor small cells assuming a portfolio of spectrum similar to that of European mobile network operators today,i.e.using low-band spectrum(sub-1GHz),lower mid-band spectrum(1500MHz,1800MH
33、z,2100MHz and 2600MHz),upper mid-band spectrum(3500MHz plus,in our study,the upper 6GHz band),and high-band spectrum(e.g.typical millimetre-wave spectrum,such as 26GHz).We estimate a per-user growth in targeted Mbit/s in the busy hour in both the downlink and the uplink direction,reaching the target
34、 speeds for MBB(in the dense urban area)and MBB and FWA(in the rural town or village)by 2030.We then estimate the 5G mobile network carbon emissions based on the output of network modelling(in terms of sites per square kilometre)we consider for each site both the embodied footprint(which relates to
35、the fabrication,construction and installation stages of the base stations and sites)and the recurring footprint(which relates to the operation of the base stations in terms of powering and maintaining the sites).The study is focused on the infrastructure-related carbon emissions,i.e.carbon emissions
36、 associated with the production and operation of base stations.In this study,we were not asked to consider the impact on the carbon footprint of mobile devices(our model assumes a natural renewal of devices in line with their expected lifetime,rather than forcing user migration),nor do we consider t
37、he enablement impact of mobile networks on other sectors(e.g.by enabling other sectors to improve the efficiency of their real-time or remote operations).2 We note that other third-party reports do address these considerations.3 Overall,the aim of our study is to consider whether the saving in carbo
38、n emissions associated with having fewer macro sites and outdoor small cells in a mobile 2 Information on the enablement effect is available in the 2023 GSMA Report:“Spectrum:the Climate Connection Spectrum policy and carbon emissions”.20 3 For example,the GSMAs report“The enablement effect”18 discu
39、sses mobile network enablement technologies and assesses six different sectors in which these mobile network enablement technologies can reduce carbon emissions for that sector.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|4 Ref:
40、728565284-work due to having more spectrum outweighs the incremental carbon emission cost of implementing and deploying the additional spectrum in 5G mobile networks.In Section 5,considering ongoing discussions on future use of the upper 6GHz band in Europe and in some other markets,we consider simu
41、lations made available to us in order to estimate the impact of increasing mid-band spectrum availability on Wi-Fi throughput.In particular,we consider whether there is an impact in terms of the number of access points needed for Wi-Fi to meet a gigabit connectivity target in typical dense urban,and
42、 rural town and village,indoor settings,if the upper 6GHz band is used alongside existing spectrum bands already available for Wi-Fi in the 2.4GHz,5GHz and lower 6GHz bands.Key findings Overall,our analysis demonstrates that the carbon footprint of future 5G mobile networks is expected to be lower i
43、f additional mid-band spectrum is made available to meet future capacity targets,by avoiding a significant densification of macro sites and outdoor small cells.This applies both in the dense urban area and in the rural town or village we have modelled in our study.More specifically,our main conclusi
44、ons are as follows:Our modelling results show that lower network carbon emissions arise in 5G mobile networks with additional mid-band spectrum available to meet the future connectivity targets we have considered in this report,compared to a situation where networks are densified through additional
45、macro and outdoor small cells without the availability of additional mid-band spectrum.We calculate that the carbon emission savings from having less densification in 5G mobile networks outweigh the incremental carbon emission costs of deploying and operating new radios(to support the additional mid
46、-band spectrum we have modelled)on macro sites and outdoor small cells for the dense urban area and for macro sites in the rural town or village.It should be noted that in the dense urban area we consider two deployment variants for densification(in the absence of additional mid-band spectrum):first
47、ly,densification primarily via macro sites(with some supporting outdoor small cells)secondly,densification primarily via additional outdoor small cells(and thus lower macro-site densification).In each case,the incremental carbon emission cost of deploying and operating new upper mid-band radios at t
48、he dense urban macro sites and outdoor small cells is lower than the incremental carbon footprint associated with the higher level of densification needed without the additional mid-band spectrum.In addition to the increased carbon footprint associated with greater densification,the levels of densif
49、ication that would be required in 5G mobile networks to meet the connectivity targets in Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|5 Ref:728565284-225.the absence of additional mid-band spectrum would be practically challengi
50、ng and also potentially technically unfeasible(due to interference between sites that are too close to each other).For Wi-Fi,based on simulations made available to us and considering the future connectivity targets for fixed broadband(i.e.an aggregated throughput of more than 1Gbit/s per premises),t
51、he availability of the upper 6GHz band would not translate into any reduction in carbon emissions,given such targets can be met via the latest Wi-Fi technology using spectrum bands already available for Wi-Fi use in Europe(2.4GHz,5GHz and lower 6GHz).While these results have been modelled assuming u
52、pper 6GHz deployment(e.g.in terms of the bandwidth available),these conclusions may apply to other upper mid-band spectrum,provided that the alternative upper mid-band spectrum exhibits similar characteristics to those modelled here.Impact of additional mid-band spectrum on the carbon footprint of 5
53、G mobile networks:the case of the upper 6GHz band|6 Ref:728565284-225.2 Introduction This report contains the results of a study conducted by Analysys Mason to investigate the environmental impacts of meeting future wireless connectivity targets in Europe with and without additional mid-band spectru
54、m.2.1 Background and context Growth in wireless traffic in outdoor and indoor environments is well documented and has been explored in several published reports.However,most published studies tend to consider the benefits of further spectrum for one use only(e.g.Wi-Fi or mobile).Various studies spec
55、ifically on the role of Wi-Fi for future connectivity refer to the energy efficiency of Wi-Fi due to its low-power transmission,but without addressing how future connectivity targets over the wider area(e.g.to all populated areas,in line with the European Commissions(ECs)Digital Decade objectives)wi
56、ll be met.4 In the outdoor environment,much of the mobile broadband(MBB)connectivity to smartphones and other connected devices in Europe is via mobile networks,now in their Fifth Generation(5G)of deployment.These 5G mobile networks use a combination of newly assigned frequency bands harmonised in E
57、urope for 5G deployment(e.g.the 700MHz and 3.5GHz bands)together with frequency bands used for previous generations of mobile network that are now progressively being used for 5G,such as the 1800MHz or 2100MHz bands.In the indoor environment,smartphones and other connected devices either use a mobil
58、e network(outside-in coverage),or,where available,wireless local area network(WLAN)connectivity,predominantly delivered today via short-range wireless technologies such as Wi-Fi.The growth in Wi-Fi traffic has been well documented,and the question of future Wi-Fi capacity needs is being addressed th
59、rough technology evolutions and,in some markets,through the addition of further frequencies in the 6GHz band.Mid-band spectrum has been a particular focus of 5G deployment,and it will remain so as 5G networks evolve and MBB traffic increases.Mid-band spectrum can be considered as intermediate spectr
60、um between the low-band spectrum used for 4G and prior generations,and high-band spectrum in the millimetre-wave range such as the 26GHz band that provides both capacity and coverage to a mobile network.Despite the global movement to reduce carbon emissions,and sustainability goals being a core part
61、 of the vision for 6G,the impact of future mobile spectrum assignments on the wireless industrys carbon footprint has not been widely addressed.This study has sought to provide a different perspective on future mobile spectrum needs,focused on the environmental implications of mid-band spectrum assi
62、gnment,in contrast to other previous 4 We elaborate on various reports we have reviewed as part of this study in Annex B.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|7 Ref:728565284-225.studies,which have considered factors such
63、 as infrastructure deployment,financial viability,coverage and quality of service.The study has focused on the addition of mid-band spectrum to 5G mobile networks,and the implications for the carbon footprint of 5G mobile networks of using this spectrum compared to the densification needed to meet f
64、uture connectivity targets without additional mid-band spectrum.Our study has considered 5G mobile networks in two typical locations:a highly populated European city a rural town or village outside the reach of fibre networks.In Europe,the ECs Digital Decade policy programme has established objectiv
65、es and targets for digital transformation in Europe,and lays out a vision to be achieved by 2030.Of most relevance to this study are targets for all end users at fixed locations to have gigabit connectivity at least equivalent in speed to that of 5G,and for all populous areas to be covered by high-c
66、apacity 5G mobile networks 6.For 5G MBB,the radiocommunications sector of the International Telecommunications Union(ITU-R,who are responsible for defining global telecommunications standards and for international frequency allocations)has defined minimum performance requirements for IMT-2020(being
67、ITU-R terminology for 5G)as including a downlink speed of 100Mbit/s and an uplink speed of 50Mbit/s 4.In accordance with the European targets described above,our study considers the infrastructure required to meet the following targets:In the dense urban area we model a 5G mobile network designed to
68、 target5 delivery of MBB services with a downlink speed of 100Mbit/s and an uplink speed of 50Mbit/s to users in the busy hour In the rural town or village we model a 5G mobile network designed to target delivery of MBB and FWA services,with the MBB targets as above,and FWA supporting 1Gbit/s downli
