1、2EFFECTIVE ENERGY SAVING STRATEGIES AND BEST PRACTICES FOR MNOs by NGMN AllianceVersion:1.0Date:07.11.2023Document Type:Final Deliverable(approved)Confidentiality Class:P-PublicProject:Green Future NetworksEditor/Submitter:Antonio De Domenico,HuaweiRishikesh Chakraborty,VodafoneNGMN Programme Office

2、:Chris Hogg,representing NGMNApproved by/Date:NGMN Alliance Board,27 October 2023 For Public documents(P):2023 Next Generation Mobile Networks Alliance e.V.All rights reserved.No part of this document may be reproduced or transmitted in any form or by any means without prior written permission from

3、NGMN Alliance e.V.The information contained in this document represents the current view held by NGMN Alliance e.V.on the issues discussed as of the date of publication.This document is provided“as is”with no warranties whatsoever including any warranty of merchantability,non-infringement,or fitness

4、 for any particular purpose.All liability(including liability for infringement of any property rights)relating to the use of information in this document is disclaimed.No license,express or implied,to any intellectual property rights are granted herein.This document is distri-buted for informational

5、 purposes only and is subject to change without notice.3EXECUTIVE SUMMARYIn recent years,Mobile Network Operators(MNOs)have faced a more complex busi-ness environment.Tightening energy markets as well as spikes in natural gas prices have contributed to a challenging energy landscape.The significant

6、increase in MNOs operational costs impacted efforts to deliver on transformational connectivity goals and meet investment targets.Fortunately,MNOs are responding collectively to mitigate this crisis by:(1)reducing energy consumption by implementing efficiency measures,and(2)by ensuring stability of

7、energy supply.To increase energy efficiency,MNOs implemented a variety of methods,and amongst these,the most effective ones are highlighted in this publication.Specifically,MNOs have suggested to shut down legacy networks,re-farming spectrum for more energy efficient radio access technologies(RATs).

8、They have also stressed the need to accelerate the modernisation of Radio Access Network(RAN)equipment to integrate more energy efficient hardware.In addition,power saving features should be implemented,such as switching off capacity carrier frequencies during periods of low traffic.Importantly,the

9、state of extreme deep dormancy could be activated as a means to further optimise energy consumption at the Base Station(BS).Other methods have been explored by some MNOs in an effort to reduce energy consumption.These included simplified architectures at a network site to reduce the need for active

10、cooling,using more advanced monitoring and control systems and looking at liquid cooling.To ensure stability of energy supply,MNOs are working to accelerate the transition to renewable sources of energy for net-work sites combined with back-up battery storage.MNOs also seek to optimise network energ

11、y flexibility.Ultimately,the collective MNO response to this challenging energy landscape yielded a range of methods that were regarded as“best practice”that could be relevant for future solutions,should the energy crisis persist.This involved shutting down 2G and 3G where possible,shifting to more

12、energy efficient RAN equipment especially in 4G and 5G de-ployments.Moreover,shifting from legacy airconditioning units towards more energy efficient cooling solutions could be an effective way to optimise energy consumption.NGMN will argue for the above recommendations to mitigate the energy challe

13、nges experienced by MNOs.Future work that builds on this publication will also look at novel technologies that can support such efforts to make mobile networks greener by impro-ving network energy efficiency.This could include looking at Artificial Intelligence(AI)for more intelligent energy managem

14、ent.It involves also looking at new opportunities for embedding software in the mobile networks.4CONTENTS01 SCOPE AND INTRODUCTION .502 CHALLENGES FOR THE TELCO SECTOR ENERGY .603 MNO RESPONSE TO THE CRISIS.704 REDUCING ENERGY CONSUMPTION.84.1 Improving the site energy efficiency through modular sit

15、e architectures .84.2 Monitoring and control systems.94.3 Improving cooling efficiency:liquid cooling.104.4 Optimising BS energy consumption.114.4.1 Carrier shutdown.114.4.2 Extreme Deep Dormancy .1205 ENSURING ENERGY STABILITY.135.1 Renewable electricity generation and battery storage.135.2 Optimis

16、ing battery backup autonomy.145.3 Improving the network energy flexibility.1406 BEST PRACTICE.1607 ABBREVIATIONS.1708 ACKNOWLEDGEMENTS.1809 REFERENCES.19501 SCOPE AND INTRODUCTION Energy efficiency,along with strategies towards significant reduction in energy con-sumption,coupled with minimisation a

