《华为&TM Forum：2024新一代智能运维白皮书2.0（英文版）（82页）.pdf》由会员分享，可在线阅读，更多相关《华为&TM Forum：2024新一代智能运维白皮书2.0（英文版）（82页）.pdf（82页珍藏版）》请在三个皮匠报告上搜索。

1、1 Back to ContentsObjective:to explore ways in which communications service providers(CSPs)can enhance service-centric operations transformation,accelerated through the use of AI,and by identifying new value metrics and measurement methodologies.AN INDUSTRY WHITEPAPERVersion 2.0JUNE 2024New-generati

2、on intelligent operations:SPONSORED BY:the service-centric transformation path2 Back to Contents Axiata XL Syakil Bin ShahabudinChina Mobile Cheng Lu,Ding Dong,Lin Chongyun,Wang Chen,Jin Dongsheng,Zheng Yifeng,Peng Chenfa,An Jiujiang,Yang Chuan,Huang Jie,Xu Ming,Kong HuamingChina Unicom Wang Yu,Zhou

3、 Ying,Zhu Hong,Fan YanlinHKT Tom Pang,Kam Shing Fung,Otto LauOrange Emmanuel Chautard,Olivier Simon,Aroussia Maadi,Tristan Trepat-Marti,Gabriel Junghiatu,Asser el Nahas,Alexis Koalla,Linda Chong ChauvotHuawei Ye Rongchun(Kevin Ye),Yang Shengdong,Ouyang Yongjian(Henry Au Yeung),Wang Hailong,Luo Tianh

4、ong(Ken Luo),Jin Kaixu,Yu Yefu,Lan Yu,Gao Peng,Liang Yongyun,Mao Shijie,Charlene Wong,Goh Khiang Chew,Ahmed Omer Abdelnour Suliman,Weng Zaixin,Wei Yuzhuo,Deng Dan,Marc Bikok,Chenzhicheng,JiaodongfengTM Forum Richard Webb,Senior Analyst Teresa Cottam,Contributing Analyst(Omnisperience)CONTRIBUTING IN

5、DIVIDUALSContributors and contributing companiesCONTRIBUTING COMPANIES3contentsExecutive summary 5Section 1 Industry landscape:service-centric operations 71.1 From whitepaper 1.0 to 2.0:charting the network-centric to service-centric journey 71.1.1 Autonomous networks and autonomous operations 71.1.

6、2 AN and AO maturity models 81.2 Industry trends driving operations transformation 91.2.1 Market challenges and opportunities 91.2.2 AI for transformation acceleration 111.2.3 The importance of data 131.2.4 Using digital twins to evolve network and service monitoring 141.3 Architectures,tools and be

7、nchmarking 141.3.1 Frameworks for transformation 141.3.2 Importance of measuring operations value 15Section 2 Defining new values and metrics for service-centric operations 162.1 Operational transformation challenges and success factors 162.2 Evaluate,Operate and Transfer(E.O.T.)for service-centric

8、operations 182.2.1 Procedures and activities 182.2.2 Fundamental changes in the E.O.T.journey 192.3 Using TM Forum assets to measure business objectives and service effectiveness 192.3.1 Defining MAMA framework to guide capital resource allocation and development 192.3.2 AOMM provides guidance to tr

9、ansformation journey 192.3.3 Value Operations Framework(VOF)three-layer model and value tree concept 202.3.4 From silo SLAs to value measurement 212.4 Value perception of new value metrics for service-centric operations 222.4.1 The landscape of services and value metrics 222.4.2 Value metrics dictio

10、nary 232.4.3 Business objectives categorized into revenue and satisfaction 242.4.4 Value metrics of data mobile service 242.4.5 Value metrics of home broadband service 242.4.6 Value metrics of private line enterprise service 252.5 Detailed value measurement formulae 252.5.1 Business values for servi

11、ce centric operation(R.I.S.E.objectives)252.5.2 KPIs for monitoring and handling anomaly events+data service 252.5.3 KPIs for monitoring and handling anomaly events+home broadband service 262.5.4 KPIs for monitoring and handling anomaly events+private line service 264contentsSection 3 Transformation

12、 approaches to realize new values 273.1 Suggested transformation framework 273.2 Recap of whitepaper 1.0:key characteristics of service-centric operations 273.2.1 Digital enablement platforms and approaches key requirements 273.2.2 Technology evolution to overcome operational challenges 283.2.3 From

13、 as-is present method of operations to future method of operations 293.3 Key evolutions since whitepaper 1.0:new value,new technology,new applications 303.3.1 CSPs are missing key operations capabilities 303.3.2 New value 323.3.3 New technology 323.3.4 New applications 323.4 Utilizing Huawei assets:

14、new platform,new technology(GenAI,DTN,EDNS)323.4.1 The to-be digital enablement platform and new progress 323.4.2 Network-centric operations to service-centric operations require introduction of 3 new core technological features 333.4.3 The connection between the six dimensions model and 3 core tech

15、nological features 343.4.4 Detailed discussion on 3 core technological features:DTN,GenAI,EDNS 353.5 Latest developments:new applications enabled by new technologies 413.5.1 To the consumer(ToC)service assurance:adding agile service restoration to incident management to maximize service loss reducti

16、on 413.5.2 To the home(ToH)service assurance:Make FTTx service&network visible and manageable,and one fault one trouble ticket 423.5.3 To the business service assurance:service-level quality visualization and proactive assurance for enterprise-grade connectivity services 433.5.4 Role-based front off

17、ice and field maintenance engineer copilot 443.6 NOC-SOC transformation as related to people and organization 463.6.1 NOC&SOC operation target model 463.6.2 Best practices for achieving NOC-SOC collaboration 473.6.3 People:roles and responsibility changes 47Section 4 Case studies showcasing CSPs ser

18、vice-centric operations transformation paths 49Orange 49China Unicom 55HKT 60China Mobile 64XL Axiata 68IOH 72AIS 77Jazz 80TM Forum 2024.The entire contents of this publication are protected by copyright.All rights reserved.The views and opinions expressed in this white paper are provided in thecont

19、ributors personal capacities and may not reflect the views of their companies.While allcare has been taken in preparation of this paper,no responsibility for any loss occasioned toany person acting or refraining from any action as a result of any material in this publicationcan be accepted by the ed

20、itors,contributors or publisher.5Communications service providers(CSPs)are having to deal with massive increases in network and service complexity and scale,whilst facing challenging market conditions and ever-higher customer expectations.To succeed in the evolving digital economy,they are having to

21、 radically re-evaluate their network and service operations,business models and how they deliver value.But this is a fundamental opportunity too.By leveraging new technological capabilities such as automation and generative AI(GenAI),and by focusing on customer satisfaction,CSPs can transform their

22、operations to generate new efficiencies,create new services and deliver new experiences.Taking advantage of this opportunity requires them to evolve their operations from being network-centric to a more service-centric approach at the heart of which is autonomous operations(AO).Central to AO is auto

23、mating business operations through intelligent system-based decision-making and providing the structure to support overall digital transformation.An important foundation for AO is autonomous networks(ANs)the automation of network configuration,operation and optimization to deliver zero-touch,zero-wa

24、it and zero-trouble services.Progressing through levels of network automation can bring major benefits to CSPs in terms of efficiencies,network resilience and performance gains and is a significant factor in relating network capabilities to commercial outcomes,which to date have not typically been c

25、losely linked.Traditionally,CSP network operations centers(NOCs)have focused on network health and repair without correlation to the customers experience of their connectivity services.In an increasingly competitive and more digitalized world,customers end-to-end service experience has greater signi

26、ficance given the growing requirements created by digital media,and complex industry and business applications and thus greater value.This has led many CSPs to implement a service operations center(SOC),to concentrate on customer needs and delivering a more valuable experience.But to fundamentally i

27、mprove customer experience,collaboration between NOC and SOC must be seamless,overcoming disconnected silos of activity between network resource and service operations functions,sharing data insights to understand network performance and translating that to service outcomes.By focusing directly on p