69、nk and 200Mbit/s uplink to users in the busy hour.2.2 Objectives of the study The main objectives of the study have been to:explore the future infrastructure needed to meet the minimum performance requirements defined by the International Telecommunication Union(ITU)for IMT-2020/5G in populated area
70、s together with EC gigabit connectivity targets for end users at fixed locations 5 In a wireless network,the data rate experienced by a user can vary depending on the users location in a cell,the loading of the cell and propagation effects.Hence when we model a target delivery,we are modelling a net
71、work designed to achieve the stated downlink and uplink speeds to users in a busy hour,in a typical urban(or rural)cell.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|8 Ref:728565284-225.consider the need for additional mid-band s
72、pectrum for both 5G and Wi-Fi,in the context of how additional mid-band spectrum can mitigate the need for infrastructure densification compare the impacts of addressing future connectivity targets on a 5G mobile networks carbon footprint(embodied and recurring),with and without additional mid-band
73、spectrum conduct sensitivity analysis to demonstrate how the results vary based on key assumptions.2.3 Structure of this document The remainder of this document is laid out as follows:Sections 3 and 4 cover the modelling of the 5G mobile networks carbon footprint Section 3 details the inputs and ass
74、umptions Section 4 discusses the results Section 5 explores the impact of additional spectrum on the carbon footprint of Wi-Fi access networks Section 6 summarises the conclusions.The report includes a number of annexes containing supplementary material:Annex A contains a detailed modelling methodol
75、ogy Annex B includes a list of other published studies considered as part of our study Annex C provides a bibliography.6 6 Where a report/other source is mentioned in the text,the inclusion of a reference of the form n indicates that the web address for that document can be found in Annex C.Impact o
76、f additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|9 Ref:728565284-225.3 5G mobile network modelling methodology,inputs and assumptions This section discusses the methodology,inputs and assumptions used in our modelling of 5G mobile networks
77、and in our assessment of the environmental impact of additional mid-band spectrum assignment.In turn,we:outline the methodology(Section 3.1)describe the characteristics of the two scenarios modelled(Section 3.2)discuss the target user experience rate and parameters involved in calculating the requir
78、ed capacity to meet the target user experience,compared to the capacity available in the network in the urban and rural settlements modelled(Section 3.3)define the inputs required to determine the capacity of the 5G mobile network(Section 3.4)detail the data used to inform our environmental assessme
79、nt of the 5G mobile network(Section 3.5).3.1 Overview of approach The model we have developed provides an estimate of the environmental impacts of different deployment models and scenarios for 5G mobile network evolution from the present time,to meet future connectivity targets defined for 2030.As s
80、hown in Figure 3.1,we use a network model to estimate the sites required per square kilometre assuming the network evolves from todays deployment to meet the European connectivity targets described in the previous section,and then we calculate the associated carbon cost.Impact of additional mid-band
81、 spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|10 Ref:728565284-225 Figure 3.1:Overview of modelling approach Source:Analysys Mason,2023 CalculationInputOutputAvailable MBB capacity supply*Available MBB capacity supply*Mbit/s/macro site and small cell/yearReq
82、uired MBB Required MBB capacity*capacity*Mbit/s/km2/yearRequired macro site density(urban,Required macro site density(urban,rural town or rural town or village),outdoor village),outdoor small cell density(urban)small cell density(urban)/km2/yearMHz of spectrum MHz of spectrum per macroper macroSite
83、sectorisationSite sectorisationMHz of spectrum MHz of spectrum per small cellper small cellSpectral Spectral efficienciesefficienciesPopulation densityPopulation density2030 connectivity 2030 connectivity targetstargetsActivity factorActivity factorHighHigh-band band offloading(urban)offloading(urba
84、n)No.of outdoor No.of outdoor small cells relative small cells relative to macroto macroTotal urban/rural town or village Total urban/rural town or village embodied embodied&recurring carbon&recurring carbonkg CO2e/km2/yearUrban outdoor small cell total Urban outdoor small cell total embodied&recurr
85、ing carbonembodied&recurring carbonkg CO2e/site/yearUrban/rural town or Urban/rural town or village macro sites village macro sites passive equipment passive equipment embodied&recurring embodied&recurring carboncarbonkg CO2e/km2/yearUrban outdoor small cells Urban outdoor small cells passive equipm
86、ent passive equipment embodied embodied&recurring&recurring carboncarbonkg CO2e/km2/yearUrban/rural town or Urban/rural town or village macro sites village macro sites active equipment active equipment embodied&recurring embodied&recurring carboncarbonkg CO2e/km2/yearUrban outdoor small cells Urban
87、outdoor small cells active equipment active equipment embodied embodied&recurring&recurring carboncarbonkg CO2e/km2/yearHigh band High band coverage relative coverage relative to low bandsto low bands1 1Urban/rural town or village Urban/rural town or village macro sites total embodied¯o sites to
88、tal embodied&recurring carbon recurring carbon kg CO2e/site/year1For rural town or village deployment*Downlink and uplinkImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|11 Ref:728565284-225.In order to calculate the macro site and
89、outdoor small cell density required to meet the target user experience rate,the target is converted to a capacity density requirement.A per-user growth rate in Mbit/s is applied per year of the model,so that the target user experience rate to be achieved by 2030 is calculated.This required MBB servi
90、ce per active user in a busy hour is then multiplied by the dense urban population density and the proportion of active users(the activity factor)to arrive at a required capacity density.In the dense urban scenario,this capacity density is then adjusted to reflect offloading of traffic to high band,
91、to arrive at the required MBB uplink and downlink capacity density that has to be met by the macro sites and outdoor small cells.The model output of number of sites is then used to provide an estimate of annualised carbon costs in terms of:embodied costs that is,carbon emissions due to the raw mater
92、ial acquisition,manufacturing,distribution and installation of passive and active equipment at a 5G site,as well as construction of the required site infrastructure recurring costs that is,carbon emissions due to providing energy to operate and maintain the sites.Total embodied and recurring costs w
93、ill change depending on the deployment models and variants adopted by different mobile network operators(MNOs)to satisfy coverage or capacity requirements.For example,densification by building new macro sites to add capacity to a 5G mobile network creates embodied costs in the site building,together
94、 with increased recurring costs due to the increased number of sites in the network.Instead,MNOs might opt to build fewer new macro sites,and instead densify by building additional outdoor small cells.These small cells would provide less coverage than a macro site,but they would have lower embodied
95、costs per small cell.However,due to the coverage footprint being lower,more small cells would be needed.There are practical difficulties in building both macro sites and small cells,which are summarised in Section 4.3 later.Alternatively,MNOs could add spectrum to existing macro sites and/or small c
96、ells,assuming additional spectrum is available,and that existing sites can accommodate further deployment of radio equipment and antennas.Finally,MNOs could upgrade antenna systems and/or baseband processing to improve the spectral efficiency of their network,thus making better use of their availabl
97、e frequency portfolio.Accordingly,our modelling methodology is based on comparing two alternative deployment models:Additional mid-band spectrum is added to existing sites,in combination with a degree of densification(of both macro sites and small cells).We assume the additional mid-band spectrum(i.
98、e.the upper 6GHz band)is available for deployment from 2027.More-extensive network densification is carried out,due to no additional mid-band spectrum being made available for 5G mobile networks.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the uppe
99、r 6GHz band|12 Ref:728565284-225.The model is designed to compare the capacity supplied by a representative 5G network to the capacity needed to meet the future connectivity targets.The capacity supplied by the network is derived from assumptions on the density of macro sites in current 5G mobile ne
100、tworks,together with low-,mid-and high-band spectrum availability,spectral efficiency,number of sectors per cell,and various design assumptions within the mobile network that we summarise in the remainder of this section.The additional capacity needed in the network to meet the European connectivity
101、 targets for MBB and for broadband to fixed locations in 2030 is derived based on the connectivity targets,population and household densities,and end-user activity factors.We model two deployment scenarios:a dense urban area,and a rural town or village,as described further in Section 3.2 below.3.2 D
102、eployment scenarios We model two deployment scenarios:a dense urban area,defined as having a population density of 15 000/km 7 a rural town or village,defined as having a population density of at least 300/km,8 with an average of 2.8 people per household.9 This scenario assumes that the population l
103、ives outside the reach of fibre networks.The characteristics of 5G mobile networks vary in each case:Macro sites are more densely deployed in urban areas than in more rural areas,for two main reasons:due to the greater density of people,a tighter mesh is required to meet aggregated capacity demand t
104、he greater density of buildings and street furniture limits signal propagation and yields lower macro-site cell radii.Outdoor small cells are beginning to be deployed in dense urban areas as an alternative method of increasing capacity,as macro densification is becoming increasingly difficult in the
105、se areas for a number of reasons(including finding suitable real estate,obtaining planning permission,and inter-site interference).FWA is being deployed in rural towns and villages and other locations to provide fibre-like speeds where fibre deployment is uneconomic.FWA deployments can use high-band
