5g 下一节：新的自动化层如何提高整体能效

INSIGHT SPOTLIGHT A shift is occurring from energy-saving solutions thatare configuration-based, semi-static andunsynchronised, to those that are intent-driven andable to orchestrate existing and new functionality. This This Insight Spotlight is the second part of a series–called 5G Next–to help better understand the nextphase of the 5G era and highlight promising technology innovations shaping networks. This edition exploresautomated solutions, also known as orchestration, Analysis Orchestrators are expected to integrate existing features from thetime domain (sleep and shutdown), spectrum domain (trafficsteering between frequencies to enable deactivation of carriers)and spatial domain (massive MIMO). For example, Ericsson has Energy-saving solutions can increase complexity Energy-saving solutions are already available for radio units,basebands, massive MIMO, data centres and passiveinfrastructure. Rapid developments in energy-saving algorithms,cheaper sensors and computing capacity, and soaring energy The impact on network economics A holistic network architecture and the concept of intent-drivennetworks offer significant potential for operators looking to improve As most of the energy used by network operators is in the RAN(76%, on average, according to GSMA Intelligence data), solutionshave naturally focused on this area. AI is delivering for operatorsmore capabilities and efficiency improvements than ever before.New solutions allow network equipment to understand traffic Opex Energy costs–According to GSMA’s Mobile Energy Efficiencybenchmarking project, mobile operators spend 20% of theiropex on energy. Orchestrators can identify a sweet spot to However, both the number of energy-saving solutions and their complexity have reached levels where operators need to re-examine their efficiency tools. Sleep modes, enhancedscheduling, optimal spectrum-layer selection, boosting and othersolutions can all work separately in silos. Alternatively, data andinsight can be shared between solutions to orchestrate and Simplifying operations–Operators spend a significant part oftheir opex on network operations, energy management andfault management. Orchestrators can use parameters and the How an intent-driven, energy performance orchestrator helps Capex Energy management is data-intensive. Operators cannot efficientlyprocess all the information and make real-time decisions at scalewithout the use of an automated, overarching platform. With a new,network-level orchestrator, they can connect and harmonisedifferent solutions, automatically share insights and simplifynetwork operations. They can also maximise network performance Purchasing and operating the orchestrator–Althoughautomation reduces dependency on labour, the solutions haveprocurement and transition costs, and require updates and new Revenue User experience: monitoring SLAs and monetising premiumlayers–Better quality of service and customer satisfaction canbe achieved through constant measuring of KPIs in real-time.This can support operators with their network-slicing ambitions Implications Mobile operators Vendors •Rethinking the approach to energy efficiency–Sinceenergy-saving solutions first came to market, operators haveused them as a bottom-up tool for specific areas. Networkengineers have configured solutions and then iterated theinput variables (network SLAs) based on outcomes (networkperformance, network availability or energy consumption). •A new approach for the ‘long tail’–Operators of differentsizes have different requirements for their network strategies.Larger operators can fall short of the necessary talent,knowledge and capacity to allow them to transform. Lack ofexpertise and capacity is even more of an issue for smaller •Country-level challenges require country-level solutions–Regulations regarding data residency vary by country. Localoperators and governments are protective of local customer •Ensuring the user experience–Energy-saving orchestrationsoftware responsible for multi-layer coordination of networkresources can heavily impact the user experience. Thesolutions find the optimum point where users receive whatthey paid for but with no equipment running unnecessarily. data and are reluctant to give full access to local data tovendors to run algorithms on their premises. Vendors should •Time to consider opening up–It is difficult to integratemany of the energy efficiency solutions currently available withthose from other technology providers (even those that aresimilar). Creating interoperable data pipelines, usingstandardised metrics, and offering compatible software is acomplex task, requiring cross-industry harmonisation and •Investing in AI and automation–There has been a lot ofhype around AI in telecoms, with high implementation costsand questionable tangible impact. In 2024, the differencebetween AI and the metaverse, for example, should be clear;AI is mature enough for operators to understand it will