28、erformance of commercial services,service-centric operations correlate the impact of the network on the customers quality of experience,as well as facilitating the prediction and resolution of problems before customers experience them.Executive summary Back to Contents|Executive SummaryProgressing t

29、hrough levels of network automation can bring major benefits to CSPs in terms of efficiencies,network resilience and performance gains and is a significant factor in relating network capabilities to commercial outcomes,which to date have not typically been closely linked.6 Back to Contents|Executive

30、 SummaryMeasuring the valueAnother vital part of operational transformation is understanding where value exists in different services.Being able to measure service values makes service-centric operations more powerful,as it enables the CSP to enhance service performance,with customer intent the prim

31、ary focus.In this whitepaper Huawei proposes new value metrics for service-centric operations together with methodologies for measuring those values to prioritize investment of resources and capital,based on TM Forums Value Operations Framework(VOF).New value metrics is a new-generation digital oper

32、ations framework comprising governance,service,enablement and system domains,plus guidance methods for digital transformation.Huaweis operations solution,which complements TM Forum architectures and maturity models(as we see in section 1),also leverages new technology capabilities such as GenAI,digi

33、tal twin networks and service impact assessment systems and evaluation algorithms.These serve as a benchmark for measuring the end-to-end effectiveness of service-centric operations.For CSPs,operational transformation is an on-going evolution which requires C-level management leadership and vision,o

34、rganizational flexibility and new skills.Service-centric operations utilize automation and GenAI capabilities and put the focus on customer intent,with service and network metrics based on customer-perceived quality and performance.As we set out in this whitepaper,this enables CSPs to take advantage

35、 of new service opportunities and better meet customer requirements,in turn positioning them to meet the challenges of this fast-changing digital world.Service-centric operations utilize automation and GenAI capabilities and put the focus on customer intent,with service and network metrics based on

36、customer-perceived quality and performance.7OverviewIn September 2023,TM Forum published a whitepaper,New-generation intelligent operations:the service-centric transformation path,launched at TM Forums flagship annual event,DTW23-Ignite,in Copenhagen.It explored ways in which communications service

37、providers(CSPs)can improve customer experiences by shifting from a network-centric to a service-centric approach,enabled by intelligent automation of network operations and maintenance(O&M).Transforming network and service operations is a foundation for profound digital transformation.CSPs can furth

38、er evolve the concept of intelligent automation by leveraging the power of AI to accelerate that transformation and measure the value of service-centric operations in customer experience.In this follow-up industry whitepaper Huawei proposes its holistic solution for measuring operational values and

39、relating these to customer experience,within service-centric operations.It discusses the identification of new value metrics and how these can be applied to specific service areas such as mobile,home broadband and enterprise connectivity,and used to guide capital resource allocation and service deve

40、lopment.The whitepaper shows how new value metrics and measurement relate to technology capabilities such as generative AI(GenAI)and digital twin networks(DTNs)and offers a roadmap for implementation which mitigates risks associated with service-centric transformation.Huaweis solution draws on and c

41、omplements TM Forum assets such as its Value Operations Framework(VOF)and Autonomous Operations Maturity Model(AOMM),Measuring and Managing Autonomy(MAMA)toolkit,and Open Digital Framework(ODF).Finally,through a diverse series of CSP case studies,the whitepaper illustrates a range of industry best p

42、ractices for CSPs making the journey to service-centric operations.1.1 FROM WHITEPAPER 1.0 TO 2.0:CHARTING THE NETWORK-CENTRIC TO SERVICE-CENTRIC JOURNEY1.1.1 AUTONOMOUS NETWORKS AND AUTONOMOUS OPERATIONS Our previous whitepaper noted that CSPs have traditionally managed their networks using network

43、 operations centers(NOCs).These focused on monitoring network-centric alarms,faults and performance,and typically did not correlate with service performance and customer experience.However,as CSPs needed more visibility of customers end-to-end service experience,many implemented a separate operation

44、s organization,the service operations center(SOC).The SOC has access to tools for service quality measurement and customer experience management and can monitor the quality of overall service,communicate with the customer about service status,and take rapid actions to rectify service degradation and

45、 outages that negatively affect service quality and experience.Organizationally,the NOC and SOC may be separate or combined into one entity,an NSOC,but that requires collaboration between the NOC and SOC teams which can result in the reliance on manual flows and handoffs between the two organization

46、s.Manual handoffs often create disconnected silos of activity within and between network resource and service operations levels and do not help resolve incidents or improve customer experience.Industry landscape:service-centric operations Back to Contents|Industry landscape:service-centric operation

47、s1.CSPs can further evolve the concept of intelligent automation by leveraging the power of AI to accelerate digital transformation and measure the value of service-centric operations in customer experience.8 Back to Contents|Industry landscape:service-centric operationsIn contrast,by focusing direc

48、tly on the performance of commercial services,service-centric operations correlate the impact of the network on the customers quality of experience,as well as facilitating the prediction and resolution of problems before customers experience service issues.Being able to measure service values in ser

49、vice-centric operations makes them even more powerful,as it enables the CSP to refine and enhance service performance.This places the customer at the heart of service operations,with customer intent being the primary driver of operations automation,and service and network metrics correlating to cust

50、omer-perceived quality and performance.Given the growing complexity of connectivity,value-added services and customer demands,service-centric transformation necessitates operational autonomy,referred to as autonomous operations(AO).Whilst autonomous networks(ANs)are defined as able to configure,oper

51、ate and optimize themselves without human intervention,providing zero-touch,zero-wait and zero-trouble services,AO includes automated business operations and the governance of operations,and provides the structure to support overall digital transformation.1.1.2 AN AND AO MATURITY MODELSA key element

52、 of TM Forums Autonomous Networks Project is a six-step maturity model that CSPs can use to measure their AN progress.Each AN level has a set of characteristics describing the evolutionary stage of the CSPs journey from fully manual to fully autonomous operations(see graphic right).Fully autonomous

53、network:The system has closed-loop automation capabilities across multiple services,multiple domains(including partners domains)and the entire lifecycle via cognitive self-adaptation.Highly autonomous network:In a more complicated cross-domain environment,the system enables decision-making based on

54、predictive analysis or active closed-loop management of service-driven and customer experience-driven networks via AI modeling and continuous learning.Conditional autonomous network:The system senses real-time environmental changes and in certain network domains will optimize and adjust itself to th

55、e external environment to enable,closed-loop management via dynamically programmable policies.Partial autonomous network:The system enables closed-loop operations and maintenance for specific units under certain external environments via statically configured rules.Assisted operations and maintenanc

56、e:The system executes a specific,repetitive subtask based on pre-configuration,which can be recorded online and traced,in order to increase execution efficiency.Manual operations and maintenance:The system delivers assisted monitoring capabilities,but all dynamic tasks must be executed manually.0143

57、25Autonomous network levelsTM Forum,2023Figure 1:TM Forums six-step AN maturity modelFigure 2:TM Forums Autonomous Operations Maturity Model9 Back to Contents|Industry landscape:service-centric operationsTM Forums Autonomous Operations Maturity Model(AOMM),meanwhile,defines CSP operational evolution

58、(see graphic on the previous page).It enables CSPs to assess their performance in applying advanced automation techniques to their value streams by using a maturity grid assessment framework.AOMM helps CSPs understand their maturity level,set the next level target,and identify and subsequently devel

59、op a list of prioritized capabilities to accelerate their operational transformation.AN provides an important foundation for AO and could be a fundamental part of a CSPs overall digital transformation strategy.However,this whitepaper predominantly focuses on service operations,and therefore mainly r

60、efers to the AOMM as the benchmark for operational transformation,rather than the AN maturity model.Some of the CSP case studies included in section 4 refer to aspects of NOC automation that have positively impacted service operations.AN and AO progress pathways have been accelerated by machine lear

61、ning and AI,but the shift to service-centric operations requires a“multidimensional”approach to transformation.As detailed in the previous whitepaper,this encompasses:People CSP culture,organizational structure,collaboration,skills,job functions Processes evolving manual processes to automated,end-t