106、 7 UN Habitat defines this as the optimum population density for a sustainable city 21,and this value sits fairly centrally in the range of cities looked at in the GSMA report 7 8 This is the population density of the rural settlement itself,not just the average population density of a largely uninh
107、abited rural area.We use the Eurostat definition of a“moderate density cluster”as having a population density of at least 300km 23.9 According to the OECD,average households range from two to four people depending on the country 24,whereas according to Irelands 2016 census urban areas had lower aver
108、age household sizes than more-rural locations,at 2.7 compared with 2.8 25.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|13 Ref:728565284-225.millimetre-wave(mmWave)spectrum to provide very high speeds,although the propagation of
109、high-band spectrum is worse than low-and mid-band spectrum,thus lowering the cell radius compared to that of a typical rural macro site.In order to model these different mobile network deployment characteristics we have defined several inputs,as described in Figure 3.2.Where N/A appears in the table
110、,this is because the values are not relevant to the analysis of a given deployment scenario.Figure 3.2:Summary of 5G mobile network deployment inputs and assumptions by settlement type Source:Analysys Mason,GSMA 7,2023 InputInput Dense uDense urban rban areasareas Rural towns and villagesRural towns
111、 and villages ValueValue SourceSource ValueValue SourceSource Macro-site inter-site distance(ISD)10 400m in 2022(corresponding to a site density of 7.2/km and a radius of 267m),which reduces over time as the model calculates increased macro-site density Analysys Mason assumption based on typical Eur
112、opean cities 3750m in 2022(corresponding to a site density of 0.08/km and a radius of 2500m),which reduces over time as the model calculates increased macro-site density Analysys Mason assumptions Small-cell radius 65m,which remains constant across the modelling period11 Huawei N/A N/A High-band rur
113、al cell radius N/A N/A 1000m in 2022,reducing linearly to 500m by 2030 as demand grows and radios prioritise throughput over coverage Analysys Mason assumptions 3.3 User experience data rate and required capacity The model calculates the 5G mobile network deployment required to meet a user experienc
114、e data rate in 2030 in line with the following targets(as explained in Section 2.1):100Mbit/s downlink and 50Mbit/s uplink for MBB 1Gbit/s downlink and 200Mbit/s uplink for FWA.10 We assume that the same ISD is also applied when additional mid-band spectrum is added.When modelling densification scen
115、arios via additional macro cells,the ISD reduces.11 Our assumptions have been designed to reflect that,in practice,outdoor small cells will be able to supply capacity in limited areas within the macro cells coverage area.However,the deployment of outdoor small cells will allow the macro cell to make
116、 an equivalent capacity supply available in other areas within the cell.For more information,see Section A.1.1.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|14 Ref:728565284-225.In line with a previous study conducted on behalf o
117、f the GSMA,entitled“Estimating the mid-band spectrum needs in the 20252030 timeframe”,we note that,whilst a particular data speed cannot be guaranteed by a mobile network,the assumption we make in our modelling is that the network is designed to target this user-experienced speed in the busy hour.Th
118、is implicitly means that a higher level of capacity is needed compared to a network designed to deliver a best-efforts service(which,for example,might have degraded speeds during the busy hour).7 Building upon these targets,we have used a number of further traffic-specific inputs and assumptions in
119、order to determine the networks required capacity.Inputs and assumptions for modelling MBB capacity demand In line with the GSMAs report,we have defined an activity factor,which is used to determine the maximum number of MBB users accessing the spectrum concurrently.This activity factor takes into a
120、ccount the variability of the user base(i.e.not all the same users access the spectrum at the same times of the day)and is linked to the busy-hour traffic of the network(i.e.the peak capacity requirement for the network).We have assumed a 5%activity factor,applied to the population of the modelled s
121、ettlement,consistent with typical usage profiles for smartphone users.Since this activity factor is a key dimensioning parameter of the model,we have also considered the impact of applying a higher activity factor(in the sensitivity analysis described in Section 4.1.3).In addition to this busy-hour
122、traffic,we have also assumed that MBB networks support traffic generated from outside-in coverage.This refers to a 10%increase in the required capacity to reflect that some devices within premises use MBB connectivity,with the traffic carried via base stations located outdoors.In the dense urban are
123、a,the high-band spectrum(mmWave)contribution to network capacity is not explicitly modelled,mainly due to mmWave deployments being concentrated within localised areas of high traffic demand but with more localised coverage compared to 5G mid-band and low-band spectrum.Rather,in the modelled dense ur
124、ban area,we assume that a portion of the required capacity is met by high-band spectrum being deployed on existing sites,which reduces the capacity to be met through additional mid-band spectrum and/or site densification.We assume that this high-band offloading increases linearly from 0%in 2022 to 1
125、0%in five years time,to reflect the progressive roll-out of mmWave small cells.Inputs and assumptions for modelling FWA capacity demand The analysis refers to a rural town or village where the population lives outside the reach of fibre networks and instead receives high-speed broadband connectivity
126、 via a 5G-based FWA service.It is assumed that there may be more concurrent FWA users in the busy hour than MBB users,mainly because fixed access usage is generally more continuous(for streaming,gaming,etc.)than mobile usage.As a result,a 10%FWA activity factor is assumed.Impact of additional mid-ba
127、nd spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|15 Ref:728565284-225.We note that not all households in the rural town or village will take up FWA.A sigmoidal penetration curve is assumed,with penetration of 80%of households being achieved by 2030(as shown i
128、n Figure 3.3).In the rural town or village,mmWave spectrum is modelled as part of the macro site spectrum portfolio,rather than as a traffic offload(which is used in the dense urban model).Figure 3.3:Assumed FWA penetration in a rural town or village receiving high-speed broadband connectivity via a
129、 5G-based FWA network Source:Analysys Mason,2023 3.4 Inputs for modelling MBB and FWA capacity supply Network capacity density estimation In order to calculate the macro-site and outdoor small-cell density required to meet a target user experience,the total network capacity density must be estimated
130、.To do so,the capacity that can be provided by each macro site and each small cell must be understood.This individual capacity is directly linked to the available spectrum per network operator(we assume a three-operator market).In our model we consider the following spectrum availability and timelin
131、es,which are consistent with typical spectrum availability in European markets(as summarised in Figure 3.4).190MHz of low-band spectrum(700900MHz)composed of:60MHz of 700MHz frequency-division duplexing(FDD)spectrum 60MHz of 800MHz FDD spectrum 70MHz of 900MHz FDD spectrum 0%20%40%60%80%202220232024
132、202520262027202820292030Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|16 Ref:728565284-225.535MHz of lower mid-band spectrum(1.52.6GHz)composed of:85MHz of 1500MHz supplementary downlink(SDL)spectrum(although only 40MHz is availa
133、ble until 2027)150MHz of 1800MHz FDD spectrum 120MHz of 2100MHz FDD spectrum 140MHz of 2600MHz FDD spectrum 40MHz of 2600MHz time-division duplexing(TDD)spectrum 400MHz of upper mid-band TDD spectrum(3.43.8GHz)700MHz of additional upper mid-band TDD spectrum(6.4257.125GHz)only available from 202712
134、2400MHz of high-band TDD spectrum(25.127.5GHz)from 2023.13 The model considers all available spectrum for public mobile use in a European country,assuming that all operators in a given market operate their own networks.The spectrum availability for public mobile use in European markets is shown in F
135、igure 3.4 below.Figure 3.4:Maximum spectrum bandwidth available for public mobile use in European markets Source:Analysys Mason,2023 SpectrumSpectrum Bandwidth available(MHz)Bandwidth available(MHz)20222022 20232023 20242024 20252025 20262026 20272027 20282028 20292029 20302030 Low bands 190 190 190
136、 190 190 190 190 190 190 Lower mid-bands FDD 410 410 410 410 410 410 410 410 410 Lower mid-bands SDL 40 40 40 40 40 85 85 85 85 Lower mid-bands TDD 40 40 40 40 40 40 40 40 40 Upper mid-bands 400 400 400 400 400 400 400 400 400 Additional upper mid-bands -700 700 700 700 High bands -2400 2400 2400 24
137、00 2400 2400 2400 2400 12 The ITUs World Radiocommunication Conference 2023(WRC-23)will discuss various frequency bands for IMT use in its agenda item 1.2,including upper 6GHz in ITU Region 1.Depending on the WRC decision,the CEPT European Communications Committee(ECC)might develop a new harmonisati
138、on decision(s)concerning use of the 64257125MHz band.Allowing time for this to occur,and for devices to become available,we have assumed a 2027 date for the upper 6GHz band deployment.13 While the 26GHz band is harmonised in Europe from 24.2527.5GHz,this report assumes that the lower part of the ban
139、d is assigned for lower-power deployments relying on local area assignments.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|17 Ref:728565284-225.In order to determine the uplink and downlink bandwidth the following assumptions are
140、made:The use of FDD spectrum is assumed to be split into 50%downlink and 50%uplink The use of TDD spectrum is assumed to be 75%downlink and 25%uplink The use of SDL spectrum is assumed to be 100%downlink.While Figure 3.4 summarises the availability of all spectrum bands over time,not all site types
141、across all geographies are assumed to use all available spectrum.Our model considers different types of site(tri-sectored macro sites vs.single-sector small cells)and uses(MBB or FWA),each with its own individual spectrum portfolio,as shown in Figure 3.5.Figure 3.5:Summary of spectrum availability b
142、y site and traffic type Source:Analysys Mason,2023 SpectrumSpectrum Available for Available for MBB on urban MBB on urban macro site?macro site?AvailAvailable for able for MBB on urban MBB on urban small cell?small cell?Available for Available for MBB on ruralMBB on rural town or village town or vil