17、nd elimination of energy waste,has become an essential requirement.This responds to environmental as well as economic sustaina-bility,particularly in the presence of the ongoing energy challenges.These challenges include the significant increase in processing of data with new and expanded network ca

18、pabilities,despite the more efficient 5G technology.The purpose of this publication is to tackle the recent energy price spike challenges confronting the telecoms industry,which are exacerbating the need to improve net-work energy efficiency.This is the first piece of the NGMN Alliance(NGMN)work tha

19、t demonstrates a collective MNO response to the challenging energy landscape through reduction in energy consumption and efforts to safeguard energy stability without ne-gatively impacting the network performance.It follows previous NGMN publications on the topic of Network Energy Efficiency 1 2.Sec

20、tion 2 contextualises the energy price volatility challenges that were acutely felt by the telecoms industry,especially for MNOs.Section 3 then introduces the MNO response to this challenging climate in two ways:firstly,in Section 4,by outlining the methods used to reduce network energy consumption,

21、and secondly,in Section 5,by presenting solutions to ensure energy stability so as not to undermine quality of service.The United Nations energy related goal ensuring access to affordable,reliable,sustainable and mo-dern energy for all requires increasing the renewable energy sources in mobile netwo

22、rk operators total final energy consumption and decreasing the energy consumption of the equipment.Finally,Section 6 handles the key learnings from this analysis,highlighting the best practice undertaken by MNOs.602 CHALLENGES FOR THE TELCO SECTOR Energy markets began to tighten in 2021 due to a var

23、iety of factors:1.the extraordinarily rapid economic rebound following the COVID-19 pandemic which increased the demand for energy,2.a long,cold winter in the Northern Hemisphere which increased energy demand across the whole world,3.a weaker-than-expected increase in energy supply.Natural gas price

24、s have seen the biggest increase,hitting historical maximums.1 The significance of the energy situation has further grown,especially in 2022,mainly due to a reduction in supply,the disruption of energy trade flows,and competition for gas supply security.These prices reached their maximum in July and

25、 August 2022.Against this challenging backdrop of sustained energy price rises,it became a priority in the telecoms sector to attempt to reduce energy costs as much as possible by securing and stabilising whatever energy was available.However,another crucial component of the telecoms industrys energ

26、y transition involves altering energy consumption patterns on the network side by increasing network energy efficiency as well as transitioning towards renewable energy sources to reduce greenhouse gas(GHG)emissions aligned with the goal to mitigate climate change while satisfying the need for expan

27、ded network capabilities.Given the unprecedented and acute nature of price rises in 2022,the telecommunications industry has been faced with significantly increased operating costs;however,attempts to secure energy supply to combat these rises in the short term should not defer other important energ

28、y transition actions.In this context,energy efficiency,particularly in the mobile networks will be an even more important part of the energy transition and es-sential to get right,helping to reduce energy costs and GHG emissions across the globe.In addition,if national power grids were temporarily u

29、nable to provide electricity,MNOs may face the threat of network outages which could lead to loss of service within hours,or even instantaneously for sites that lack back-up electricity.These two risk factors combined could undermine the telecommunication industrys ambitious societal targets to deli

30、ver for citizens,governments,and businesses.1Source:IEA-What is behind soaring energy prices and what happens next?703 MNO RESPONSE TO THE CRISISMNOs took steps to manage the energy crisis.Broadly,this falls into two categories:(1)reducing energy consumption,and(2)ensuring energy stability to mainta

31、in service continuity.This includes developing solutions to support the transition to greener and more resilient networks and to optimise network energy efficiency,i.e.,the number of delivered useful bits per unit of consumed energy.To satisfy the request for novel mo-bile services and stringent per

32、formance requirements,MNOs have deployed new sites.Although this contributes to increase the energy consumption,many MNOs kept the overall net energy consumption broadly flat by gradually shifting to more energy efficient technologies.These include further deployment of new radio access technologies

33、(RATs),where 4G and 5G are at least five times more energy efficient than 3G when carrying the same amount of traffic and the application of 5G Massive Multiple Input Multiple Out-put(M-MIMO)which is at least three times more energy efficient than 4G M-MIMO 2.Consolidation of fixed access and the co