62、o-end services,tying network and customer processes together,and developing new relevant performance metrics Platforms enabling the digitalization of a CSPs businesses,including operations Technologies including AI,digital twins,intent-driven operations,and API-based access.1.2 INDUSTRY TRENDS DRIVI

63、NG OPERATIONS TRANSFORMATION1.2.1 MARKET CHALLENGES AND OPPORTUNITIESTo reap the most benefit from 5G and future technologies,CSPs must radically rethink the way they operate in both the network and service domains.Whilst better coordination and correlation of network and service issues through tigh

64、ter collaboration of network and service operations brings real benefits to CSPs,if they are to improve customer experience they need to reimagine their network,cloud and IT infrastructure as digital enablement platforms:AI-native and intent-based.Evolving from the NOC-SOC collaboration approach to

65、a truly service-centric operational approach requires the replacement of siloed,overly manual processes and inflexible architectures with a more radical operational combination of frameworks,architecture,systems and tools.These must enable the aggregation and analysis of customer-led,service-orienta

66、ted data,and derive actions which improve customer experience.CSPs must also enable a fail fast,learn,refine environment for application development and new service creation,so they can dynamically customize services for specific customers.The appeal of platform models to CSPs is clear,but they requ

67、ire automated(and standardized)service-centric operations as a vital foundation.For example,5G slicing in mobile networks will include customer-specific requirements for quality of experience.But to fully take advantage of this opportunity the telecoms industry must drive end-to-end standards.CSPs n

68、eed to agree on a common architecture and open interfaces in order to automate ordering,provisioning,orchestration and assurance of services across multiple domains(RAN,core,edge,IP,optical)and reduce operational and service complexity.Rising complexityCutting complexity across the business is key t

69、o helping CSPs address a range of challenges,which include:technical investment debt;an adverse macroeconomic environment;industry regulation that can inhibit their ability to partner;cultural issues related to virtualization and disaggregation of the network;a lack of software skills;and an increas

70、ing need for standardization.Whilst better coordination and correlation of network and service issues through tighter collaboration of network and service operations brings real benefits to CSPs,if they are to improve customer experience they need to reimagine their network,cloud and IT infrastructu

71、re as digital enablement platforms:AI-native and intent-based.10 Back to Contents|Industry landscape:service-centric operationsThe rising complexity of networks and operations is driving the need for cloud-based networks that rely on intent-based management enabled by AI.AN gives CSPs the ability to

72、 scale whilst intent-based management helps CSPs manage thousands of dynamic 5G network slices or millions of IoT devices using automated processes.Another reason to deploy ANs is the promise of cost savings,particularly to reduce operational expenditures related to energy usage,as well as to reduce

73、 carbon footprint.How AI is driving operations efficiencyAt DTW23-Ignite,China Mobile championed a Moonshot Catalyst project demonstrating several approaches to reducing energy consumption in 5G cell sites and equipment rooms including use of solar and wind power to supply 5G base station equipment

74、and provide backup power.The project which won the award for outstanding Catalyst,use of TM Forum assets also focused on development of an AI intelligent power-saving model,with an aim to achieve a 10%reduction in energy consumption in 5G base stations.But there are key challenges CSPs must overcome

75、 if they are to transform to service-centric operations.These include:Increasingly stringent,deterministic service performance requirements Rapid increases in network and service complexity Growing,continuous need for domain model algorithms and knowledge Scenario-based development to support agile

76、services Open platforms and ecosystems to accommodate multi-vendor,heterogeneous IT and network environments.Huawei has proposed six dimensions of technology evolution needed to address these challenges and support service-centric digital operation,leveraging the capabilities delivered by the digita

77、l enablement platform outlined in the previous whitepaper:Deterministic network assurance(DNA)Hyper Automation(HA)Model and algorithm of telecom(MAT)Telecom knowledge platform(TKP)One Trustworthy DevOps(OTD)Ecosystem Enablement(EE).Section 3 has a detailed review of these challenges and the technolo

78、gy evolution needed to address them,showing how they can be applied to operational transformation.Closing the skills gapAnother challenge to service-centric operations transformation highlighted in the previous whitepaper was that of skills within the CSP.In a software-orientated world,CSPs need acc

79、ess to a range of data science and software engineering capabilities if they are going to move fast enough to build the platforms and solutions they need for service-centric operations,as well as understand the methodologies for deploying AI,automation and analytics technologies.Resolving this skill

80、s gap requires digital leadership,which includes C-level vision,organizational flexibility,and equipping experienced operations personnel with the necessary training to execute the transformation.Many cloud-native operations concepts such as DevOps,AI-based operations(AIOps)and intent-driven automat

81、ion have come from IT operations that are not linked to a particular vertical such as telecoms.Working with professional services partners is one way to help CSPs evolve toward service-centric operations and lower the risk of the transformation journey.In a software-orientated world,CSPs need access

82、 to a range of data science and software engineering capabilities if they are going to move fast enough to build the platforms and solutions they need for service-centric operations.11 Back to Contents|Industry landscape:service-centric operations1.2.2 AI FOR TRANSFORMATION ACCELERATIONSimplificatio

83、n of digital operations is a mantra CSPs have been repeating for a decade.The drivers have not changed,though they sometimes shift in importance:improve customer experience,reduce costs,shorten time to new revenue.This is setting the stage for a big leap forward for intent-based automation and AI,re

84、imagining architecture for the future with AI at the center.AI is already enabling CSPs to respond rapidly and efficiently to customers inquiries triaging problems faster and augmenting agent performance by suggesting the next-best action.Since the previous whitepaper,one of the biggest changes in t

85、he industry has been the disruptive impact of generative AI(GenAI),which has seen rapid adoption across CSP resources and activities,driving change in platform capabilities and processes.Perceptions of GenAI have expanded rapidly from an AI-based content creation tool towards a strategic platform th

86、at is fast becoming front and center of thinking for CSPs worldwide.While much emphasis has been placed on CSPs leveraging GenAI for customer experience(CX)use cases,such as customer-facing chatbots,AI in general is now moving from optimizing reactiveness to supporting proactivity,becoming increasin

87、gly predictive.This has dynamic potential for operations too:not only giving the means to detect problems and then proactively resolve them before the customer is aware,but also to predict where problems are most likely to occur in the future and prevent them.For example,AI might be used to predict

88、where congestion is likely to occur in a network and prioritize network expansion to avoid the problem.Using AIOps for efficiencyAIOps combines big data and machine learning to automate digital operations processes,including event correlation,anomaly detection and causality determination,and is at t

89、he heart of CSP usage of AI for operational and energy efficiency.Use case examples include:Network optimization and management.GenAI can help CSPs improve network performance and reliability by generating code and analyzing data to detect problems.For example,a large language model(LLM)can generate

90、 network topology diagrams and learn patterns from historical data to identify anomalies and the root cause of problems.It can also automate troubleshooting and predictive maintenance.Field operations.GenAI can use data to enable dynamic and real-time routing to optimize routes for technicians.This

91、reduces travel time and fuel costs and improves service delivery.AI can also be used for NOCs,SOCs and field operations.By leveraging machine learning and graphic representations of the network,CSPs can drill down into a service issue and feed information to technicians so they can be efficient in p

92、roblem resolution.Security and fraud detection.GenAI can analyze terabytes of network data to detect anomalies,enhance security and improve fraud detection.For example,it can identify fraudulent activity such as SIM swap attacks or unauthorized access to customer accounts.Writing software code.GenAI

93、 models can be trained to provide code completion suggestions as developers write code or to generate entirely new code.For example,a model could analyze context,infer a developers intentions,and generate suggestions for completing code fragments,or it could be used to create code from scratch.Indee

94、d,AI is now an ever-present theme within the telecoms industry.In the following sections the whitepaper explores how CSPs can harness GenAI-derived capabilities to transform their service-centric operations,through new value measurement and enhanced customer experience.While much emphasis has been p

95、laced on CSPs leveraging GenAI for customer experience(CX)use cases,such as customer-facing chatbots,AI in general is now moving from optimizing reactiveness to supporting proactivity,becoming increasingly predictive.12 Back to Contents|Industry landscape:service-centric operationsBut implementation