143、lage macro site?macro site?Available for Available for FWA on rural FWA on rural town or village town or village macro site?macro site?Low-band(700900MHz)13 1 Lower mid-band FDD(1.52.6GHz)Upper mid-band(3.43.8GHz)Additional upper mid-band(6.4257.125GHz)High-band(25.127.5GHz)2 We assume that typicall
144、y only one low band will be used for FWA,to ensure sufficient capacity in the mobile network for wide-area MBB traffic.Since there are three low-bands in our model,we assume that 1/3 of those,i.e.one band,is available for FWA use As explained in Section 3.3,the high band is not explicitly modelled i
145、n dense urban areas,but rather a high-band offloading factor is used.Spectrum roll-out assumptions,and time period of the model Figure 3.4 highlights the points in time when we assume new spectrum bands become available for use.We assume an operator will generally deploy new spectrum onto existing s
146、ites progressively across its network,progressing from dense urban areas(where capacity requirements are more stringent)to less dense rural areas,in line with customer needs,and reflecting any regulatory requirements(e.g.roll-out obligations in a spectrum licence)as well as market competition.This p
147、rogressive roll-out reflects operational and economical constraints:while spectrum may be available nationally,it is not necessarily deployed immediately across all the sites in a national network.To account for this progressive roll-out of new frequency bands,two spectrum deployment profiles are mo
148、delled for each new band one for each settlement type(see Figure 3.6 and Figure 3.7).Since it was only recently made available,the upper mid-band(3.43.8GHz)is included in these progressive roll-out profiles.It is assumed that the additional lower mid-band SDL spectrum can be deployed on all sites wi
149、th an existing 1500MHz deployment,so no roll-out profile is shown for this Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|18 Ref:728565284-225.band.Because high-band spectrum is not explicitly being modelled for urban areas,it is
150、not included in Figure 3.6.The spectrum roll-out in our model is completed by 2030 and hence the charts below show the profiles up to the final roll-out in 2030.This final roll-out then applies thereafter,until 2032.The rationale for the spectrum roll-out is as follows:We assume that operators have
151、deployed existing upper mid-band spectrum(e.g.3.5GHz)on all dense urban sites in the initial stages of 5G roll-out and hence the upper mid-band roll-out in the dense urban case is 100%throughout the model.In the rural towns and villages,we assume that operators are still completing 3.5GHz roll-out a
152、nd hence the upper mid-band roll-out to rural town or village macro sites increases from 2022 until roll-out is complete in 2026.We assume that additional upper mid-band spectrum becomes available from 2027(see Figure 3.4),and is progressively rolled out to sites in both the dense urban and rural to
153、wn or village scenario.Device roll-out assumptions together with site roll-out assumptions mean that the additional upper mid-band is progressively added to sites from 2027 to 2030.The high-band roll-out in the rural town or village refers to mmWave spectrum(see Figure 3.4).In the dense urban modell
154、ing scenario we do not explicitly model the mmWave roll-out profile,but instead assume that a portion of the required capacity is met through mmWave.We assume that 20%of rural town or village sites are not suitable for mmWave deployment,and thus roll-out stops at 80%by 2026.Figure 3.6:New mid-band s
155、pectrum roll-out on dense urban macro sites and small cells Source:Analysys Mason,2023 Figure 3.7:New mid-band and high-band spectrum roll-out on rural town or village macro sites Source:Analysys Mason,2023 0%20%40%60%80%100%20222024202620282030Proportion of sitesUpper mid-bandsAdditional upper mid-
156、bands0%20%40%60%80%100%20222024202620282030Proportion of sitesUpper mid-bandsAdditional upper mid-bandsHigh bandsImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|19 Ref:728565284-225.In addition,we assume that newly assigned spectru
157、m becomes useable once compatible devices are available in sufficient volumes,and that the penetration of devices with new upper mid-band spectrum included will increase over time(see Figure 3.8 below).MBB device penetration growth is assumed to be the same in both urban and rural settlements.As upp
158、er mid-band and high-band are already available when FWA deployment begins,we assume that all FWA consumer premises equipment(CPE)is compatible with these bands.However,some CPE will be installed before the additional upper mid-band spectrum is available,and this will only be gradually upgraded to b
159、e compatible with the additional upper mid-band spectrum,in line with the MBB device take-up curve.Any new CPE installed after the additional upper mid-band spectrum is available will be compatible.The device penetration curves shown below are Analysys Mason estimates,guided both by our understandin
160、g of historical device take-up trends and by current 5G device types.For example,we assume that devices compatible with mmWave will not reach the same levels of penetration as those compatible with upper mid-band,due to a range of factors such as current availability and device cost.We assume that a
161、dditional upper mid-band spectrum might be added to existing upper mid-band devices,with penetration levels rising more rapidly,assuming that a European harmonisation decision is in place around 2026.Figure 3.8:Penetration of MBB devices compatible with upper mid-band,additional upper mid-band and h
162、igh-band Source:Analysys Mason,2023 Note:High-band device penetration is only used in our rural settlement scenario 0%20%40%60%80%100%202220232024202520262027202820292030Upper mid-bandsAdditional upper mid-bandsHigh bandsImpact of additional mid-band spectrum on the carbon footprint of 5G mobile net
163、works:the case of the upper 6GHz band|20 Ref:728565284-225.The available capacity of the macro and outdoor small cell sites modelled in the network are calculated from the spectrum bandwidths and spectral efficiencies.Spectral efficiencies vary between downlink and uplink,and with site type/use,but
164、not by settlement type(see Figure 3.9,Figure 3.10 and Figure 3.11).Spectral efficiency improvement can come from various sources such as technology improvements,antenna design,network design improvements or other changes affecting the transmitted waveform.Figure 3.9:MBB macro-site spectral efficienc
165、ies Source:Analysys Mason,GSMA 7,14 Huawei,2023 SpectrumSpectrum Downlink spectral Downlink spectral efficiency(bit/s/Hz)efficiency(bit/s/Hz)Uplink spectral Uplink spectral efficiency(bit/s/Hz)efficiency(bit/s/Hz)20222022 2032030 0 20222022 2032030 0 Low bands(700900MHz)*1.87 1.87 1.03 1.23 Lower mi
166、d-bands(1.52.6GHz)FDD*1.87 3.50 1.03 1.68 Lower mid-bands(1.52.6GHz)SDL*1.87 3.50 N/A N/A Lower mid-bands(1.52.6GHz)TDD*2.34 3.28 1.05 1.05 Upper mid-bands(3.43.8GHz)5.01 7.15 3.31 4.73 Additional upper mid-bands(6.4257.125GHz)5.51 7.87 3.64 5.20 High bands(25.127.5GHz)*3.10 4.65 1.50 2.25*Non-activ
167、e antenna system(AAS)base stations are assumed for frequencies below 3.4GHz.*Only applicable to rural macro-site modelling,as described in Section 3.3.Figure 3.10:Urban small-cell spectral efficiencies15 Source:Huawei,Analysys Mason,2023 SpectrumSpectrum Downlink spectral Downlink spectral efficienc
168、y(bit/s/Hz)efficiency(bit/s/Hz)Uplink spectral Uplink spectral efficiency(bit/s/Hz)efficiency(bit/s/Hz)20222022 2032030 0 20222022 2032030 0 Upper mid-bands(3.43.8GHz)1.67 2.38 1.10 1.58 Additional upper mid-bands(6.4257.125GHz)1.84 2.62 1.21 1.73 Figure 3.11:FWA macro-site spectral efficiencies Sou
169、rce:Huawei,Analysys Mason,2023 SpectrumSpectrum Downlink Downlink spectral spectral efficiency(bit/s/Hz)efficiency(bit/s/Hz)Uplink spectral Uplink spectral efficiency(bit/s/Hz)efficiency(bit/s/Hz)20222022 2032030 0 20222022 2032030 0 Low band(e.g.900MHz)2.06 2.20 1.13 1.45 Upper mid-bands(3.43.8GHz)
170、5.51 8.40 3.64 5.56 Additional upper mid-bands(6.4257.125GHz)6.06 9.24 4.01 6.11 High bands(25.127.5GHz)4.65 6.98 2.25 3.38 14 Note that the lower values for lower mid-bands compared to those for upper mid-bands are because a non-active antenna system base-station deployment is being assumed for low
171、er mid-bands.15 Improvements in the spectral efficiency of urban small cells are assumed,to reflect advances in technology and antennas leading to improved performance.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|21 Ref:72856528
172、4-225.In addition,a 15%design margin is applied to the capacity(i.e.only 85%of the capacity can actually be achieved)to account for the reality that in practice capacity at a site/cell in the busy period cannot be fully utilised.This approach is consistent with that used in the GSMA report on estima
173、ting mid-band spectrum needs 7 referred to previously.The model makes assumptions about the number of small cells deployed per macro cell in the urban deployment.Given that individual MNOs will make different deployment choices depending on their network strategy,customer needs and other factors ass
174、ociated with the density of their existing network grids,we model two deployment variants(densification focused on macro cells,and densification focused on small cells),described in Sections 4.1.1 and 4.1.2.3.5 Inputs and assumptions for modelling the environmental impact Figure 3.12 details the env
175、ironmental inputs used in the model.Environmental inputs have been derived from public sources 7 8 9 10 11 12,but as far as possible they have been verified as reasonable by Huawei.In some cases estimations have been made(e.g.for future additional upper mid-band equipment).As noted in Section 3.1 ea
176、rlier there are two types of carbon cost embodied and recurring.Embodied carbon costs are incurred once per lifetime,and they have been amortised over that lifetime.In contrast,recurring carbon costs are incurred annually.To calculate the required infrastructure,we have assumed that the mid-band,upp
177、er mid-band and high-band spectrum is deployed on a proportion of sites.For these sites,there is therefore an additional fixed recurring carbon cost associated with the relevant active equipment being used,and there is an incremental recurring carbon cost that varies depending on network loading(bot
178、h across the day,which has been accounted for in the input value,and in terms of users who can use the spectrum represented by the compatible device penetration).In our calculation of recurring costs we assume a carbon intensity(a measure of carbon dioxide and other greenhouse gases in energy produc
179、tion)for electricity used by the telecoms sector of 100g/kWh,as we believe that MNOs use a more-green energy mix than that of the typical grid in Europe(currently 250g/kWh)and that there will be some improvements in emissions by 2030.13 The model calculates the number of macro sites and small cells
