~~ HUAWEI yr el oy cr eee ray GREENSG W WHITE PAPER |sel tain Boley aN aae als CONTENTS add 1) eee 14 fel e]Ro- re Sor Tee Eee 01a] | aioe auilel ne dso Kabel Lc) iT Major Characteristics of Green 5G Networks ------ CS errand Co ee Co oe 4,4 Full-ifecycle Environment-Friendliness 05 [areas elle LaLa ier Cats 5.1 Direction 1: Multi-Antenna RF Reduces Per-Bit Energy Consumption and er uc i See eae ee ee ead Be Tt ee eer ena Prarie nine peepee eer sow or on Ca eee ee ee Gea cs] a at ST ee nee ee oa me aces Cae eee eee ey ee ed eee ee eco I Vision of Wireless Networks 2030 - - er eae apa vs ars aT) yaa 22 reyClimate change sa global challenge. Countries must cooperate closely to devel Ce aed Cs ue ee peer en ea ee ee et iy tee ie ee Ue eet er sees Pee aro ecm es un ese gS policies. As a developing country, China has pledged to peak its carbon emis sions by 2030 and become carbon neutral by 2060, After returning to the agree- eee eee Sa ent Cat Tas Pr nIi ac} The ICTindustryis the first to respond tothe Paris Agreement, moving both itself Bree ere enact te Communications contribute to non-ICT carbon emissions. As the world has ee eae Eee ewe ECT) voice cals and text messages to HD voice calls, mobile videos, personal applica- tions, and platform services in many new fields, becoming an essential part of ‘our daily lives. Such a diverse range of communications has always played an Cee ee ede ene Ee Le om ec structure to enable digital intelligence in both production and operations across. ‘a wide range of industries, helping them optimize processes, control precision, EO enue eae ee cus Decca ene aca Laas closer to the low carbon targets. In its report The Enablement Effect, GSMA showed that "with a European and North American scope, mobile technologies Eerie eee ren ee aa age ae ea Lee on ereenee Suen reduction of electricity consumption in other industries. GSMA believes this ratio. Pe ae S OCLs Rae eee eee care ae carbon neutrality. Many operators, such as Deutsche Telekom, Vodafone, France Telecom, Telefonica, China Mobile, China Telecom, and China Unicom, have given "green" growth ahigh priority on their strategic agendas, ‘Trends and interest towards carbon neutrality are growing, The TU hasmade the reduction of ICT carbon emissions by more than 45% a clear target for 2030. CeO E ee er cetera) eo ee eg etc ee eet eee ea Ee Soe eter oceans network energy efficiency and to explore new technical directions for green networks, helping achieve carbon peak and carbon neutrality.? Green 5G Leads to a Low-Carbon Future 5G networks are developing rapidly, More than 176 operators have deployed 56 networks, bringing 5G services to more than 460 million people worldwide. 50 of these operators have even deployed 5G FWA services. 5G has been applied extensively throughout vertical industries, and 5G home broadband subscrib- ‘ers now exceeds 1.2 million, While gaining fast momentum in ToC, ToB, and ToH, 5G networks continuously ‘evolve to better fulfil the increasing demand for network services. Operators need to develop networks that can deliver premium services to both individual and enterprise Users while supporting the ambition towards a carbon peakand carbon neutrality. This means that, in adai- tion to chasing high performance in experience and coverage, operators will ned to focus heavily on going "green" while they scale up 5G business. ‘We have moved to purpose a green network assessment system for energy efficiency and define major char- acteristics and eight technological directions for building green 5G target networks. These suggestions will be key to helping carriers achieve their strategic carbon goals while evolving their networks ‘We suggest energy efficiency (E) as the bass for assessing network energy efficiency. High device integration, site simplification, intelligence, and full-lifecycle environmental friendliness are the ‘our major characteristics of green networks. Inaddition to these, eight technological directions will help guide operators developing green networks. Y ss ey >>> ‘sGREEN 5G WHITE PAPER 3 Energy Efficiency: Basis of Green 5G Networks Energy Efficiency Assessment Spans Across a Network's Lifecycle Appropriate systems for indicating a network's energy performance and reasonable energy efficiency objectives should be set up during planning and construction to meet the evolution and green requirements. All technologies and solutions should also be planned and deployed to support these objectives. Energy efficiency assessment