96、 of GenAI within CSPs to drive operational transformation has challenges too.GenAI works by processing huge volumes of data to find patterns and determine the best possible response to a question or situation,which it then generates as an output.By feeding AI immense amounts of data,it is able to de

97、velop an understanding of correlations and patterns within the data.Whereas conventional approaches to machine learning require data scientists to develop AI from scratch,GenAI involves the use of foundational models deep-learning neural networks as a starting point to develop models that power new

98、applications more quickly and cost-effectively.Experimenting with,and adopting,GenAI for specific use cases is very different from what CSPs have used in previous versions of AI.With GenAI,outputs can be created immediately,but the challenge then is to decide how much fine-tuning is needed to improv

99、e the results and what to do with the output.There are also potential problems and risk factors associated with GenAI(see chart below).These include:Shadow AI(non-authorized employee usage)Privacy and security Phishing and fraud Lack of truth function Misinformation and manipulation Bias and fairnes

100、s Legal considerations Leaks of proprietary data.Read our Benchmark report to find out more about CSPs AI strategies and challengesBENCHMARKTM Forum|March 2024building anAI strategy:foundations in placetelcos put theAuthor:Mark Newman,Chief Analyst Editor:Ian Kemp,Managing EditorSponsored by:What ar

101、e the main risks for your organization of using GenAI?Big riskSmall riskShadow AI(i.e.non-authorized employee usage)Privacy and security(e.g.leaks of sensitive data via GenAI apps)Phishing and fraud(e.g.phishing emails or deep fakes)Lack of truth function(i.e.creating factually incorrrect answers)Mi

102、sinformation and manipulation(e.g.creation of fake content,documents)Bias and fairness(because real-world data is often biased)Legal considerations(through use of public data)Leaks of proprietary data48%52%20%80%36%64%26%74%38%62%34%66%26%74%27%73%TM Forum,2023Whats more,regardless of where the LLM

103、is running,customizing pre-trained models and tailoring them with CSPs(often proprietary)data requires robust data infrastructure and governance to ensure that the foundation model is trained on high-quality data.Even when the right safeguards and governance are in place,when GenAI is deployed at sc

104、ale,and across the organization,new systems and more agile processes may be needed.Figure 3:Risks to CSPs of using GenAI13 Back to Contents|Industry landscape:service-centric operations1.2.3 THE IMPORTANCE OF DATAAI is reliant on data.In order to fully exploit the potential of AI,operational transfo

105、rmation must prioritize making data the strategic and operational lifeblood of the CSP.CSPs typically organize their digital operations around linear software processes managed by separate teams:fulfilment,service activation,assurance and so on.These silos exist in every domain,but leverage data in

106、a piecemeal fashion,yielding limited benefits.As automation is introduced,the same software that sets up a service or a network element also monitors and heals it and handles scaling.Breaking down operational silos needs a holistic approach to data,across all systems and processes.Essentially,this m

107、eans acting as a single entity,and not functionally separate in terms of where data is processed and resides in the organization.Within CSPs this means evolving from vertically separate,tactical implementations of data to a horizontally integrated model in which data systems can talk to each other a

108、cross domains.Bringing network,operations and customer-centric data together in a single repository gives each business function within a CSP increased visibility of overall activities,from ordering to fulfilment,orchestration,management,assurance,optimization and billing.There are many early use ca

109、ses in GenAI particularly ones that use copilots,virtual assistants used to boost productivity and efficiency that exploit unstructured data.But most GenAI use cases will still need structured data from CSPs support systems the majority of GenAI use cases need BSS/OSS data.The challenge for operator

110、s is how to integrate this sensitive customer data with the data that sits in public LLMs,and how to ensure its accuracy given that it is being drawn from both public and private sources.Access to and quality of data is still by far the top AIOps challenge,according to vendor and CSP respondents to

111、TM Forums Digital Transformation Tracker 7 report(see chart).A data-driven architecture is key to maintaining precision and consistency across domains within the CSP,using common data models to ensure data flows smoothly through all systems and accurately feeds all automated workflows.TM Forums Open

112、 APIs are central to helping with this need,freeing data from silos so that it can be used across different parts of the CSP business.A data-driven architecture is also critical to the ability to measure value in service-centric operations.A data-driven architecture is key to maintaining precision a

113、nd consistency across domains within the CSP,using common data models to ensure data flows smoothly through all systems and accurately feeds all automated workflows.Challenges to implementing AIOpsSuppliersCSPsInconsistent&fragmented dataLack of data analytics expertiseExplainability&governanceLack

114、of mature network components&support systemsLack of standards for end-to-end managementConcerns about securityLack of software expertise74%78%52%53%50%39%41%33%31%33%27%30%27%31%TM Forum,2023Figure 4:Data is still by far the top AIOps challenge14 Back to Contents|Industry landscape:service-centric o

115、perations1.2.4 Using digital twins to evolve network and service monitoringThe digital twin concept offers a powerful framework for service-centric simulation,prediction and diagnosis,to improve operational processes,mitigate traffic loss,detect faults and enhance quality and accuracy of services.In

116、 essence,a digital twin is a digital or virtual representation of a physical object used by businesses for practical purposes,such as simulation,testing,monitoring and maintenance.Digital twins provide an immersive environment that enables CSPs to monitor network operations in real time and predict

117、events that might affect service quality,such as high traffic loads or network failures.They can also be used to build topology restoration and network resource models,network and operations state models,and to simulate dynamic service and network resilience supporting both performance optimization

118、and predictive mitigation of service disruption.TM Forums Digital Twin for Digital Intelligence(DT4DI)initiative aims to define an industry standard decision intelligence framework that integrates digital twin,AI and other technologies with business processes to help CSPs analyze,diagnose,predict an

119、d make decisions faster,more consistently and more accurately.The potential importance of digital twins as a tool for leveraging network planning,performance monitoring and fault remediation,and in measuring value in service-centric operations,is explored in detail in section 3.1.3 ARCHITECTURES,TOO

120、LS AND BENCHMARKING1.3.1 FRAMEWORKS FOR TRANSFORMATIONCSPs need to be able to orchestrate zero-touch,zero-wait and zero-trouble services end to end across network domains,including their partners,at scale.This requires automation of the entire service lifecycle,from ordering to fulfilment,activation

121、,management,assurance,optimization and billing.TM Forums Open API program was born out of this need.But end-to-end automation has remained elusive because of interoperability issues,architectural fragmentation and customization.Based on TM Forums definition,Level 4 autonomous networks mark the trans

122、ition between traditional automation of human-defined process behavior and autonomous behavior,in which systems make decisions independent of humans.Today,most operators are between Levels 2 and 3,but some including China Mobile,China Telecom,China Unicom,MTN and Orange are aiming to achieve Level 4

123、 autonomy for some processes by 2025.The TM Forum Open Digital Architecture(ODA)provides a migration path from legacy IT systems and processes to modular,cloud-native software orchestrated using AI.The framework comprises tools,code,knowledge and standards(machine-readable assets,not just documents)

124、.Developed by TM Forum member organizations through Collaboration Community and Catalyst proofs of concept,ODA is being used by leading service providers and software companies worldwide.In December 2023,TM Forum launched a new data reference architecture for the telecoms industry that aims to encom

125、pass both new AI-enabled business models and network operations.The new project reflects network operators need to build AI-driven autonomous processes,put data and AI at the core of operations,and bring external data into their meta-ecosystems.It also sets out to tackle the question of what a moder

126、n data architecture should look like for AI-enabled telecoms operations.The Modern Data Architecture for Telecom Operations Project was launched in collaboration with eighteen CSPs,in addition to suppliers and systems integrators.REPORTSponsored by:leveling up:Author:Mark Mortensen,Contributing Anal