180、required,on the assumption all spectrum is utilised per site.However,because we consider a three-operator market,in reality each calculated macro site would comprise three macro sites one per operator sharing the available spectrum.As such,the environmental inputs shown in Figure 3.12 are the sum fo
181、r three operators each using a third of the total spectrum.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|22 Ref:728565284-225.Figure 3.12:Summary of environmental inputs Source:Huawei,Analysys Mason,2023 InputInput UrbanUrban mac
182、ro sitemacro site UrbanUrban small cellsmall cell RuralRural macro sitemacro site Baseline site(excluding upper mid-bands,additional upper mid-bands and high bands)passive equipment embodied carbon 150 000kg CO2e 6000kg CO2e 190 000kg CO2e Baseline site(excluding upper mid-bands,additional upper mid
183、-bands and high bands)active equipment embodied carbon 12 000kg CO2e N/A 12 000kg CO2e Upper mid-band passive and active equipment embodied carbon 9000kg CO2e 900kg CO2e 9000kg CO2e Additional upper mid-band passive and active equipment embodied carbon 9000kg CO2e 900kg CO2e 9000kg CO2e High-band pa
184、ssive and active equipment embodied carbon N/A N/A 4500kg CO2e Lifetime of passive equipment 20 years 20 years 20 years Lifetime of active equipment 8 years 8 years 8 years Baseline site(excluding upper mid-bands,additional upper mid-bands and high bands)recurring carbon cost 12 000kg CO2e/year 150k
185、g CO2e/year 12 000kg CO2e/year Upper mid-band recurring carbon cost fixed component(irrespective of loading)4000kg CO2e/year 250kg CO2e/year 4000kg CO2e/year Upper mid-band recurring carbon cost variable component(loading dependent)2000kg CO2e/year 150kg CO2e/year 2000kg CO2e/year Additional upper m
186、id-band recurring carbon cost fixed component(irrespective of loading)4000kg CO2e/year 250kg CO2e/year 4000kg CO2e/year Additional upper mid-band recurring carbon cost variable component(loading dependent)2000kg CO2e/year 150kg CO2e/year 2000kg CO2e/year High-band recurring carbon cost fixed compone
187、nt(irrespective of loading)N/A N/A 2000kg CO2e/year High-band recurring carbon cost variable component(loading dependent)N/A N/A 1000kg CO2e/year Annex A provides further details of the methodology,including descriptions of where each input is used.Impact of additional mid-band spectrum on the carbo
188、n footprint of 5G mobile networks:the case of the upper 6GHz band|23 Ref:728565284-225.4 5G mobile network modelling carbon footprint results This section presents results from the mobile network modelling for the urban and rural settlements(in Sections 4.1 and 4.2 respectively.We then go on to disc
189、uss some of the practical issues impacting mobile network densification of the sort modelled in this study,in Section 4.3.Finally we consider how technological improvements are likely to improve the power efficiency of mobile networks as technologies evolve,which is discussed in Section 4.4.4.1 Dens
190、e urban area results Figure 4.1 shows the progression of the modelled uplink and downlink MBB capacity target for the urban settlement over time.The required downlink MBB capacity reaches nearly 75Gbit/s/km by 2030.Figure 4.1:Dense urban MBB required capacity Source:Analysys Mason,2023 In comparison
191、,Figure 4.2 and Figure 4.3 show the dense urban MBB capacity supply per average macro site and small cell respectively.In both figures it can be seen that the capacity provided increases gradually until 2026 as spectral efficiencies increase,and upper mid-band device take-up increases.However,from 2
192、027,the introduction of additional upper-mid band spectrum brings a significant increase in capacity as the new spectrum is rolled out on sites and compatible device penetration increases.202220232024202520262027202820292030020406080Gbit/s/km2DownlinkUplinkImpact of additional mid-band spectrum on t
193、he carbon footprint of 5G mobile networks:the case of the upper 6GHz band|24 Ref:728565284-225.Figure 4.2:Dense urban MBB capacity supply per average macro site Source:Analysys Mason,2023 Figure 4.3:Dense urban MBB capacity supply per average small cell Source:Analysys Mason,2023 As discussed below,
194、we have modelled two dense urban deployment variants,the first with a maximum ratio of three small cells per macro site(in line with the GSMA report),and the second with a limit on macro-site densification.4.1.1 Deployment variant 1 densification primarily via macro cells Figure 4.4 and Figure 4.5 s
195、how the macro-site and outdoor small-cell density,respectively,required when the level of small-cell densification is limited to three small cells per macro site.With additional upper mid-band spectrum no macro-site densification is required(as shown on Figure 4.4,macro-site density remains constant
196、 at 7.2/km).As shown on Figure 4.5,at 18.6 small cells/km the outdoor small-cell densification does not reach the limit of three small cells per macro site;there is a gradual increase in small cells until the new spectrum is deployed,at which point small-cell deployment plateaus for several years wh
197、ile the additional capacity required is met by the additional spectrum,and after this small-cell deployment resumes.However,without additional upper mid-band spectrum,the small-cell densification cap results in significant macro-site densification in later years in order to provide the required capa
198、city:as shown 2022202420262028203005101520Gbit/sDownlink capacity withoutadditional upper mid-bandsDownlink capacity withadditional upper mid-bandsUplink capacity withoutadditional upper mid-bandsUplink capacity withadditional upper mid-bands202220242026202820300.00.51.01.52.0Gbit/sDownlink capacity
199、 withoutadditional upper mid-bandsDownlink capacity withadditional upper mid-bandsUplink capacity withoutadditional upper mid-bandsUplink capacity withadditional upper mid-bandsImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|25 Ref
200、:728565284-225.on Figure 4.4,it increases by 83%by 2030,to reach 13.2 macro sites/km(corresponding to an ISD of 296m).This macro-site densification is in addition to the maximum allowance of small cells(three per macro site),reaching 39.6 small cells/km by 2032(as shown in Figure 4.5).Figure 4.4:Den
201、sity of dense urban macro sites Source:Analysys Mason,2023 Figure 4.5:Density of dense urban outdoor small cells Source:Analysys Mason,2023 With the assumed downlink to uplink spectrum ratios,it is the case that in order to meet the uplink target,the modelled capacity required is such that the downl
202、ink target for the deployment is substantially exceeded.Downlink and uplink capacity assumptions for upper mid-band and additional upper-mid band spectrum are based on currently adopted downlink-centric frame structures16 in 5G MBB networks.It is noted that alternative frame structures might be adop
203、ted in the future as a wider variety of services with bespoke connectivity requirements become more widely available.Figure 4.6 shows the annual carbon savings achieved by using additional upper mid-bands for IMT-2020/5G(i.e.the amount by which annual carbon footprints are reduced relative to a scen
204、ario without additional upper mid-bands)by category per urban km.By 2030,the dense urban networks carbon footprint is 131tCO2e17/km lower with additional upper mid-bands than without.Figure 4.6 also shows that the carbon savings from having less densification in 5G mobile networks outweigh the incre
205、mental carbon cost of deploying and operating additional mid-band spectrum at existing macro(and small)cells.16 Frame structure in a 5G mobile network refers to TDD and FDD transmissions variously used in different frequency bands used for 5G networks.The frame structure for 5G mid-band spectrum is
206、assumed to be based on TDD transmission.In TDD transmission,uplinks and downlinks operate in the same frequency channel at different points in time.The frame structure defines the time allocated to the downlink compared to the uplink.We assume a downlink-oriented frame structure in which downlink tr
207、ansmissions occupy 75%of the available bandwidth and uplink transmissions occupy 25%.17 Tonnes of carbon dioxide equivalent.054202620282030Macro sites per km2Without additional upper mid-bandsWith additional upper mid-bands022024202620282030Outdoor small cells per km2Without ad
208、ditional upper mid-bandsWith additional upper mid-bandsImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|26 Ref:728565284-225.It is noted that the carbon savings will continue beyond 2030.The precise level of these savings is not cal
209、culated in our model,due to uncertainty over site build after 2030,and the potential introduction of new mobile access technologies(e.g.6G)for which performance is as yet undefined.In Figure 4.6 we show what the carbon savings would be in 2031 and 2032 assuming the same level of savings in those yea
210、rs as calculated in our model for 2030.Figure 4.7 shows the cumulative carbon savings per urban km,reaching 475tCO2e/km by 2032(on the assumption that annual savings in 2031 and 2032 are the same as in 2030).Figure 4.6:Annual carbon savings per urban km Source:Analysys Mason,2023 Figure 4.7:Cumulati
211、ve carbon savings per urban km Source:Analysys Mason,2023 Figure 4.8 and Figure 4.9 illustrate the magnitude of the savings per urban inhabitant,with reference to the modelled urban scenario described in this report(i.e.a population of 15 000 per km).In 2030,9kgCO2e is saved.-50050100150200tCO2e/km2
212、2027202820292030201131Macro site embodied carbonMacro site recurring carbon costMacro site recurring carbon cost upper-mid bandsMacro site recurring carbon cost additional upper-mid bandsSmall cell embodied carbonSmall cell recurring carbon cost2027tCO2e/km4420304752
213、0321982213Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|27 Ref:728565284-225.Figure 4.8:Annual carbon savings per dense urban inhabitant Source:Analysys Mason,2023 Figure 4.9:Cumulative carbon savings per dense urban inhabitant S
214、ource:Analysys Mason,2023 4.1.2 Deployment variant 2 densification using a higher proportion of small cells In the second urban deployment variant,macro-site density is capped at 9.4/km,an increase of 31%relative to current macro-site density(corresponding to an ISD of 350m),as shown in Figure 4.10.
215、Figure 4.11 shows the resulting substantial deployment of small cells required to meet the target exceeding 100 small cells/km.-5051015kg CO2e/inhabitant2027 2028 2029 2030 2031 2032014999Macro site embodied carbonMacro site recurring carbon costMacro site recurring carbon cost upper-mid bandsMacro
216、site recurring carbon cost additional upper-mid bandsSmall cell embodied carbonSmall cell recurring carbon cost01kg CO2e/inhabitant20272029202820322030Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|28 Ref:728565284-225.