34、re/data centre network infrastructure improves energy efficiency.Another mechanism to optimise energy consumption is via dynamic use of low power modes.To assist in the delivery of the above energy efficiency optimisation aims,MNOs are strongly reducing the energy consumption of the RAN,which accoun

35、ts for almost three-quarters of total network energy consumption by shutting down 2G and 3G networks,wherever possible,2 and refarming 2G and 3G spectrum for more energy efficient RATs(e.g.,LTE and NR),3 accelerating modernisation of RAN equipment to benefit from next generation energy efficient sol

36、utions,implementing power saving features to switch off capacity layers during periods of low traffic demand without negatively affecting the user performance.The following subsections describe in more detail methods being used by MNOs to manage the energy challenges.2According to Analysis Mason,MNO

37、s that run 2G,3G,4G and 5G with separate BSs could potentially reduce their energy consumption by up to 24.4%at each macro site by switching off 3G 83https:/ REDUCING ENERGY CONSUMPTIONThis section describes the methods used by MNOs to reduce or optimise energy consumption as part of efforts to deal

38、 with a challenging energy landscape.The main methods used,each of which is described in turn,were:improving site energy efficiency,deploying monitoring and control systems to check energy consumption and network performance parameters,improving power and cooling efficiency,and optimising energy con

39、sumption at the base station through novel sleeping features.4.1 IMPROVING THE SITE ENERGY EFFICIENCY THROUGH MODULAR SITE ARCHITECTURES As discussed in 2,the Radio Access Network(RAN)ac-counts for around three-quarters of all mobile network electricity consumption with the Base Station(BS)typically

40、 resulting for more than half of the electricity consumption of a typical cell site.In addition to the BS,the energy con-sumption of the site support equipment is significant.In traditional sites,based on the architecture shown below in Figure 1,the site support equipment power consump-tion,mainly d

41、riven by the air-conditioned equipment room,can be as great as 40%of the total site energy consumption.All power consumed at a site is the sum of BS power consumption and site support equipment power consumption,with the precise ratio being influenced by a range of environmental factors.In a cooler

42、environment with passive cooling the air conditioning load is small,in a warmer external environment the air conditioning load would be significantly higher.Novel site design solutions can determine which sites really need air conditioning and a temperature control system,so that these are deployed

43、only where required.Figure 1:Traditional site with indoor site support equipmentGiven that the total site energy consumption is the sum of BS and site support equipment consumption,this can be limited by reducing energy consumption of either or both of these two components.The following subsection f

44、ocuses on improving site energy efficiency by reducing site support equipment energy consumption.Partly,this optimisation involves limiting the power consumption of multiple site support equipment components.Historically,access site deployment involved placing site support equipment indoors while th

45、e telecoms equipment(e.g.,BS,RRUs and antennae)were located outdoors.With modular site architectures,site support equipment could be moved from traditional indoor sites to outdoor sites as shown in Figure 2.Shifting some of this equipment outdoors together with further deployment of outdoor poles re

46、sults in reduced energy consumption by outdoor cooling.This approach also reduces power loss due to the conversion for power continuity,which can be further minimised with the deployment of state-of-the-art power solutions that have high conversion efficiency.Switching off legacy radio access techno

47、logies(RATs)has been associated with transitions to simpler architectures and with novel hardware able to run at higher tempe-ratures,which allows the deployment of more modular components outdoors rather than indoors.9A variant of this approach involved placing network com-ponents,e.g.,antennas,rad

48、io,and Baseband units(BBUs),which can support multiple technologies,in one centrali-sed piece of equipment,without adversely affecting user experience on the live network.4 Figure 2:Traditional and simplified sitesIt is important to highlight that achieving low-cost heat dissipation control,efficien

49、t heat management,and high reliability in a simplified site operating in different environmental conditions,hardware configurations and processing requirements,is a challenge,which requires system design methods based on accurate thermal si-mulation design and integrated performance verification cap

50、abilities.Going forwards,to evaluate the energy-saving effective-ness of these simplified architectures fully,it is necessary to reach a consensus on the scope of the site energy efficiency(SEE).Should it focus on the RAN only or also include the transport network?The collection of site energy consu

51、mption data is a pre-requisite in the SEE evaluation.Potentially,this could be assisted by automation of some processes,whose energy consumption impact should be considered.54.2 MONITORING AND CONTROL SYSTEMSTo analyse and then improve the overall energy efficiency,is key to deploy monitoring system