127、ystEditor:Dawn Bushaus,Contributing Editor achieving Level 3 autonomous networks and beyondAugust 2023September 2023|www.tmforum.orgAuthor:Dr.Mark H Mortensen,Contributing Analyst,TM ForumEditor:Dawn Bushaus,Contributing Editor,TM Forumsponsored by:REPORTSponsored by:leveling up:Author:Mark Mortense

128、n,Contributing AnalystEditor:Dawn Bushaus,Contributing Editor achieving Level 3 autonomous networks and beyondAugust 2023Read the report to find out more about operators AN progress:15 Back to Contents|Industry landscape:service-centric operationsODA includes:An architecture framework,common languag

129、e and design principles Open APIs exposing business services Standardized software components A reference implementation Guides to navigate digital transformation Tools to support the migration from legacy architecture to ODA Maturity models and readiness checks to baseline digital capabilities.Sect

130、ion 4 of this whitepaper features case studies contributed by CSPs which showcase their experiences and successes in transforming their digital operations.As well as using ODA to underpin these strategies they are using new technology capabilities such as GenAI and digital twins to drive evolution,a

131、nd using new metrics to determine the value of operational change on customer experience.Taken as a whole,these case studies demonstrate how CSPs can leverage AI to create compelling new services and deliver customer satisfaction.1.3.2 IMPORTANCE OF MEASURING OPERATIONS VALUE The tech-focused approa

132、ch of digital transformation,with its goals of cost savings and process optimization,has begun to evolve into a more human-centered focus that sees technology and processes delivering richer,transformative experiences.This new approach promises to empower human experiences,interactions and relations

133、hips while supporting a new way of thinking and a cultural shift that puts customers and their needs at the heart of everything.The upshot is that transformations now need to deliver against all these goals:improving customer experience,reducing costs and speeding time to new revenue.As the transfor

134、mation of service-centric operations progresses,service impact measurement is getting more important,both in terms of network availability/efficiency such as for resource-saving related network KPIs and service performance such as for service level assurance(SLA)compliance and service key quality in

135、dicators(KQIs).Identifying new value streams and metrics for service-centric operations and creating a value measurement methodology forms the central theme of this whitepaper.These topics are analyzed in depth in sections 2 and 3.Figure 5:TM Forums ODA provides a migration path for digital transfor

136、mation16 Back to Contents|Defining new values and metrics for service-centric operationsDefining new values and metrics for service-centric operations2.OverviewThe previous section outlined current market drivers and challenges for service-centric operations and underlined the importance of AI and d

137、ata in accelerating operational transformation.This section looks at transformation pathways for CSPs and identifies values and metrics by which they can define their business objectives and measure the success at the service and business level.In particular,it shows how CSPs can evolve their transf

138、ormation beyond the traditional focus on network operations which mainly concentrate on efficiency,network quality and opex to more service-centric,outcome-based KPIs,which relate more to revenue and customer experience.2.1 OPERATIONAL TRANSFORMATION CHALLENGES AND SUCCESS FACTORSAccording to cross-

139、industry analysis from Boston Consulting Group,only around 30%of enterprise high-level management believes digital transformation has succeeded,in terms of meeting or exceeding goals and achieving a sustainable impact on the organization.Based on this research,some of the key digital transformation

140、challenges encountered by enterprises include:04204080 Success phase 2less than successfulPercentage of Cases 1702644%30%26%Limited value created(50%of the target)and no sustainable changeValue created but not achieved,limited sustainable changeReaching or exceeding the target value;Creat

141、es sustainable changeWinWorryWoeIndustry-wide,only 30%of enterprises have achieved digital success1.A comprehensive set of cases from internal and external data sets(n=895)2.Success scores are calculated based on the percentage of goals and values achieved,whether goals were achieved on time,and the

142、 results achieved compared to expectations.and comparison with other successful transformations 3.110%of the target value vs 66%of the target value.Percentage of the target value,for example,110%to 66%of the target value.4.For example,by upgrading the skills of existing staff or introducing new staf

143、f.Source:BCG AnalysisFigure 6:According to BCG only 30%of digital transformations are successfulAccording to cross-industry analysis from Boston Consulting Group,only around 30%of enterprise high-level management believes digital transformation has succeeded,in terms of meeting or exceeding goals an

144、d achieving a sustainable impact on the organization.17 Back to Contents|Defining new values and metrics for service-centric operations Lack of alignment with the business strategy,resulting in divergence between the strategy and transformation objectives and targets Lack of measurable indicators to

145、 demonstrate the outcome of transformation Ambitious plans with all-in-one solutions Lack of phased,step-by-step or progressive schedules,resulting in an imbalance between the outcomes and effort/investment Lack of process and service integration/consolidation,which introduces segments and breakpoin

146、ts causing back and forth between different organizations and eventually weakening the outcomes Investment in platforms and technologies but overlooking the development and transformation of people and processes.To implement an effective and efficient evolution from traditional network-centric to se

147、rvice-centric operations,Huawei proposes the design of a reference architecture to determine the objectives,scope,processes,models and technologies within which new values and metrics are established.This can help ensure that values align with strategic priorities within the CSP and can be applied a

148、gainst a Value Operations Framework(VOF)to implement,operate and measure the solutions against outcomes and return on investment to close the loop of operation transformation.E:EvaluateO:OperateT:TransferMetricsKBIInstance of R.I.S.E business objectivesKPIService EffectivenessE.Efficiency Improvemen

149、tOPEXOperate NEs per CapitaNetwork QualityR.I.S.ENetwork O&M(Efficiency)ComplaintService PerformanceFaultMBBFBBEnergy/DCNOCFLMEnergyAIService-centric Operation(E+RIS)HBBDATAPRIVATE LINEToCToHToBComputingR.Revenue Loss ReductionI.Innovation ImprovementS.Customer Satisfaction ImprovementRevenue Assura

150、nce(Churn,Data Pack)Service Availability(Data)Service Loss(Data)Maturity EvaluationSolution DeliveryCustomer Survey&Questionnaire Scoring for network faultsRepetitive Customer Complaint reductionBusiness result-oriented closed-loopGB1040&IG1291Align with CSP value chain for business operationService

151、 Effectivenessclose-loopIG1291&IG1294Service AvailabilityCompliance with service contractual SLA Extent of User&Service ImpactVOFGB1041&IG1292From single E to E+RISECorrectivePreventiveE.O.T Transformation JourneyFigure 7:Transformation reference architecture of service-centric operationsThe referen

152、ce architecture shown in the graphic covers four key success factors:Evaluate,Operate,Transfer(E.O.T.)a new roadmap for successful transformation.An end-to-end process to translate CSP strategy into measurable values using the VOF value tree,and implement/transfer transformation solutions,ensuring a

153、n effective transformation from network-centric to service-centric operations.Using VOF to measure new values(business objectives and service effectiveness).This is the key focus of service-centric operation accountability,supporting the business objectives in terms of revenue,satisfaction and innov

154、ation,creating an operations closed loop.End-to-end process and capability.A consolidated value stream to provide a holistic view for services such as monitoring and rectifying anomaly events,including fault management,performance management and customer complaints processes and eliminate process an

155、d capability breakpoints.Outcomes commitment.After the holistic and accurate evaluation,design and deployment of the transformation,the service outcome KPIs should now be recognized as value and benefit commitments for senior leadership.18 Back to Contents|Defining new values and metrics for service

156、-centric operationsSection 2 of this whitepaper explores each success factor within this reference architecture and identifies new value metrics and how they can apply to service-centric KPIs.2.2 EVALUATE,OPERATE AND TRANSFER(E.O.T.)FOR SERVICE-CENTRIC OPERATIONS2.2.1 PROCEDURES AND ACTIVITIESE.O.T.