during network operations and optimization helps operators better understand their networks and ‘identify weak points in energy efficiency. This will help operators target the reconfigurations required! to enhance their energy-sav- ingdeployments. During network upgrade and reconstruction, once the areas with the lowest energy efficiency are recognized and the causes are found out, operators will better understand the potential for energy saving on their networks and will be able tofind the exact solutions needed to upgrade or reconstruct their networks. Multi-Dimensional Comprehensive Assessment ‘The key to assessment is to find appropriate indicators to describe a network's energy performance. Traditionally, energy efficiency is defined as the effective output of per unit energy usage. However, in the real world, traffic volume is used to evaluate the energy efficiency. That is, energy efficiency is measured based on the ratio of traffic volume to the energy used. ‘This is generally referred to as trafic energy efficiency. ‘Ample research has been conducted and practices explored for using traffic energy efficiency to analyze and improve net- work energy efficiency. New services and scenarios are emerging along with the develop- ment of networks, with each having different requirements on data rates. This requires operators to focus on not just growing traffic demand but also service exp. 2GEra | 3GEa | AGE! SG Era baa tse seesIn partnership with operators, we have analyzed the energy efficiency of mobile networks in many areas around the globe. One of our key findings is that users perceived vastly different. data rates from some networks even with a close level of energy efficiency. This shows that traffic energy efficiency, focusing on traffic volume, cannot precisely reflect the quality of services. Traffic energy efficiency with different user-perceived rates A 64 63 B 62 96 c eT 20.1 Therefore, the evaluation system needs to improve. We cannot rely simply on traffic volume and must expand to more dimensions, such as both traffic volumes and user experience. ‘A more comprehensive evaluation system will help operators evaluate network energy efficiency more appropriately based on development stages and scenario-specific service priorities. In this way, their mobile networks will be able to handle more data traffic while consuming less energy and improving user experience. eet 1.0 Ere Energy consumption Traffic volume _Energy consumption Traffic volume User experience ce, = —_Datvolume fe = Sete Ta brerenc) Energy Consumption Energy Consumption New types of ervices, such as URLLC? and mMTO, and new scenarios, such as networksslicing, are developing.GREEN 5G WHITE PAPER Category-based Objective Evaluation There are both objective and subjective factors affecting energy efficiency. The objective factors include population density, ‘weather conditions, and geographic environments, and the subjective factors refer to construction standards, architecture, hard- ‘ware configuration, and feature deployment. To minimize the impact of objective factors on evaluation, they should be normal- ized in advance or classified so that category-based objective evaluation will be possible to recognize weaknesses more precisely and create conditions for better evaluation results For example, operators classify their networks frst based on population density and then further divide the areas of a network based on site spacing before evaluation. As such, the networks in urban areas that have small site spacing are evaluated as one category, while the rural networks with a large spacing are evaluated as another category. Dynamic Changes of Network Energy Efficiency Network energy efficiency tends to fluctuate vastly during peak hours and off-peak hours n high-traffic hours, trafic energy eff- ciency is high, but the user-percelved rate might be quite low, and the opposite wll be true in hours with low traffic Network energy efficiency is also likely to differ as a network develops at different stages. In the early days, the traffic energy effi ciency generally stays at a low level. Later on, with growing traffic volume, the traffic energy efficiency gradually increases. Key Factors Affecting Network Energy Efficiency ‘The following are generalized categories of factors that affect network energy efficiency. Networking architecture: suchas the proportions of centralized equipment rooms and outdoor equipment roomsand band-varying spectrum polices Device efficiency: frequencies at which hardware is replaced, ultra-broadband and high- integration devices, and ener- gy-saving features. Radio performan interference and spectral efficiency and the use of high-gain antennas. Services and loads: traffic demand, camping ratio of 5G UEs, and service types and distribution. The key factors may differ depending on the scenario and area. Therefore, network energy efficiency requires area-specific evalua- tion to ensure accurate results,04 Major Characteristics of Green 5G Networks Poteet ret eer cee weGREEN 5G WHITE PAPER 4.2 Site Simplification An information flow cycle involves terminals, air interfaces, AAU'/RRU! main equipment, BBL’, transport networks, and core net- ‘works. Therefore, higher efficiency of information flows and less energy consumed by other devices and components will both help reduce site-level energy consumption. Site simplification aims to reduce the energy consumed by other devices. In a traditional D-RAN? site, each physical site uses an ‘equipment room, each requiring air conditioning, This leads to considerable energy consumption and it isnot productive. With sites simplified, operators need fewer equipment rooms and air conditioners, and power supply and transmission deployments are streamlined, helping boosting the overall efficiency at stes.. HEE ise RAN AlLoutdoor bladesite All-outdoorcabinetsite 4} semeronare . re ~ craintcen Re U ~ ia HA ne LS ‘tonal equpmentoemsandstes Site simplification primarily means simplified site construction. As traditional equipment rooms are replaced by cabinets and blade power supplies are used, site footprint is downsized by a huge margin to drastically decrease wasted energy. Based on China ‘Tower's statistics, after three cabinets were integrated intoa single one in China's southwest Guizhou province, the rent and elec- tricity bills reduced by CNY4,400 per month, cutting O&M costs by 7546. Blade power supply modules were deployed in Bejing and Langfang, improving site energy efficiency to 96% and reducing T00 kWh of cooling per year. They also saw a 50% reduction in the heat loss over cables. Site simplification also means streamlined power supply. AC/DC integrated power supply, multiple output modes in one module, and linkage between power supply and energy storage systems will help streamline power supply. Furthermore, site simplification covers simplified transmission. Indoor transport modules are replaced by outdoor ones to facil tate all outdoor simplified site deployment.4.3 Network Intelligence Network energy efficiency changes with service traffic as it enters peak and off-peak periods in a day. Software adjustment is the key to improving energy efficiency of off-peak hours by putting resources — such as carriers, channels, and symbols —tosleep and tweaking transmit power. Operators have long shown mixed attitudes towards software-based energy saving solutions: excit- ed at their potential for diminishing network energy usage but also showing great concern for their negative impact on network performance. PEER 2 Production > Trnsporation > Operation Sen Network sidedevice Userside Cee ena device Terminal product 1 1 1 Hi i +t + 1 0% 20% 40% 60% 80% 100% Recycling-oriented product lifecycle management helps the industry minimize its dependence on natural resources by selecting environment-riendly materials, using waste-minimization product design, optimizing packaging and transporta- tion, maximally prolonging service life, and using recyclable and easy-to-disassemble design. Recycling-oriented product fecycle management. Minimization —= optimized package Litt atest ‘and transport Inpvee rece Woigtiimersions Resse ee ceo Environment- Friendly Prolonged service life Renenable ae Moda design La ieor| Durability design Recycling Easyto disassembleEight Directions: Technical Development Trends of Green 5G Networks TEC a cu Ly emu Sr sa oe eg Se Cerrar em oe Ceca SEY Nee Nene ee Bcheens yi feT UB RMU reed ee iia aga Consumption and Increases Transmission Efficiency Cee renee oe oo enc eee ee ene a this consumption is the key to developing green 56 net- roe Figure 10 oe org a ro DTU Cer as 60% Senn COn ieee eS Cae ee Cure en Nts Ce De eee eon Peers ny Cie Figure 11 om aT Gall co ee Cee eo nee eco te eee to ice Memes cy eee eee ate een a) toconcentrate radio energy on narrower beams that point to Users more precisely, Increasing energy transmission effcien cyand greatly raising per-bit energy efficiency. In addition, we De ects en cy multi-disciplinary