52、s on energy usage,traffic,and network performance.A monitoring system allows data management and analytics by collecting,processing,and storing data feeds from electricity sup-pliers and smart meters.Effective monitoring of energy consumption,data traffic,and configuration data feeds have paved the

53、way to embed some degree of automation to control and optimise energy efficiency.4As an example,for a like for like upgrade comparing adding a new single band radio to legacy radio vs deploying new multiband radios,we can often achieve a 30%energy reduction(Source:results reported to NGMN by Vodafon

54、e).5NGMN expects to track this developing area in future reports focusing on metering for sustainable networks.In this context,some MNOs have already used the fol-lowing technologies leveraging the availability of energy and network data:Energy saving features to,for example,put the RAN sites in low

55、 power mode during low traffic periods.These features allow further monitoring,tracking,and assessing the impact on energy savings in the network and business Key Performance Indicators(KPIs).Passive infrastructure optimisation to monitor energy consumption metrics related to BS support equipment an

56、d identify relatively energy inefficient sites.Smart sites remote control of passive infrastructure to optimise energy consumption,Operation and Mainte-nance(O&M),and Capital Expenditure(CAPEX)planning.Identification of energy saving opportunities via site energy efficiency analysis.Dynamic thermal

57、management to control cooling in real-time by balancing the airflow demand with the supplies from the air conditioning units in Core and Data Centre sites.BBU Traffic monitoring to activate BSs in the extreme dormancy state in realtime.It should be noted that while automation alone is no silver bull

58、et since it has its own associated energy footprint,the above examples are areas where MNOs have discovered automation to be effective in helping to reduce energy consumption without affecting negatively network per-formance.104.3 IMPROVING COOLING EFFICIENCY:LIQUID COOLINGCooling systems at a netwo

59、rk site,which remove heat generated from the ICT equipment can have a significant impact on the site energy consumption.MNOs have tried to minimise cooling-induced energy consumption using the following techniques:Replacing refrigerants with lower Global Warming Potential(GWP)solutions.Deploying new

60、 cooling technologies that are significantly more efficient.Maximising the usage of free cooling,thus guaranteeing minimal consumption.Regarding cooling solutions,free cooling can be exploited in out-door sites;however,new cooling technologies are still required for indoor sites(see Section 4.1).In

61、the following subsections,we present in more detail advantages and challenges when deploying liquid cooling technologies.Liquid cooling technology has been regarded by some MNOs as an efficient way to reduce the energy consumption in the site support equipment room,especially with the increasing con

62、centration of BBU deployments.In the following,we discuss the four types of liquid cooling technologies available for equipment rooms that MNOs have explored:Cabinet liquid cooling:The BBU equipment is placed in the liquid cooling cabinet with liquid cooling pipes on the back of the cabinet.A Coolan

63、t Distribution Unit(CDU)and a water pump are used to circulate the liquid in the cooling pipes continuously so that the heat generated by the BBU equip-ment is absorbed into the liquid of the cooling pipes,then discharged into the atmosphere by the external radiator,resulting in heat dissipation.Dir

64、ect cold plate liquid cooling(DCLC):A liquid co-oling pipeline is fixed to a BBU board card.There can be a customised design for different boards,with the liquid cooling pipeline mainly fixed onto a high-power heating device.The coolant is kept flowing uniformly and circularly through the liquid coo

65、ling plates of each BBU component to export the heat generated by the heating device through the liquid cooling pipe.The heat generated by the equipment is dissipated by the external heat sink due to heat exchange between the indoor and outdoor environments respectively.Immersion liquid cooling:The

66、BBU equipment is immersed directly in the coolant,so that the he-at-generating electronic components(e.g.,CPU,motherboard,memory module,hard disk,etc.)are in direct contact with the insulating and chemically inert coolant.Thus,the heat generated by these electronic components are dissipated through

67、the circulating coolant.Spray liquid cooling:The BBU equipment is placed in a spray cabinet,which is structured in layers.The cooling liquid is stored within the layers of the spray cabinet,with open holes on the top of each layer.The coolant is sprayed directly onto the equipment to allow it to coo

68、l.Table 1 provides a comparison of the four ty-pes of the above discussed liquid cooling tech-nologies regarding heat dissipation capabilities,measured in Power Usage effectiveness(PUE),corresponding requirements for the equipment room,and compatibility with existing equipment.6 6 6Note that the PUE