217、Figure 4.10:Density of dense urban macro sites Source:Analysys Mason,2023 Figure 4.11:Density of dense urban outdoor small cells Source:Analysys Mason,2023 As in deployment variant 1,it is the case that in order to meet the uplink target,the modelled capacity required is such that the downlink targe
218、t for the deployment is substantially exceeded.Figure 4.12 to Figure 4.15 show there are still carbon savings from using additional upper mid-bands for IMT-2020/5G under deployment variant 2,but they are around 30%lower than under deployment variant 1.0246802620282030Macro sites per km2Wi
219、thout additional upper mid-bandsWith additional upper mid-bands020406080024202620282030Outdoor small cells per km2Without additional upper mid-bandsWith additional upper mid-bandsImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GH
220、z band|29 Ref:728565284-225.Figure 4.12:Annual carbon savings per dense urban km Source:Analysys Mason,2023 Figure 4.13:Cumulative carbon savings per dense urban km Source:Analysys Mason,2023 -50050100150tCO2e/km22027202820292030200Macro site embodied carbonMacro site recurring carb
221、on costMacro site recurring carbon cost upper-mid bandsMacro site recurring carbon cost additional upper-mid bandsSmall cell embodied carbonSmall cell recurring carbon costtCO2e/km2202920322027242133313Impact of additional mid-band spectrum on the carbon footprint of 5G mobile netwo
222、rks:the case of the upper 6GHz band|30 Ref:728565284-225.Figure 4.14:Annual carbon savings per dense urban inhabitant Source:Analysys Mason,2023 Figure 4.15:Cumulative carbon savings per dense urban inhabitant Source:Analysys Mason,2023 4.1.3 Sensitivity analysis for the dense urban scenario Figure
223、4.12 and Figure 4.13 show how the cumulative carbon savings to 2030 vary with activity factor and high-band offloading for dense urban deployment variants 1 and 2 respectively.As expected,a higher activity factor increases the carbon savings.This is because the difference between the volume of site
224、build required with and without additional 5G upper mid-band spectrum is exacerbated as demand on the network increases.Similarly,as high-band offloading increases the carbon savings decrease.This is because there is less demand on the macro sites and small cells,so less densification is required an
225、d thus the impact of additional spectrum in terms of mitigating the need for additional site build is reduced.However,we note that high-band offloading is only possible if sites at those specific locations where demand is highest can accommodate high-band deployment,and/or where new mmWave sites can
226、 be deployed.It is also noted that the mmWave coverage is lower than the coverage achieved from either 5G mid-bands or from other,lower,frequency bands used in 5G mobile networks.Hence,mmWave offloading can complement 5G mid-band deployment but does not provide a direct substitute for it due to cove
227、rage differences.-50510kg CO2e/inhabitant2027 2028 2029 2030 2031 2032012666Macro site embodied carbonMacro site recurring carbon costMacro site recurring carbon cost upper-mid bandsMacro site recurring carbon cost additional upper-mid bandsSmall cell embodied carbonSmall cell recurring carbon cost0
228、03912027kg CO2e/inhabitant202820302029Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|31 Ref:728565284-225.Figure 4.16:Cumulative carbon savings to 2030 activity factor and high-band offloading sensitivity on dense urban
229、 deployment variant 1(data point used in Section 4.1.1 highlighted)Source:Analysys Mason,2023 22%24%26%02004006008004%120014000%2%6%8%10%12%14%16%18%20%1000tCO2e/km2213213Activity factor0%high band offloading5%high band offloading10%high band offloading15%high band offloading20%high band offloading2
230、5%high band offloading30%high band offloading35%high band offloading40%high band offloading45%high band offloadingImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|32 Ref:728565284-225.Figure 4.17:Cumulative carbon savings to 2030 ac
231、tivity factor and high-band offloading sensitivity on dense urban deployment variant 2(data point used in Section 4.1.2 highlighted)Source:Analysys Mason,2023 4.2 Rural town or village results Figure 4.18 shows the required uplink and downlink MBB capacity for the rural town or village over time.Lik
232、ewise,Figure 4.19 shows the FWA required capacity,which accounts for the majority of the total rural required capacity.The required capacity for MBB is calculated based on the assumed population density for the rural town or village,which is 300/km.The required capacity for FWA assumes households in
233、 the town or village are beyond the reach of a fixed(fibre)network.We note that not all households will take up a broadband service and hence we assume a penetration of 80%by 2030.14%2%0%4%8%06%10%12%16%18%20%22%24%60026%2004008000tCO2e/km2Activity factor1335%high band offloading0%high ba
234、nd offloading15%high band offloading10%high band offloading20%high band offloading45%high band offloading25%high band offloading30%high band offloading35%high band offloading40%high band offloadingImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the up
235、per 6GHz band|33 Ref:728565284-225.Figure 4.18:MBB required capacity for rural town or village Source:Analysys Mason,2023 Figure 4.19:FWA required capacity for rural town or village Source:Analysys Mason,2023 In comparison,Figure 4.20 and Figure 4.21 show the MBB and FWA capacity supply per macro si
236、te in a rural town or village respectively.Both are fairly similar(there are slight variations due to the different spectrum portfolios and spectral efficiencies),but in reality,rural macro sites will use a combination of MBB and FWA,rather than MBB-only or FWA-only as illustrated here.2022202420262
237、02820300.00.51.01.5Gbit/s/km2DownlinkUplink202220242026202820300246810Gbit/s/km2DownlinkUplinkImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|34 Ref:728565284-225.Figure 4.20:Maximum MBB capacity supply per macro site for rural tow
238、n or village Source:Analysys Mason,2023 Figure 4.21:Maximum FWA capacity supply per macro site for rural town or village Source:Analysys Mason,2023 In order to meet the total required capacity,macro-site densification is required(see Figure 4.22).With additional upper mid-bands rural macro-site dens
239、ity increases to 0.4/km by 2030(an ISD of 1.6km).In comparison,without additional upper mid-band spectrum,rural macro sites reach twice that density,to reach 0.8/km(an ISD of 1.2km).202220242026202820300510152025Gbit/sDownlink capacity withoutadditional upper mid-bandsDownlink capacity withadditiona
240、l upper mid-bandsUplink capacity withoutadditional upper mid-bandsUplink capacity withadditional upper mid-bands20222024202620282030051015202530Gbit/sDownlink capacity withoutadditional upper mid-bandsDownlink capacity withadditional upper mid-bandsUplink capacity withoutadditional upper mid-bandsUp
241、link capacity withadditional upper mid-bandsImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|35 Ref:728565284-225.Figure 4.22:Density of macro sites for rural towns or villages Source:Analysys Mason,2023 As in the dense urban scenar
242、io the downlink target is exceeded;the uplink target is the limiting factor.Figure 4.23 and Figure 4.24 show the annual and cumulative carbon savings achieved by using additional upper mid-bands for IMT-2020/5G.It is noted that the carbon savings will continue from 2030.However,the precise level of
243、these savings is not calculated in our model,due to uncertainty over site build after 2030,and the introduction of new mobile access technologies(e.g.6G)for which performance is as yet undefined.In Figure 4.24 we show what the carbon savings would be in 2031 and 2032 assuming the same level of savin
244、gs in those years as calculated in our model for 2030.0.00.20.40.60.81.0202220232024202520262027202820292030Macro sites per km2Without additional upper mid-bandsWith additional upper mid-bandsImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6
245、GHz band|36 Ref:728565284-225.Figure 4.23:Annual carbon savings per rural town or village km Source:Analysys Mason,2023 Figure 4.24:Cumulative carbon savings per rural town or village km Source:Analysys Mason,2023 While from a network point of view these numbers are much lower than the equivalent va
246、lues in the dense urban scenario,the carbon savings per inhabitant are significantly greater for rural towns or villages(see Figure 4.25 and Figure 4.26).The annual savings per inhabitant are at least three times higher than for a dense urban inhabitant(30kgCO2e in 2030 in the rural scenario compare
247、d with 9kgCO2e and 6kgCO2e for urban deployment variants 1 and 2 respectively).-2.50.02.55.07.510.012.5tCO2e/km22027202820292030299Macro site embodied carbonMacro site recurring carbon costMacro site recurring carbon cost upper-mid bandsMacro site recurring carbon cost additional upper-mi
248、d bandsMacro site recurring carbon cost high bands2030tCO2e/km220312027 20280192838Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|37 Ref:728565284-225.Figure 4.25:Annual carbon savings per rural town or village inhabita
249、nt Source:Analysys Mason,2023 Figure 4.26:Cumulative carbon savings per rural town or village inhabitant Source:Analysys Mason,2023 4.3 Practical issues in building additional macro and small cells Whilst densifying either macro sites or small cells enables operators to meet increased demand,both op
250、tions present practical issues in their design,implementation and cost.In terms of macro-site densification,we summarise below examples of the key practical issues:Suitable macro-site locations in urban areas are increasingly hard to find few locations that would be suitable for improving existing c
251、overage have sufficient physical space planning issues can delay site acquisition densification may be considered unsightly or the cause of additional electromagnetic radiation,leading to public resistance As macro-site density increases,so too does inter-site interference,which reduces the effectiv
252、e site capacity.-kg CO2e/inhabitant2027 2028 2029 2030 2031 203241119303030Macro site embodied carbonMacro site recurring carbon costMacro site recurring carbon cost upper-mid bandsMacro site recurring carbon cost additional upper-mid bandsMacro site recurring carbon cost high bands415346
253、49520322031kg CO2e/inhabitant2027 2028Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|38 Ref:728565284-225.In terms of small-cell densification,we summarise below examples of the key practical issues:Small cells provide
254、lower coverage compared to macro sites,and so many more small cells would be needed to meet future traffic demand,but there may be public resistance to such proliferation Identifying enough suitable locations for small cells(which are typically installed on urban furniture or building facades,rather
255、 than rooftops)may be difficult Local authority co-ordination/planning issues might affect deployment timescales,or even influence whether sites are viable or not,due to the time taken for planning issues associated new sites and/or modifications to existing sites to be authorised Small-cell costs c
256、an be high relative to the capacity provided,leading to a potentially unsustainable network deployment model as the number of small cells increases.4.4 Power efficiency measures in mobile network architectures As mobile networks expand and user demands grow,energy efficiency and energy saving featur
257、es will become even more important to optimise power consumption as networks are densified,and further spectrum is added.Improved technology capabilities to manage power consumption and energy efficiency are an important aspect of how 5G mobile network equipment vendors are evolving their systems to
258、 limit the carbon footprint of their clients networks.Whist not an exhaustive list,the following are examples of technology features that might be used in 5G mobile networks to manage power consumption and energy efficiency(noting that these features are vendor specific,we include an illustrative li
259、st below with specific vendor solutions indicated in the footnote18):Transmission on demand for example,dynamic onoff functions in antennas and radio frequency(RF)chains.These dynamic onoff features can be assisted by artificial intelligence(AI)and/or user feedback to anticipate user and traffic pat
260、terns and to enable real-time network optimisation(e.g.by analysing historical traffic patterns,busy hours,cell conditions and user needs,and using these in intelligent network optimisation)More-efficient power control in networks and devices,adjusting radiated power to actual coverage needs Innovat
261、ive solutions for massive MIMO19 antennas to increase deployment efficiency and lower power consumption(e.g.through novel optimisation algorithms)20 18 For example,see https:/ and https:/ MIMO refers to multiple-input,multiple-output,which is the use of multiple transmission and receiving antennas o