convergent innovation and improving ere aoe ee ee ee eee sumption. As networks continue to develop, this huge poten: Peculiar t for operators to meet continuously growing traffic demand Ge Mere ean test Meee ee a eee ed increasing, the total radio transmit power of base stations has Peet ae eae ene et cd Deets coiGREEN 5G WHITE PAPER {A1-dB increase in base station transmit power leads to.an approximate 10% increase in base station power consumption 100% Radio alr ‘conditioner Powersupply Others ns dono eA ening Improve energy Improve energy conversionefficiency transmission efficiency = = p> Massive MIMO AAU BEEBE nebo etna ry Power cansumption ~30% Y vessporer ‘consumes Time — Tradtonalassve MIM —— New Masive IMO Improving energy transmission efficiency s the new track for AAU energy saving development. The ultra-large antenna array — a new innovation created by combining progress in software and hardware, including new base- band algorithms and antennas — has been adopted to maximize the potential of antennas for energy savingand guarantee both better experience and coverage. Theoret cally, AAU products using the ultra-arge antenna array consume 30% less electricity while delivering the same level of coverage and experience for users in cell-edge areas. Trait Deicepover| Donte Uni consumption | coverage coverage vacuo ft tT: ‘ft > ‘The average user experience at the cell edge improvesby 20%, Single-UE rate — Traditional MassveMMo —— NewMassive MNO‘As networks develop, more bands are used, leading to De eee eu eget) Eee ieee PMU RCo Meee ee OT Le) Higher integration enables a device module to expand its RF See eee eco ee ecu eee) See cece) construction from deployment ofjust one band on one RRU or ‘AAU module to one integrated module that supports multiple bands. For operators, the benefit is that one module can be sed to provide various services that previously required mult ple modules to implement, helping lower costs. More impor- ee een tie see tc Seen ‘Take the six bands 700 MHz, 800 MHz, 900 MHz, 1800 MHz,2100 Tee eee een eet esa Cee ene eee eae car RRUs/AAUs are required in the ultra-broadband solution, with Cen cue eee asc Sees Cotes n eeu nee ee Td RF modules to provide services on the 800 MHz and 900 MHz See eee eee eet cece emus to provide services on the 700 MHz, 800 MHz, and 900 MHz bands with essentially the same level of power consumption. Sn ose gaat et ee Ce anne ersten(ita) WHITE PAPER Currently, RF hardware transmits with a higher eee ee es Bree Cee cee Cee eee y working mode and cannot automatically adjust Rete ies eee eee eee See eee eed frequency and baseband parts, are designed to een eee teed Nonlinear hardware efciency change ith Ou ee ena PEE evincicscicves components can be adjusted to work in sleep eee arate ate As such, the hardware power consumption changes nearly in direct proportion with load ae Sen ec eg (PA) largely contributes to the total power con- Bena hm Rey improve its efficiency in medium-load and light-load scenarios. Dee Omen demonstrated that deep dormancy helps ee occ eee ec) Noload —Lightload —Mediumload _-HeavyloadDirection 4: and Air Conditioners implified Sites Without Equipment Rooms There are a large number of D-RAN sites around the world. The energy consumption of air conditioners in equipment rooms accounts for approximately 40% ofthe site's total energy consumption, dragging down the overall site energy efficiency toa low level. @ With centralized BBU deployment, one air conditioner can be used for the cooling of multiple sites rather than ‘one for each as is the case in separate BBU deployment. To further save electricity, liquid-based and other natural cooling methods are good options for equipment rooms. Stuales in China have shown that 17,000 kWh of electrii- ty.can be saved at a single site each year with centralized BBU deployment. ne site, one cabinet” deploymentin China © “one site, one cabinet” and “blade site deployment” can be used as an alternative to one air conditioner at each site, With “one site, one cabinet’, cooling media works in close proximity to heat-producing devices, facilitating precise and targeted cooling to siash energy consump. tion. "Blade site deployment" focuses on saving energy by mounting devices on poles — a deployment mode that can maximize the effect of natural cooling to help remove air conditioners from sites. This will allow a 60% to 97% increase in energy efficiency on sites.GREEN 5G WHITE PAPER Direction 5: Linked Site Energy Improves Comprehen