69、 values indicated in Table 1 are typically larger than those reported in literature when presenting recent advancements in cooling applied to data centers,where equipments are usually more concentrated and have higher power consumption,which leads to lower PUE values(around 1.05)11Table 1:Comparison

70、 of different types of liquid cooling technologies based on trial and qualitative analysis.Cabinet liquid coolingCold plate liquid coolingImmersion liquid coolingSpray liquid coolingPUE for a number of BBUs1.4(10 BBUs)1.35(5 BBUs)1.2(10 BBUs)ReliabilityVery goodGood if care is taken to avoid leakage

71、 due to corrosionGood if care is taken with the optical module and high-speed connectors.MaintainabilityVery goodVery goodGood-however causticity of the liquid has an impact on installation.Dry cham-bers are used to prevent contact with the equipment and drip trays are deployed to prevent any room s

72、pillage.Requirements for equipment roomNo requirementsNo requirementsLoad bearing and space needs considerationLarger space requiredCompatibilityGoodPoor,it cannot be used for existing BBUsPoor-but can be transformed4.4 OPTIMISING BS ENERGY CONSUMPTIONTo reduce the energy consumption of the RAN whil

73、e maintaining stable network KPIs,MNOs have explored the deployment of new RAN equipment with reduced load independent power consumption as well as the optimisation of network capacity according to traffic de-mand by switching off temporarily underutilised carriers.Time variation and nonuniform spat

74、ial distribution of the system traffic results in significant energy waste,because the energy-consuming BS components keep working even when there is no data to transmit.Different levels of BS dormancy are of benefit to varying degrees,often invol-ving turning off part of the active component of Act

75、ive Antenna Units(AAUs)and of Radio Remote Units(RRUs)in accordance with service requirements,as described in 2.In this section,we report the energy saving gains that have been achieved deploying carrier shutdown and extreme deep dormancy.4.4.1 CARRIER SHUTDOWNResults from China Mobile lab tests and

76、 real network tests demonstrated that in zero traffic load scenarios,AAU/RRU carrier shutdown yields power consumption reductions of 20%50%.It should be noted,however,that extending the lifecycle of BSs-a typical BS lifecycle is approximately 8-15 years-is in conflict with swapping legacy BS compone

77、nts for more energy-efficient solutions.The optimal choice depends on the specific conditions of a site and its associated traffic profile.An example of a carrier shutdown process used by MNOs is described in Figure 3.Figure 3:Example of carrier shutdown in 4G network12Figure 4 shows the cell level

78、traffic pattern for both high loaded and low loaded cells and the corresponding power saving window when activating the strategy described in Figure 3.The window size corresponds to the amount of energy that is saved in a cell.Figure 4:Traffic profiles of 4G cells with corresponding power saving win

79、dow4.4.2 EXTREME DEEP DORMANCY To investigate the possibility of optimising BS energy con-sumption further,MNOs explored the concept of extreme deep dormancy.Even when the BS is in deep dormancy,the active components of AAU/RRU still consumes dozens of watts of power.To achieve extreme energy saving

80、,the BS could be put in a state of extreme dormancy.This is defined as the state of reduced power consumption achie-ved by muting all the modules of the AAU/RRU except for the inner auxiliary source,timing module,microcontroller and supporting control circuit of the power supply.The power consumptio

81、n of the AAU/RRU will be significantly lower when extreme dormancy is activated,and the time that it takes to wake up to normal operations will not be longer than 5 minutes.Tests from China Mobile highlighted that there was not any drop rate when users were transitioned to the coverage cell.Neverthe

82、less,the 5 minutes wake up might be a limiting factor for MNOs to implement this feature.Adoption could be greater if the wake-up time were to be reduced.Current BSs support extreme dormancy based on a predefined timing setting,as well as real-time wake-up/sleep modes through BBU traffic monitoring.