157、is designed as a roadmap for service-centric operations,to translate business strategy,define an appropriate value stream,value tree,as-is baseline,to-be objectives,and implement relevant solutions in a phased and manageable approach.Figure 8:Evaluate,Operate and Transfer journey for service-centric

158、 operationsThe table below sets out descriptions and activities for each of the E.O.T.procedures.ProcedureActivityDescriptionE:EvaluateAlign with strategy,transformation evaluationUnderstand the CSP strategy:n Base the transformation on senior-level strategic requirementsn Detailed understanding of

159、strategy background,key indicators of business objectives,expectation,plan,timeframe etc.Translate strategy into measurable business objectives(KBIs)and service outcomes(KPIs),establishing the baseline of the transformation:n Align with the strategy and determine scope of consultancy,deep inside to

160、stakeholders,processes,procedures,platform and resourcesn AO to evaluate and assess the as-is autonomous maturity of each value stream phase and activityn VOF to evaluate the as-is business and service outcomes of value streamsGap analysis,identify transformation targetsBased on the assessment resul

161、t and values collected:n as-is gaps identify gaps against the CSP strategy,industry baseline and/or best practice,and analyze to determine the root causen to-be targets determine realistic,achievable and progressive targetsPhased transformation roadmap and metrics designDesign short-term and long-te

162、rm roadmap through a manageable and phased approach,and reach an agreement at senior level:n Design the value metrics to be achieved in each phasen Preliminary high-level design of the solutions to achieve the targetsn Create the phased plan and schedules for implementationn Utilize VOF to create a

163、visible value tree for investment justification.O:OperateTransformation solution designn Low-Level design of the solutions based on the roadmap,and reach agreement on processes,platform/systems and people transformationn Create the implementation planTransformation solution deployment&trainingn Impl

164、ement value metrics and solutions according to the plan and schedulesn Value show-case and return on investment of the transformation in each phasen Continually drive the optimization of the value metrics to maximize the business and service outcomes of value streamsT:TransferEmployees,knowledge and

165、 system transfern List of materials to be transferredn Process and platform trainingn Operation guidelines,system user manual etc.Post-transfer CSIn Shadowing and continual monitoring and analysis of the outcome of the transformation for one monthn Provide consultancy and support CSP service improve

166、ment19 Back to Contents|Defining new values and metrics for service-centric operations2.2.2 FUNDAMENTAL CHANGES IN THE E.O.T.JOURNEYOverall,the assessment follows the guideline of TM Forums MAMA-defined CSP value stream for autonomous operations IG1291&IG1292,which outlines the following key changes

167、 to steer operational transformation direction and outcomes:Expansion of key stakeholders at high-level management.Unlike traditional network operation,the evaluation and assessment of service-centric operation focuses on requirements such as churn reduction,NPS,margin and so on.End-to-end view of v

168、alue streams and business activities.Targeting the requirements of CXO strategy,all relevant operations processes are evaluated from end to end to identify elements and breakpoints the key to achieving KPIs and R.I.S.E.(revenue,innovation,satisfaction,efficiency)business objectives.Incremental value

169、s.Identifying potential improvements in value streams,based on feasibility and preferences,including the operational performance measurements(OPMs),KPIs and R.I.S.E.objectives.Figure 9:Expansion of key stakeholders of service-centric operation2.3 USING TM FORUM ASSETS TO MEASURE BUSINESS OBJECTIVES

170、AND SERVICE EFFECTIVENESS2.3.1 DEFINING MAMA FRAMEWORK TO GUIDE CAPITAL RESOURCE ALLOCATION AND DEVELOPMENTA well-established framework is essential to guide digital capital resource allocation for CSPs strategic development plans and measure transformation success.TM Forum members have defined the

171、Measuring and Managing Autonomy(MAMA)framework to provide a systematic approach to digital capital allocation with governance in place.2.3.2 AOMM PROVIDES GUIDANCE TO TRANSFORMATION JOURNEYAutonomous Operations(AO)is a transformation model which moves CSPs towards zero-touch operations and automates

172、 capabilities for learning,self-governance and self-adaptation of its value streams.TM Forums Autonomous Operations Maturity Model(AOMM)assesses the performance of a CSP applying advanced automation techniques to its value streams,by applying a maturity grid assessment framework that links to the Fo

173、rums Digital Maturity Model(DMM).A well-established framework is essential to guide digital capital resource allocation for CSPs strategic development plans and measure transformation success.20 Back to Contents|Defining new values and metrics for service-centric operations2.3.3 VALUE OPERATIONS FRA

174、MEWORK(VOF)THREE-LAYER MODEL AND VALUE TREE CONCEPTA well-built value model can help CSPs explore relationships between assumptions and likely outcomes.Defined in IG1292,VOF is based on a three-layer hierarchical model in which operational metrics and KPIs translate into business values defined with

175、in a four-level objective model:R.I.S.E.(revenue,innovation,satisfaction,efficiency).The analysis of operations value is not a simple one-to-one mapping,as operations logic for most CSP business processes is quite complex.Huawei recommends using the value tree model to visualize the operations logic

176、 based on business objectives and to translate this into KPIs,operations activities and operational performance measurements(OPMs).The Value Operations Framework(VOF),based on a value modeling approach and value tree analysis,provides a structural and graphical overview of the relevant decision-maki

177、ng element,enabling decision-making for each potential outcome preference.Value RealizationValue Operations Framework VOFAutonomous Operation Strategic InitiativesCapability Optimization BlueprintBusiness MetricsAdaptivity QuotientOperational MeasuresIntent Detection AccuracyImpactAutonomous Operati

178、on Maturity Model AOMMAOMM AssessmentPain point and gap analysisService capability roadmapMVP DesignAOMM 5-Level Maturity Value Stream Value Stream StageValue Stream StageValue Stream StageCustomer JourneyBusiness CapabilitiesBusiness CapabilitiesPeopleProcessResourcesInformationAO Maturity Benchmar

179、kAI-CLAANPAI-CLAProcess ImprovementApplications&APISystem&Data IntegrationTalent ReskillingBusiness GoalsCapital ReturnsRevenue&Margin GenerationEfficiencyFigure 10:Value realization utilizing VOF and AOMMAO Assessment:Gaps analysis will be performed to identify what capability to be enhancedAssessm

180、ent,Analysis,and RoadmapInside-Out Analysis:From stakeholder point of view to unfold the interrelationship between process,platform&supporting partiesOutside-In Analysis:From user/customer point of view to collect user journey,experience&expectationAOMM Operation DimensionAO MaturityCurrent capabili

181、tiesGap AnalysisBenefit and impact analysisCapability Enhancement High Level Plan Sub-DimensionsCriteriaNetwork anomaly resolutionMonitor network and detect anomalyPrevent network anomalyDemarcate and locate root-causeFormulate and decide solutionExecute solutionCustomer Network Experience Assurance

182、Detect customer network experience issueDetermine root-causeDetermine corrective actionImplement corrective actionCustomer network complaints resolutionDiagnose complaintsIdentify complaint root-causeDevelop corrective action planExecute action planFigure 11:AO assessment,analysis model and roadmapT

183、he Value Operations Framework(VOF),based on a value modeling approach and value tree analysis,provides a structural and graphical overview of the relevant decision-making element,enabling decision-making for each potential outcome preference.21 Back to Contents|Defining new values and metrics for se

184、rvice-centric operations2.3.4 FROM SILO SLAS TO VALUE MEASUREMENTTraditional network operation has limited insight because:It mainly focuses on network quality and operation efficiency(in terms of%improvement).Insights are limited to the immediate network capability value only,such as mean time to r

185、epair(MTTR),length of time spent onsite and so on.Most of the time the impact on potential business benefits such as financial KPIs is unclear.KPIs are interpreted in operational silos in each domain or process,such as RAN,core,transport and so on.As a result,each domain or process is only responsib

186、le for itself,with no ultimate responsibility for assurance of the end-to-end process,KPIs and business objectives.Figure 12:VOF three layers and seven steps for investment justificationFigure 13:Silo operation in traditional network functionsBusiness objectives can be expanded from efficiency only(

187、E)to revenue(R),innovation(I)and satisfaction(S)within service-centric operations,by providing a cross-domain holistic view of processes.These still include network quality and operation efficiency objectives,which are mandatory and fundamental.But it also adds a view across the whole CSP value chai

188、n to focus on the quality of the services(revenue),experience of subscribers(satisfaction),and the introduction and support of new technologies(innovation).22 Back to Contents|Defining new values and metrics for service-centric operationsAcross the CSP value chain,from networks to services and ultim