Energy Efficiency Traditionally, power supply, power storage, and power usage components ata site are unable to sense and collaborate with ‘one another or sense current service loads and running status. As a result, end-to-end power supply and backup efficien is low, which greatly increases the site's power consumption level ‘These components are now linked so that they can collaborate to adjust temperature and power operations in line with ser- vice loads, thereby reducing energy waste. For operators, this will help them better fulfill site energy saving targets. Clean energy has been gradually and efficiently introduced, Traditional sites are mainly powered by mains power or diesel generators. In the future, power sources will be more diversified and cleaner. New energy, especially solar power, will be gradually incorporated into site power supply. Linking services to solar energy and mains power will enable PV panels to work atthe highest efficiency and offer diversified optimal power supply solutions. The temperature in equipment rooms is adjusted based on service status and device capability. A 10° C increase in the tem- perature of an equipment room means a 20% decrease in the power consumption of that room, Power modules and backup power can be adaptively adjusted based on the predictions of power consumption and battery state of charge (SOC)/state of health (SOH)to achieve optimal staggered usage of mains electricity, thereby saving on elec: tricity costs, The intelligent staggering has already been implemented in Zhejiang, China, where it has helped the operator cut 17.1% electricity fees per site every yearDirection 6: Intelligence for Energy Saving and Network Performance Operators usually power off some active hardware when cells are lightly loaded to reduce power consumption. In the past, there were two energy-saving mechanisms for operators: carrier-level shutdown and channel-level shutdown. That is, some carriers or RF channels are shut down ata fixed time or when the load is below a certain threshold, thereby reducing the power consumption of equipment. However, mobile networks employ multiple radio access technologies and multiple bands. Coverage of multiple sites and cells may overlap. Coverage scenarios and service characteristics are different among cells. Therefore, shutdown thresholds need to be set and managed independently for each site and each cell. Otherwise, network quality and user experience cannot be ensured, Network energy saving solutions are becoming more intelligent. Al technologies are used to calculate shutdown thresholds and allocate network and spectrum resources in real time based on network mode, band, scenario, service characteristics, and network running status. In this way, the best energy saving solution can be employed, while still retaining optimal net- work performance. Al-based power saving solution: rT MM historical service load characteristics More celts for power saving after accurate Traffic volume forecast based on historical co-coverage analysis, traffic characteristics Last week’ trafic (PRB) Parameter optimization is ac the optimal trade (oie! “Traffic volume forecast (PRB)GREEN 5G WHITE PAPER Capacity-layer cell Coverage-layer cell CTE Luce Mito The intelligent network has been put into practice in Changsha, China. An intelligent network reduces its energy consump- tion by more than 25% while retaining network performance. The continuous development of intelligent networks means that energy consumption will also be on a constant downward trend,ncy 4G is seven to ten times more energy-efficient than 3G, while 5G is 20 times more energy-efficient than 4G. 5G protocols are more conducive to energy saving. For example, there is no CRS, the period of common signals can be adjusted, and UE-level power control is supported. The spectral efficiency of 56 is higher than LTE and will be continuously improved upon in the future. @ _56uses better coding schemes and polar codes to help reduce energy consumption. © Massive mimo beamforming effect ly improves the signal-to-noise rati (SNR) and energy efficiency. Thanks to @ wider bandwidth and better spectral efficiency, data transmitted per unit time in 5G is 10 times more than LTE, lowering per-bit power consumption, After services are migrated to 5G, the energy efficiency of the entire network is improved and energy consumption is reduced. The more services migrated, the more si icant the decrease in power consumption. Power consumption comparison of 2 four-band site before and after service migration to 5G o 1300 e =| £3 100 | ———| $3 — 23 900 Ee ZZ ee 700 A i 500 0% 10% 20% 30% 40% 50% Load — Powerconsunption before migration —— siisofserices re migatedtoSG © —— sso services re migatedto 5G Moreover, existing bands can be effectively refarmed for SGafter services are migrated to 5G to maximize its value. 