83、The latter approach is more flexible and efficient for energy saving while avoiding potential service quality degradation but requires certain dedicated hardware to receive wake-up signals from the BBU.1305 ENSURING ENERGY STABILITYThe second pillar of the MNO response to the energy challenge is to

84、ensure energy stability against a backdrop of energy supply,so that quality of service(QoS)could be maintained.MNOs have proactively worked towards ensuring a resilient and affordable energy supply using the following methods:Responding to calls for load shedding preparations and power cut/outage ma

85、nagement during winter.Deploying equipment for production of renewable energy.Leveraging hydrogen,ammonia and fuel cell techno-logies to produce carbon-free energy.Optimising battery usage for RAN sites.Improving the resilience of the electrical network and avoiding unnecessary electricity expenses

86、through energy flexibility initiatives.5.1 RENEWABLE ELECTRICITY GENERATION AND BATTERY STORAGEFor MNOs,renewables are a key initiative to leverage onsite power generation to reduce the power from the grid,as well as reduce carbon emissions from diesel generators for off-grid RAN sites.To maximise o

87、nsite electricity generation,MNOs are eva-luating the possibility of using solar and wind together.Wind energy may complement solar energy and the energy generated from both can be stored in batteries,which can be used when neither solar nor wind is enough to meet energy demand or during energy cons

88、umption peaks.Unfortunately,both energy production technologies still have high deployment costs,and only some sites are suitable for renewable implementation due to limiting factors e.g.,shading,space,wind speed,etc.Moreover,it may be difficult to obtain the necessary permits,whereby these may also

89、 require long waiting times to be granted by local regulators.MNOs have sought innovative solutions to the challenge of generating renewable power directly at RAN sites in rural or remote areas without access to power grids or where there are frequent power outages while avoiding using fossil fuels

90、onsite.Together with renewables,the deployment of battery equipment on site,is key to ensuring continuous service and maximising the usage of renewable sources.Electri-city networks in advanced markets have been assumed to be stable and resilient based on historical records de-monstrating many years

91、 of largely uninterrupted supply.This has led to a reduction in backup solutions onsite,leading to increased dependency and vulnerability to grid outages.Today,MNOs are deploying more back-up battery capacity as a smarter way to reduce dependence on electricity grids.Both the RAN and Core network in

92、frastructure can be equipped with batteries,and in some cases,even with generators or fuel cells offering protection against vul-nerability to power outages.While this initiative does not necessarily save energy alone,it does support MNO business continuity plans to guarantee QoS for customers.145.2

93、 OPTIMISING BATTERY BACKUP AUTONOMYMaintaining uninterrupted service to customers in the face of energy supply risks necessitates improving ener-gy backup autonomy for RAN sites.The energy backup situation varies significantly across markets.In markets with good quality in energy supply,the majority

94、 of RAN sites have a battery lifetime ranging from approximately 15 to 30 minutes.By activating intelligent energy saving features,e.g.,using the Operational Support System(OSS)through cell switch-off,the required power can be reduced,subsequently increasing battery autonomy,while maintaining stable

95、 KPIs 2.The energy consumption of RAN sites has increased due to capacity upgrades,often involving the addition of new radios on the same site to cater to growing traffic demands.This introduces the concept of radios for coverage and radios for capacity,where coverage bands provide better propagatio

96、n conditions,encompassing frequencies up to 900 MHz for example,while capacity bands have a large amount of available resources,typically including frequen-cies above 900 MHz(see Figure 5).To address energy crisis scenarios,a gradual switch-off of frequencies from capacity bands to coverage bands ca

97、n be implemented.For example,a sequential switch-off order can operate as follows:2.6 GHz,2.1 GHz,1.8 GHz,and finally 700 GHz,where each respective carrier frequency is switched off in turn,depending on traffic conditions.Figure 5:Example of capacity and coverage bands for different RATsThe cell swi

98、tch-off feature,when used to deactivate a carrier frequency,can yield a potential power gain of approximately 100W-200W per frequency per site(across 3 sectors).77Source:Analysis reported to NGMN by Orange.However,the potential gains from frequency band switch-off may not be substantial due to the i

99、nherent limitations in completely turning off the power amplifiers(PAs).This is because a single PA commonly handles multiple frequencies.This is generally a more energy efficient approach than using a distinct PA per frequency band during running time but it places a ceiling on how much power reduc

100、tion through frequency band switch-off alone can be achieved 2.Moreover,complete powering down of the PAs is not feasible when rapid wake-up is required for ensuring uninterrupted service availability.Moreover,the energy saving gain is highly dependent on the traffic load and the type of radio modul

101、es employed on the site,including factors such as the product type,generation,and multi-band capability.Furthermore,even when cells are switched off,equip-ment such as RRUs continue to consume significant energy.Additional savings(approximately 150 W/RRU)can be achieved by completely switching off t