189、ately to customers,R.I.S.E.business objectives are relevant to subscriber churn,revenue loss and customer complaints and are designed to support the measurement of service-centric operations.The entire landscape of a service-centric operation can be presented as shown in the table.Services marked wi

190、th a tick indicate the scope covered in this whitepaper:Data of an individual(e.g.mobile)service see section 2.4.4 for illustration of value metrics,and section 2.5.2 for illustration of KPIs for monitoring and handling anomaly events Home broadband service see sections 2.4.5 and 2.5.3,respectively

191、Private line of public sector&enterprise service see section 2.4.6 and 2.5.4,respectively.Categorization of outcome KPIsDifferent enterprises may use slightly different terms for their KPIs and weight them differently,but similar to the definition of R.I.S.E business objectives,Huawei proposes the u

192、se of the three categories below to map outcome KPIs in the context of service-centric operation:1.Service availability.Measuring the uptime of the service without performance degradation and latency to demonstrate the quality of services2.SLA compliance.Measuring the performance of SLAs as defined

193、in a contract between the CSP and its subscriber,implying customer satisfaction3.Extent of user and service impact.Measuring intensity of negative or positive impact on the service and customer.Objective:Boost Revenue&Margin Objective:Enhance Customer Satisfaction&CXService Availability Rate(Data)Se

194、rvice Availability Rate(HBB)Service Availability Rate(Private Line)Service Recovery SLA Compliance Rate(Data)Customer Complaint Handling SLA Compliance Rate(HBB)Customer Complaint Handling SLA Compliance Rate(Private Line)Service Loss Rate(Data)Repetitive Customer Complaint Rate(HBB)Less repeating c

195、omplaintSReduced Complaints=%repetitive customer complaint improved x total number cases Secured SLA penaltyRReduced Revenue Loss=%SLA improved x total number of complaints reported x SLA penalty per breach Less Gain-back costRGain-back Cost Saving=#churn reduced*%selective customer*gain-back cost p

196、er subscriber#churn reduced=#churn last year*%service availability improved*churn rate for unsatisfied quality reason*churn rate for unsatisfied quality reason:38%of customer will tend to churn for unsatisfied network experience;CSP may use their own reference dataLess Churn of subscribersRChurn Los

197、s Reduction=#churn reduced*ARPUR.I.S.E objectivesOutcome KPIsAvailabilityComplianceExtentSecured Prepaid Revenue LossRReduced Revenue Loss=prepaid subscribers between 5:00 and 23:00 EDNS_traffic degree x average price per GB*EDNS_traffic degree:please refer to IG1294,the average loss degree of users

198、 who is provided data traffic less than their expected data traffic in kB/MB/GB.This is accumulated loss degree of users.Figure 15:New value metrics of service-centric operationFigure 14:Service landscape and scopeSubflowof 1.0 WPIG1291 Value StreamIndividual ServiceHome ServicePublic Sector&Enterpr

199、ise ServiceVoiceDataSMSHomeBroadbandIPTVPrivateLine5G Private NWIP VPNIoTFaultMonitor and Handle Anomaly EventPerformanceServiceComplaint2.4 VALUE PERCEPTION OF NEW VALUE METRICS FOR SERVICE-CENTRIC OPERATIONS2.4.1 THE LANDSCAPE OF SERVICES AND VALUE METRICS Before starting to define and design the

200、value metrics of service-centric operation,it is important to determine the overall landscape of services and the scope covered in this whitepaper.Like a traditional network operation,which is categorized by the combination of network technologies and value streams,the entire service-centric operati

201、on can also be categorized by CSP services and value streams.Field Name123 Back to Contents|Defining new values and metrics for service-centric operations2.4.2 VALUE METRICS DICTIONARYTo properly define and present the OPMs,KPIs and R.I.S.E.objectives,a value metrics dictionary is provided in the ta

202、ble below.It is designed with 12 mandatory fields,intended as step-by-step instructions.#DescriptionExampleIDUUID for the value metrics(unique value)MHAE_DATA_012NAMEThe name of the value metricService availability rate(data)3CATEGORYThe category of the value metricR.I.S.E.business objective:n Reven

203、uen Innovationn Satisfactionn EfficiencyService outcome KPIs:n Availabilityn SLA compliancen Extent of user&service impactAvailability4DESCThe description of the value metricMeasure the average percentage of the duration for which mobile users can seamlessly access the data service without breakdown

204、,latency,performance degradation etc.5FORMULAThe formula description of the value metrics1-(impacted users x unavailable time)/(average users x 60 x 24 x 30)x 100%6VALUE_STREAMRefer to IG1291,the value stream which the metric belongs toMonitor and handle anomaly event7SVC_OR_NWThe service or network

205、 domain which the metric belongs toTraditional network operation:n RANn Coren TXn IPService-centric operationn Voice servicen Data servicen SMS servicen Home broadband servicen IPTV servicen Private line servicen 5G private network servicen IP VPN serviceData of individual service8VOF_LEVELR.I.S.E.b

206、usiness objectives/KBI/KPI/OPMKPI9PERIODThe default calculation period of the value metrics,e.g.day/month/quarter/yearMonthly10UNITPercentage%,Number#%11BASELINEThe best practice of the value metrics99.98%12VERSIONThe version of value metrics1.0.024 Back to Contents|Defining new values and metrics f

207、or service-centric operationsCATEGORYNAMEDESC PERIODUNITChurn Loss ReductionRevenue(churn)With improved service quality experience,subscribers are more likely to stay with current CSP,supporting aim to reduce churn lossQuarterlyCurrencye.g.CNY,USDGain-back Cost SavingRevenue(churn)Because user churn

208、 is reduced due to improved service quality experience,the cost of regaining potential churn subscribers is now savedQuarterlyCurrencye.g.CNY,USDReduced Revenue Loss(prepaid subscribers)Revenue(loss)The prepaid subscribers are more likely to consume less data in times of service unavailability durin

209、g normal hours(5:00 23:00),which may result in less consumption of their mobile data package,translating to less revenueQuarterlyCurrencye.g.CNY,USDReduced Revenue Loss(SLA penalty in contract)Revenue(loss)With improved SLA performance,the CSP is less likely to have to pay penalties due to breaches

210、of SLAsQuarterlyCurrencye.g.CNY,USDReduced Complaints(due to the reduction of repetitive cases)SatisfactionWith reduction of repetitive cases,there is a reduction of overall complaints from subscribersMonthly#complaints2.4.4 VALUE METRICS OF DATA MOBILE SERVICECATEGORYNAMEDESC PERIODUNITService Avai

211、lability Rate(data)AvailabilityMeasures the average percentage of the duration for which mobile users can seamlessly access their data service without breakdown,latency,performance degradation etc.Monthly%SERVICEData ServiceService Recovery SLA Compliance Rate(data)SLA ComplianceMeasures the percent

212、age of data service case complaints in which the service recovery time(the time taken for service state to be restored from breakdown,latency or performance degradation to normal)complies with pre-defined SLAsMonthly%Data ServiceService Loss Rate(data)Extent of User&Service ImpactMeasures the percen

213、tage of total data in GB not consumed by users during breakdown,latency or performance degradation,indicating the traffic lossDaily%Data Service2.4.5 VALUE METRICS OF HOME BROADBAND SERVICECATEGORYNAMEDESC PERIODUNITService Availability Rate(HBB)AvailabilityMeasures the average percentage of duratio

214、n for which subscribers can use home broadband service without breakdown,latency,performance degradation etc.Monthly%SERVICEHome Broadband ServiceCustomer Complaint Handling SLA Compliance Rate(HBB)SLA ComplianceMeasures the percentage of home broadband service cases or complaints of which the servi

215、ce handling time(from reporting to recovery,including the on-site visit)complies with the contractual SLAs signed with subscribersMonthly%Home Broadband ServiceRepetitive Customer Complaint Rate(HBB)Extent of User&Service ImpactMeasures the percentage of complaints reported by the same home users wi