5G network use cases in multiple places show that the continuous introduction of new 5G technologies drives the improvement of network energy efficiency. Exponential traffic growth that will be brought on by AR#/VR¥, smart factories, smart cities and other applications is inevitable, meaning that service migration o'5G will tap into its full potential for energy efficiency and minimize net- ‘workenergy consumption.Direction 8: Full-Lifecycle Recycling Reduces Dependence on Natural Resources The life cycle of ICT products, including design, production, ‘manufacturing, and use, involves energy consumption and carbon emissions. The recycling philosophy should be incor- porated into produc lifecycle management to minimize the dependence on natural resources and realize a low-carbon productlifecycle During manufacturing and transportation, environmentally friendly materials are used. Waste minimization design as well as packaging and transportation optimization is lever- aged to reduce resource consumption. The service life of a product is extended through modularization and durability design. Recyclable materials and easy-to-disassemble design make recycling easier, further reducing resource consump. tion and lessening the dependence on natural resources, ight recyclable plastic steel pallet6 Vision of Wireless Networks 2030 5G is ushering in an era of wireless communications that transform a wide range of industries at unprecedented depth and breadth. The commercial use of 6 is already spreading rapidly around the world, further improving communication capabilities and expanding not only the “human connection" butalso the "physical connection” between terminalsin a wide range of indus- tries. not only better connects people, but also establishes "oT" among terminals of various industries, marking the transforma tion of mobile communication from "connected people" to "connected everything’. In the next decade, new services will emerge, such as immersive cloud XR, holographic communication, sensory interconnection, intelligent interaction, harmonized communi cation and sensing, native intelligence, digta twin, and all-domain coverage. This will drive wireless networks to move beyond human and physical connectivity tothe smart connectivity of everything and bring about a hundredfold increase in network traffic cea! See ua Re a ea CR eno Ree should also be improved a hundredfold to avoid watts from increasing with bits. To achieve this goal, we need to exploreall-round energy-savingand carbon-reduction methods, ee ae oe ea airto-ground integration, near-field coverage, on-demand delivery, comprehensive energy supply integrated with wind and solar power, selfclosed Loop energy/consumption eet ee genet ee ea oy Ogee een Inthe future, through the continued introduction of green design concepts and native Al optimization capabilities, we are commit ted to improving end-to-end energy efficiency by 100x while ensuring ultimate performance and experience. This will help assist the ICT infrastructure and terminal devices in consuming no more energy than 5G, contributing to sustainable human develop- Dees comet eaGREEN 5G WHITE PAPER Acronym and Abbreviation PTO ECU Full Name Fixed Wireless Access 2 URLLC Ultra Reliable Low Latency Communication | | mMTC Massive Machine Type Communication 4 OPEX Operating Expense ia AAU Active Antenna Unit 6 RRU Remote Radio Unit ma BBU Baseband Unit 8 DRAN Distributed RAN on KPI Key Performance Indicator 10 AR ‘Augmented Reality = VR Virtual RealityGeneral Dislsimer Theiefomation inthis ocuren may cotanpesctivestatementinlding. without tation, sateen egg te tre fandl and operating ests future roduc potloe new tecnologise Thereaeanunberofactors that could cause ctl eats and developments to die materi om hose exprsed or pled inthe pedi tstrentsTeretr, eh information provided for reeence purpose only andconstiutes ether an ofr nr an acceptance, Huawel may change the Intrmaion atantime wtheuntice opyght :2021 HAVEL TECHNOLOGIES, LTO. Alihts Reserved Nopartol thi document may be reproduced or tranemitedin any frm or yay meanswthut rr writen consent Huse Technologies Co, id