102、he equip-ment.8 However,remote and automated switching off of equipment is currently not possible due to the need for additional hardware and the requirement to address potential reliability issues.The deployment of smart circuit breakers allows remote and automated switch-off of capacity cells in o

103、rder to provide up to six hours of autonomy,for example in voice services operating on the 700 MHz frequency band.5.3 IMPROVING THE NETWORK ENERGY FLEXIBILITYEnergy flexibility refers to the ability to adapt electricity generation and/or consumption for a specific period to contribute to the supply-

104、demand balance of the electri-city system.It offers several benefits,including reducing energy bills,optimising the utilisation of renewable ener-gy,participating in the electricity market to support grid stability,and leveraging storage investments.8Source:Analysis reported to NGMN by Orange.15For

105、MNOs,there are implicit and explicit solutions to valorise flexibility actions:Implicit energy flexibility:MNOs are encouraged to regulate their power consumption to optimise their bills by leveraging dynamic electricity tariffs.These tariffs may range from simple day and night prices to highly dyna

106、mic prices based on hourly wholesale prices.Implicit flexibility can be implemented in various ways,through e.g.,automated software,including Peak Shaving,Time Shifting,and Flexibility in Time and Amount,as depicted in Figure 6.Explicit energy flexibility:MNOs can also benefit from flexibility indiv

107、idually,either by selling energy blocks in wholesale markets or through contracts with aggregators.Energy flexibility opportunities can be explored across various energy assets,including networks,tertiary sites,Electric Vehicle(EV)fleets,and clients domains.Among the identified scenarios,leveraging

108、backup batteries in net-work assets-or even deploying additional batteries-has been found to be the most profitable and mature option.A detailed case study could be considered for exploring different mechanisms,for example,time-shifting using Li-ion batteries on RAN sites.Figure 6:Types of implicit

109、solutions for energy demand side flexibility1606 BEST PRACTICEReducing energy usage now is more important than ever.In this report,we have described how MNOs have taken smart measures at an accelerated pace to further reduce energy consumption in the short term,while simultaneously maintaining desir

110、able network QoS and handling a continuous increase of data traffic volume:Accelerate the transition from legacy networks(e.g.2G and 3G)to more energy efficient technologies(e.g.,4G and 5G)whenever possible.Deploying additional 5G/4G spectrum refarmed from 3G to increase network data throughput and

111、capacity,and to keep up with the high data growth experienced during the past years.Leveraging the capabilities of the latest RAN technologies to achieve higher energy efficiency.Implementing Intelligent RAN energy saving features to save more energy in both 4G and 5G layers.Updating RAN sites with

112、state-of-the-art energy efficient RAN equipment and antennas which facilitate all mobile generations,2G to 5G.Replacing all onsite airconditioning units with energy efficient cooling.Deploying renewable energy sources and battery storage units,which reduce the net-work dependency by electrical grids

113、.Use intelligent energy saving features together with battery storage equipment to further increase the network stability.For example,deployment of some of these solutions has allowed KPN to keep energy consumption flat over the years 2018-2022,while data traffic did increase by 33%annually.Similarl

114、y,in the past five years,in the Vodafone network,data traffic has increased over 600%yet the absolute energy consumption has not changed,representing a reduction of energy consumed per unit of data traffic of nearly 90%.Besides these best practices,we have identified in this publication relevant con

115、straints that limit the adoption of energy saving features and equipment leading to improved energy efficiency,specifically:The adoption of extreme dormancy features could be greater if the BS wake-up time could be reduced.The target of extending the BS lifecycle is in conflict with swapping legacy

116、BS compo-nents for introducing more energy-efficient solutions.Potential energy saving gains from frequency band switch-off may not be substantial due to the inability in completely turning off the multi-carrier Pas.Deploying renewable energy solutions could be accelerated if the process for obtaini

117、ng the necessary permits could be simplified.1707 ABBREVIATIONSAAUActive Antenna UnitAIArtificial IntelligenceBBUBaseband UnitBSBase StationCAPEXCapital ExpenditureCDUCoolant Distribution UnitCO2Carbon DioxideCPUCentral Processing UnitCUCentralised UnitDUDistributed UnitECEuropean CommissionETSEmiss