262、ver the same radio channel.A key effect of massive MIMO in 5G is that the capacity of the wireless connection is significantly improved 20 For example,see https:/ Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|39 Ref:728565284-225
263、.Real-time capture and evaluation of network key performance indicators(KPIs)as an input to network planning,rather than more traditional methods for non-real-time capture of KPIs(e.g.through drive testing).Other non-technology-related approaches to reduce the carbon footprint of 5G mobile networks
264、may include:Use of complementary green energy power sources(e.g.solar or wind power)to limit the use of grid power and replenish back-up batteries Power efficiency through spectrum assignment,such as considering the most efficient way to assign spectrum for new mobile technologies(e.g.energy efficie
265、ncy from wider contiguous carriers rather than aggregation of narrower carriers).Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|40 Ref:728565284-225.5 Relationship between Wi-Fi carbon footprint and the 6GHz band 5.1 Introduction
266、Wi-Fi technology is overwhelmingly used for indoor traffic,within homes,public locations or businesses,where the premises are connected to high-speed fixed broadband access networks(generally fibre,but also wireless links or legacy copper technologies)and where Wi-Fi is used to provide high-speed wi
267、reless connectivity throughout the premises.As the market penetration and performance of fixed broadband networks is increasing,with comprehensive fibre roll-out and the progressive introduction of high-speed business-to-consumer(B2C)networks,there is,in turn,a reliance on the evolution of Wi-Fi tec
268、hnologies to deliver the required speed increases within homes and premises.In Section 4 we assessed the impact on the carbon emissions associated with mobile networks for the scenario in which the upper 6GHz band was used for 5G,as well as the alternative scenario in which the band would not be mad
269、e available for 5G.In this section we assess the impact on carbon emissions associated with Wi-Fi for the scenario in which the upper 6GHz band is used for WLANs,such as Wi-Fi,as well as the alternative scenario in which the band is not made available for WLANs.The combination of results from Sectio
270、n 4 and from this section allows us to assess the overall impact on carbon emissions associated with 5G mobile networks and Wi-Fi in the case where the upper 6GHz band is made available for 5G,and the case where the band is made available for WLANs.Similarly to the analysis presented in Section 4,we
271、 have considered the impact(in terms of the number of Wi-Fi access points(APs)needed)of meeting the ECs Digital Decade policy programme target for all end users at fixed locations to have gigabit connectivity at least equivalent in speed to that of 5G.The analysis assumes connectivity within homes a
272、nd businesses uses Wi-Fi connected to a fixed broadband connection,and we consider Wi-Fi capabilities both with and without the availability of upper 6GHz spectrum,alongside the bands already available for Wi-Fi in Europe(in the 2.4GHz,5GHz and lower 6GHz bands).In this section,we further discuss th
273、e evolution of fixed broadband connectivity and the role of Wi-Fi,and we elaborate on the implications of the results of modelling performed by Huawei to assess the performance of Wi-Fi APs and terminals(STAs)in typical dense urban and rural dwellings as a function of the available spectrum.We then
274、consider the implications of the modelling outputs in terms of the carbon impact of having additional spectrum available for Wi-Fi use,compared to not having the additional spectrum available.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6
275、GHz band|41 Ref:728565284-225.5.2 Fixed broadband connectivity targets and technology evolution in Europe A key objective of the ECs Digital Decade vision is for all European households to have access to gigabit connectivity by 2030.Fixed broadband(FBB)technologies are being rolled out across Europe
276、,with fibre to the premises(FTTP)being one of the key solutions to provide gigabit connections close to houses and business premises.The adoption of fibre-optic technology for FBB connections has developed rapidly in the last 15 years,with the introduction of a new FTTx technology generation every 8
277、 to 10 years,delivering end-user speeds that are around four times higher than the previous generation.For example,G-PON(i.e.gigabit-capable passive optical network,or PON)technology is widely available today delivering up to 1Gbit/s service packages to end users,whereas 10G-PON became available in
278、2017 and is expected to reach large-scale take-up by 2026,delivering 1-5Gbit/s service packages to end users.The next generation of PON is the standard defined by the telecommunications sector of the ITU(i.e.ITU-T),called 50G PON.This became available in 2021,21 and is expected to deliver 10-20Gbit/
279、s service packages to end users.The first 50G PON products are expected to become commercially available before the end of 2023 and the technology is expected to reach a large-scale market by 2029.Preliminary research is underway for the next PON technology generation(although the ITU-T work towards
280、 standard development has not yet started),which will be required to deliver higher speeds to residential users on average,should market demand materialise in the future.However,the market launch for this new technology is unlikely to occur until the next decade.Accordingly,for the purposes of our a
281、nalysis,we use a 2030 target of 1Gbit/s within homes and business premises,consistent with the EUs Digital Decade vision.5.3 Additional spectrum and Wi-Fi throughput for typical premises Huawei has developed a model for the purpose of understanding the impacts on throughput in typical dense urban an
282、d rural premises of both the densification of Wi-Fi APs and the use of additional spectrum for Wi-Fi.This model simulates the operation of the latest type of Wi-Fi equipment,Wi-Fi 6,radios at the physical(PHY)and medium access control(MAC)layers,and accounts for the impact of co-channel and non-co-c
283、hannel interference in quantifying the achievable data throughputs.5.3.1 Modelled buildings The model has addressed two types of premises,representative of typical urban apartments and rural houses(see Figure 5.1):21 See https:/www.itu.int/hub/2021/06/new-itu-standards-to-boost-fibre-to-the-home-fro
284、m-10g-to-50g/Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|42 Ref:728565284-225.a middle-floor apartment with four 5m5m rooms,with Wi-Fi used in an apartment block containing nine other similar apartments on the same floor,ten si
285、milar apartments on the floor above and ten similar apartments on the floor below an isolated single-storey rural house with six 5m5m rooms.Radio propagation is modelled based on the New Radio(NR)indoor hotspot(InH)channel simulation model specified by the 3GPP.22 Penetration loss is modelled as 11d
286、B for the inner walls and 18dB for the floors.Figure 5.1:Modelled environment dense urban apartments and rural houses Source:Huawei,2023 5.3.2 Spectrum bands,frequency re-use and antenna technologies The operation of each Wi-Fi AP and terminal is modelled at the level of the PHY/MAC layers,accountin
287、g for co-channel and non-co-channel interference from all other modelled Wi-Fi equipment.In the case of the dense urban apartments,this implies modelling a total of up to 120 Wi-Fi APs(up to four per apartment)and 240 Wi-Fi terminals(two per room).To assess the impact of available spectrum on aggreg
288、ate throughput,the availability of the following bands for Wi-Fi is considered:2.4GHz,5GHz,lower 6GHz and upper 6GHz.In Europe,all of these bands apart from the upper 6GHz are already available for Wi-Fi use.The model implements frequency re-use amongst the frequencies used by each Wi-Fi AP,with dif
289、ferent re-use factors considered in different bands,and with each AP simultaneously using one channel per band.In addition,the model assumes that APs use a wired backhaul link(i.e.no Wi-Fi spectrum is used for communication between the APs and the on-premises internet router/modem).Regarding antenna
290、 technology,the model accounts for 44 MIMO in the Wi-Fi APs and 22 MIMO in the Wi-Fi terminals.This is a relatively conservative assumption,considering that larger numbers 22 TR 38.889 and TR 38.901.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the
291、upper 6GHz band|43 Ref:728565284-225.of antennas(e.g.8),which lead to higher throughputs,are expected to be commonplace in top-of-the-range Wi-Fi APs in future.5.3.3 Simulation methodology The simulation methodology is based on a Monte Carlo approach,whereby a large number of trials are performed,ea
292、ch using different locations for Wi-Fi terminals within the premises,in order to build up a statistical distribution of the achievable throughput.For each trial:A total of two Wi-Fi terminals per room are randomly positioned in the apartment or house The throughput between the terminal and its servi
293、ng Wi-Fi AP is calculated,under the assumption that all available bands are used,with one channel per band The aggregate throughput of all Wi-Fi terminals is calculated in the central apartment or the house(8 terminals in the apartment and 12 terminals in the house),by counting the number of success
294、fully delivered packets over a period of 1 second.The key parameters for each scenario include:The number of Wi-Fi APs,ranging from one to four APs per apartment in the dense urban setting,and from one to six APs per house in the rural setting,with the constraint that there is no more than one AP pe
295、r room The amount of spectrum available to the Wi-Fi APs and terminals The antenna technology(44 MIMO for APs and 22 MIMO for terminals).All Wi-Fi APs are assumed to serve their respective terminals at the same time(commonly known as a“full buffer”scenario),which represents a conservative assumption
296、.The result for each modelled environment is a cumulative distribution function giving,under the set of parameters selected,the probabilistic distribution of the aggregated throughput in the apartment or house.5.3.4 Simulation results For dense urban apartments,the modelling indicates that in 90%of
297、iterations of the model within a household,the target throughput of 1Gbit/s can be achieved with the spectrum currently available for Wi-Fi in the 2.4GHz,5GHz and lower 6GHz bands.This conservatively23 assumes that there is only one 44 MIMO Wi-Fi AP per apartment simultaneously serving a total of ei
298、ght 22 MIMO Wi-Fi terminals per apartment,relying on multi-user MIMO technology.To reach or exceed the target throughput of 1Gbit/s in 99%of iterations of the model within an apartment would require two 44 MIMO Wi-Fi APs per apartment,whether or not the upper 6GHz band is made available to Wi-Fi.23
299、Conservative,in the sense that larger numbers of antennas can be commonly expected in top-of-the-range Wi-Fi APs in future.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|44 Ref:728565284-225.For rural households,the simulations in
300、dicate that in 99%of all iterations of the model,the target throughput of 1Gbit/s can be readily achieved with the spectrum currently available for Wi-Fi(2.4GHz,5GHz and lower 6GHz bands)with only one 44 MIMO Wi-Fi AP per household simultaneously serving a total of 12 22 MIMO Wi-Fi terminals,relying
301、 on multi-user MIMO technology.5.4 Implications for the environmental impact The results of modelling show that the spectrum currently available to Wi-Fi in the 2.4GHz,5GHz and lower 6GHz bands is sufficient to deliver the Digital Decade target,and that the use of additional spectrum such as the upp
302、er 6GHz band would not result in a lower carbon footprint for Wi-Fi installations,since the same number of access points would be required to reach such a target,irrespective of the utilisation of the upper 6GHz band.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile network
303、s:the case of the upper 6GHz band|45 Ref:728565284-225.6 Conclusions Overall,our analysis demonstrates that,from a carbon footprint standpoint,it would be more beneficial to make additional mid-band spectrum available to 5G cellular networks than to rely exclusively on network densification to meet
304、future connectivity targets(in the absence of additional mid-band spectrum).This applies both in the dense urban area and the rural town or village we have modelled in our study.A summary of our other main conclusions is as follows:Our modelling results show that lower network carbon emissions arise
305、 in 5G mobile networks with additional mid-band spectrum available to meet the future connectivity targets we have considered in this report,compared to a situation where networks are densified through additional macro and outdoor small cells without the availability of additional mid-band spectrum.