118、ions Trading SystemEUEuropean UnionEVElectric VehicleGHGGreenhouse gasICTInformation and Communication TechnologyIoTInternet of ThingIPCCIntergovernmental Panel on Climate ChangeISVindependent Software VendorsKPIKey Performance IndicatorMIMOMultiple Input Multiple OutputMLMachine LearningMNOMobile N

119、etwork OperatorO&MOperation and MaintenanceOSSOperational Support SystemPAPower AmplifierPRBPhysical Resource BlockPUEPower Usage EffectivenessQoSQuality of ServiceRANRadio Access NetworkRATRadio Access TechnologyRRURemote Radio UnitSSESite Energy EfficiencyTDPThermal design power1808 ACKNOWLEDGEMEN

120、TSThe following people and companies were instrumental to developing this publication.Javan Erfanian,Bell CanadaJianhua Liu,China MobileCong Zhang,China MobileSaima Ansari,Deutsche TelekomPaul Pfundt,Deutsche TelekomAntonio De Domenico,HuaweiWang Man,HuaweiGeng Xinli,HuaweiGary Li,IntelRichold van d

121、er Wal,KPNHommad el Allali,KPNOlivier Guyot,NokiaYuanyuan Huang,Orange?zzet Salam,TurkcellRishikesh Chakraborty,Vodafone1909 REFERENCES1 NGMN Alliance,Green Future Networks:NETWORK ENERGY EFFICIENCY Phase I,2021.Online.Available:https:/www.ngmn.org/wp-content/uploads/211009-GFN-Network-Energy-Effici

122、ency-1.0.pdf2 NGMN Alliance,Green Future Networks:NETWORK ENERGY EFFICIENCY PHASE 2,2023.Online.Available:https:/www.ngmn.org/publications/green-future-networks-network-energy-efficiency.html3 ETSI,ES 203 228 V1.4.1,Environmental Engineering(EE);Assessment of mobile network energy efficiency,20224 O

123、pen Compute Project,Design Guidelines for Im-mersion-Cooled IT Equipment,20205 S.Sarangi,E.D.McAfee,D.G.Damm and J.Gullbrand,Single-phase immersion cooling performance in Intel ser-vers with immersion specific heat sinks,in 38th Semicon-ductor Thermal Measurement,Modeling&Management Symposium(SEMI-T

124、HERM),2022 6 Intel,Intel Makes Key Investments to Advance Data Center Sustainability,2022.Online.Available:https:/ Grand View Research,Immersion Cooling Market Size&Share Report,2030,2022.Online.Available:https:/ K.Friderike,J.F.Adolfsen,E.M.Lis and A.Meyler,Energy price developments in and out of t

125、he COVID-19 pandemic from commodity prices to consumer prices,ECB Economic Bulletin,no.4,20229 Analysys Mason,Decommissioning legacy networks will be key to reducing operators energy usage,August 2022.Online.Available:https:/ vision of the NGMN Alliance is to provide impactful guidance to achieve in

126、novative and affordable mobile tele-communication services for the end user with a particular focus on supporting 5Gs full implementation,Mastering the Route to Disaggregation,Sustainability and Green Networks,as well as 6G.MISSIONThe mission of the NGMN Alliance is To evaluate and drive technology

127、evolution towards 5Gs full implementation and the three major priorities for 2021 and beyond:Route to Disaggregation:Leading in the development of open,disaggregated,virtualised and cloud native so-lutions with a focus on the end to end operating model.Green Future Networks:Building sustainable and

128、environmentally conscious solutions.6G:Emergence of 6G highlighting key trends across technology and societal requirements plus use cases to address.to establish clear functional and non-functional requi-rements for mobile networks of the next generation.to provide guidance to equipment developers,s

129、tandar-disation bodies and cooperation partners,leading to the implementation of a cost-effective network evolution to provide an information exchange forum for the industry on critical and immediate concerns and to share experiences and lessons learnt for addressing technology challenges to identif

130、y and remove barriers for enabling successful implementations of attractive mobile servicesNEXT GENERATION MOBILE NETWORKS ALLIANCENGMN,established in 2006,is a global,operator-led alliance of over 80 companies and orga-nisation spanning operators,manufacturers,consultancies and academia.2023 Next Generation Mobile Networks Alliance e.V.All rights reserved.No part of this document may be reproduced or transmitted in any form or by any means without prior written permission from NGMN Alliance e.V.