306、We calculate that the carbon emission savings from having less densification in 5G mobile networks outweigh the incremental carbon emission costs of deploying and operating new radios(to support the additional mid-band spectrum we have modelled)on macro sites and outdoor small cells for the dense ur
307、ban area and for macro sites in the rural town or village.It should be noted that in the dense urban area we consider two deployment variants for densification(in the absence of additional mid-band spectrum):firstly,densification primarily via macro sites(with some supporting outdoor small cells)sec
308、ondly,densification primarily via additional outdoor small cells(and thus lower macro-site densification).In each case,the incremental carbon emission cost of deploying and operating new upper mid-band radios at the dense urban macro sites and outdoor small cells is lower than the incremental carbon
309、 footprint associated with the higher level of densification needed without the additional mid-band spectrum.In addition to the increased carbon footprint associated with greater densification,the levels of densification that would be required in 5G mobile networks to meet the connectivity targets i
310、n the absence of additional mid-band spectrum would be practically challenging and also potentially technically unfeasible(due to interference between sites that are too close to each other).Sensitivity analysis suggests that increasing the activity factor for MBB use in the dense urban environment
311、from 5%to 20%increases the carbon savings.This is because the difference between the volume of site build required with and without additional 5G upper mid-band spectrum is exacerbated as demand on the network increases.Similarly,as high-band offloading Impact of additional mid-band spectrum on the
312、carbon footprint of 5G mobile networks:the case of the upper 6GHz band|46 Ref:728565284-225.increases the carbon savings decrease.This is because there is less demand on the macro sites and small cells,so less densification is required and thus the impact of additional spectrum is reduced.However,we
313、 note that high band offloading is only possible if sites at those specific locations where demand is highest can accommodate high-band deployment,and/or where new mmWave sites can be deployed.It is also noted that the mmWave coverage is lower than the coverage achieved from either 5G mid-bands or f
314、rom other,lower,frequency bands used in 5G mobile networks.Hence,mmWave offloading can complement 5G mid-band deployment but does not provide a direct substitute for it due to coverage differences.For Wi-Fi,based on simulations made available to us and considering the future connectivity targets for
315、 fixed broadband(i.e.an aggregated throughput of more than 1Gbit/s per premises),availability of the upper 6GHz band would not translate into any reduction in carbon emissions,given that such targets can be met via the latest Wi-Fi technology using spectrum bands already available for Wi-Fi use in E
316、urope(2.4GHz,5GHz and lower 6GHz).While these results have been modelled assuming upper 6GHz deployment(e.g.in terms of the bandwidth available),these conclusions may apply to other upper mid-band spectrum,provided that the alternative upper mid-band spectrum exhibits similar characteristics to thos
317、e modelled here.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|A1 Ref:728565284-225 Annex A 5G mobile network modelling methodology This annex describes the detailed modelling methodology used for the dense urban area(Section A.1)
318、and the rural town or village(Section A.2).The methodology uses a number of inputs that are described in this annex.The calculations performed by the model are repeated for spectrum configurations with,and without,additional upper mid-band spectrum,and then the difference in potential carbon emissio
319、ns between the two spectrum configurations is calculated.For the urban settlement the model is run twice to reflect two potential deployment variants in the absence of adding spectrum densification primarily through macro sites,and densification primarily through outdoor small cells.A.1 Modelling me
320、thodology for dense urban area In order to calculate the embodied and annual carbon emissions of a particular dense urban access network deployment the density of macro sites and outdoor small cells must be determined.There are several steps to the calculation,as explained below.Note that the diagra
321、ms follow on from one another and that any inputs that have been calculated in a previous diagram are identified via a dark border and the relevant diagram is listed in the legend.A.1.1 Calculation of dense urban macro-site and outdoor small-cell density The density of macro sites and outdoor small
322、cells is determined by comparing the uplink and downlink MBB capacity that can be supplied for one macro site and its supplementary outdoor small cells24 with the required MBB capacity.Figure A.1 shows the wider processing of inputs required to arrive at this overarching calculation(shown in the das
323、hed box with the darker grey outline).This wider processing has been broken down into three constituent components,discussed in turn below.Values used in the calculation are end-of-period(EOP)unless stated for example,beginning of period(BOP)is used when calculating macro-cell radius,as shown in Fig
324、ure A.1 below.24 We assume an average ratio of small cells per macro site in a given location,noting that in practice the ratio will vary,as small cells are likely to be less evenly distributed than macro sites.Impact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the
325、case of the upper 6GHz band|A2 Ref:728565284-225 Figure A.1:Calculation of dense urban macro-site and outdoor small-cell density Source:Analysys Mason,2023 Activity factorActivity factor%High band High band offloadingoffloading%/yearUrban population Urban population densitydensity/km2Adjustment to A
326、djustment to account for account for outsideoutside-in in coveragecoverage%Available Available spectrum*spectrum*MHz/yearMBB device takeMBB device take-up up(upper mid-band and additional upper mid-band)%/yearSpectrum rollSpectrum roll-out out profile profile(upper mid-band and additional upper mid-
327、band)%sites/yearDesign marginDesign margin%3 sectors per 3 sectors per macro sitemacro siteMacro site MBB Macro site MBB spectral spectral efficiency*efficiency*Mbit/s/MHz/yearMacro site Macro site available MBB available MBB capacity*capacity*Mbit/s/site/yearOutdoor small cell Outdoor small cell av
328、ailable MBB available MBB capacity*capacity*Mbit/s/cell/yearSpectrum Spectrum available per available per sector*sector*MHz/sector/year1 sector per 1 sector per outdoor small celloutdoor small cellOutdoor small cell Outdoor small cell MBB spectral MBB spectral efficiency*efficiency*Mbit/s/MHz/yearMa
329、cro site radiusMacro site radiuskm/year(BOP)Outdoor small Outdoor small cell radiuscell radiuskmMaximum outdoor Maximum outdoor small cells per small cells per macro sitemacro site/yearAvailable MBB Available MBB capacity supply*capacity supply*Mbit/s/macro site and small cell/yearRequired MBB Requi
330、red MBB capacity*capacity*Mbit/s/km2/yearMacro site and Macro site and outdoor small cell outdoor small cell densitydensity/km2/year*Downlink and uplinkCalculationOutputInputMBB service MBB service growth*growth*User MBB User MBB required capacity*required capacity*Mbit/s/active user/year2022 MBB 20
331、22 MBB service*service*Mbit/s/active user2030 MBB 2030 MBB target*target*Mbit/s/active userABCAvailability of Availability of dense urban dense urban macro cell capacitymacro cell capacityTotal available dense Total available dense urban capacity supplyurban capacity supplyCalculation of required ca
332、pacityCalculation of required capacityImpact of additional mid-band spectrum on the carbon footprint of 5G mobile networks:the case of the upper 6GHz band|A3 Ref:728565284-225 Calculation of available dense urban site/cell MBB capacity supply For each macro site and outdoor small cell,the available
333、downlink and uplink MBB capacity supply is calculated on the basis of the spectrum available to the average site/cell,the number of sectors and the corresponding spectral efficiency.As illustrated in Figure A.2,there are a number of inputs that determine the spectrum deployed on an average site:The proportion of sites on which a given band has been deployed for modelling purposes this is taken to