216、thin 30 daysMonthly%Home Broadband Service2.4.3 BUSINESS OBJECTIVES CATEGORIZED INTO REVENUE AND SATISFACTION25 Back to Contents|Defining new values and metrics for service-centric operations2.4.6 VALUE METRICS OF PRIVATE LINE ENTERPRISE SERVICECATEGORYNAMEDESC PERIODUNITService Availability Rate(Pr

217、ivate Line)AvailabilityMeasures the average percentage of the duration for which subscribers can seamlessly use private line service without breakdown,latency or performance degradation etc.Monthly%Private Line ServiceCustomer Complaint Handling SLA Compliance Rate(Private Line)SLA ComplianceMeasure

218、s the percentage of the private line service cases or complaints in which the service handling time(from reporting to recovery,including the on-site visit)complies with contractual SLAs signed with subscribersMonthly%Private Line ServiceRepetitive Customer Complaint Rate(HBB)Extent of User&Service I

219、mpactMeasures the percentage of complaints reported by the same home users within 30 daysMonthly%Home Broadband ServiceSVC OR NW2.5 DETAILED VALUE MEASUREMENT FORMULAE2.5.1 BUSINESS VALUES FOR SERVICE CENTRIC OPERATION(R.I.S.E.OBJECTIVES)CATEGORYNAMEFORMULAChurn Loss Reduction%Revenue#churn last yea

220、r x%service availability improvement x churn rate for unsatisfactory quality reason x ARPUAccording to a CSP survey conducted by Huawei,38%of customers will tend to churn due to unsatisfactory network experience;CSPs may use their own reference dataGain-back Cost SavingRevenue#churn last year x%down

221、time reduced x churn rate for unsatisfactory quality reason x%selective customer x gain-back cost per subscriberAccording to a CSP survey conducted by Huawei,38%of customers will tend to churn due to unsatisfactory network experience;CSPs may use their own reference dataFrom CSP data:%selective cust

222、omer:only certain percentage of users may be chosen to gain back based on CSPs best practice.Reduced Revenue Loss(prepaid)Revenue(traffic loss in GB of anomaly event(5:00-23:00)by site per hour x unavailable time x impacted prepaid subscribers)x average price/GBReduced Revenue Loss(SLA penalty)Reven

223、ueSLA improved%x total number of cases/complaints reported x SLA penalty per breachReduced Complaints%SatisfactionImproved Repetitive Customer Complaint Rate%x total number cases2.5.2 KPIS FOR MONITORING AND HANDLING ANOMALY EVENTS+DATA SERVICECATEGORYNAMEFORMULAService Availability Rate(data)Availa

224、bility1-(impacted users x unavailable time)/(average users x 60 x 24 x 30)x 100%Service Recovery SLA Compliance Rate(data)SLA Compliance Total number of cases recovered within SLA/total number of cases x 100%Service Loss Rate(data)Extent of User&Service ImpactEDNS traffic degree/(average site data c

225、onsumption per hour x 24 x total sites)x 100%EDNS traffic degree,please refer to IG1294,the average loss degree of users who provided data traffic less than their expected data traffic in KB/MB/GB.This is accumulated loss degree of users.26 Back to Contents|Defining new values and metrics for servic

226、e-centric operations2.5.3 KPIS FOR MONITORING AND HANDLING ANOMALY EVENTS+HOME BROADBAND SERVICE2.5.4 KPIS FOR MONITORING AND HANDLING ANOMALY EVENTS+PRIVATE LINE SERVICESo far,we have explored the concept of E.O.T.(Evaluate,Operate,Transfer)as a roadmap for successful operations transformation,and

227、outlined an end-to-end process for identifying new metrics to translate CSP strategies into measurable values using the VOF value tree.Section 3 of the whitepaper now explores approaches to transformation.It reviews the operational challenges and requirements for digital enablement platforms,and det

228、ails how CSPs can apply new values,technologies and applications to achieve service-centric outcomes.CATEGORYNAMEFORMULAService Availability Rate(HBB)Availability1-(home broadband unavailable time of each subscriber per month)/total subscribers x 60 x 24 x 30)x 100%Customer Complaint Handling SLA Co

229、mpliance Rate(HBB)SLA Compliance n VIP users:total number of cases of private line VIP users recovered within SLA/total number of VIP cases x 100%n All users:total number of private line cases recovered within SLA/total number of cases x 100%Repetitive Customer Complain Rate(HBB)Extent of User&Servi

230、ce Impact(the cases reported by home users again within 30 days)/total number of cases x 100%CATEGORYNAMEFORMULAService Availability Rate(Private Line)Availability1-(private line unavailable time of each subscriber per month)/total subscribers x 60 x 24 x 30)x 100%Customer Complaint Handling SLA Com

231、pliance Rate(Private Line)SLA Compliance n VIP users:total number of cases of private line VIP users recovered within SLA/total number of VIP cases x 100%n All users:total number of private line cases recovered within SLA/total number of cases x 100%27 Back to Contents|Transformation approaches to r

232、ealize new valuesTransformation approaches to realize new values3.OVERVIEWOperations transformation drives change across processes,platforms,people and technology.This section recaps some of the concepts covered in our previous O&M whitepaper and discusses how technologies have continued to evolve s

233、ince last year.And it focuses on the new components and capabilities proposed by Huawei to drive service-centric transformation outcomes using the value metrics identified in section 2.3.1 SUGGESTED TRANSFORMATION FRAMEWORKHuawei has proposed a new-generation digital operations framework which integ

234、rates processes,platforms,people(organizations),governance and digital transformation guidance methods,and supports the implementation of a digital operations methodology,leveraging TM Forums Open Digital Framework(ODF).Transformation GuidanceDigital Operation Implementation MethodSystem DomainEnabl

235、ement DomainInformation System ArchitectureTech.ArchitectureInformation FrameworkFunctional FrameworkMicro-service Open APIsAdaptionData ModelsIncident To Resolution MechanismRequest To Fulfillment MechanismPlan To Execution MechanismDeployment Mgmt.Governance DomainDigital OperationVOFDigital Opera

236、tionAOMMEnterprise Risk Mgmt.Information Security ManagementBusiness Continuity ManagementCommunication Management Digital Organization&PersonnelDevelopmentEffective Enterprise Mgmt.Service Continuous Improvement Planning&Program ManagementService Level ManagementDigital Asset ManagementData Governa

237、nceKnowledge ManagementDevelopment&OperationService DomainOperationBillingMeditation And RatingIntereconnection SettlementInvoicingAccount ReceivableDebt CollectionIntegrated SurveillanceIncidentMgmt.Problem Mgmt.Performance Mgmt.PreventiveMain.Capacity Mgmt.Inventory Data GovernanceSite AcceptanceS

238、pare Parts Mgmt.Field Main.Supplier Mgmt.Change Mgmt.ServiceProvisioningService DeskService CatalogueStrategy,Infra.&ProductsManage Network DesignManage Network ConstructionAvailabilityMgmt.Network Strategy PlanningManage Network PlanningNetworkService Dev.NetworkDemond Mgmt.Energy Consumption Mgmt.

239、MechanismScenario PlanningRequirement Mgmt.Digital Asset OperationDevelopment Mgmt.Data Modeling Mgmt.Notes:BSS service desk functions including billing are grayed out as they are out of scope of this report AOMM is TM Forums Autonomous Operations Maturity Model,VOF is its Value Operations Framework

240、Figure 16:Huaweis new generation digital operations framework3.2 RECAP OF WHITEPAPER 1.0:KEY CHARACTERISTICS OF SERVICE-CENTRIC OPERATIONSThe previous whitepaper described a pragmatic step-based approach to service-centric operations evolution compatible with the long-term(ten-year plus)visions of m

241、ultiple standards development organizations,including TM Forums AO/AN and ETSIs Zero-touch Service Management(ZSM).3.2.1 DIGITAL ENABLEMENT PLATFORMS AND APPROACHES KEY REQUIREMENTS In addition to the foundation provided by autonomous networks and operations,evolving from a network and service opera

242、tions center(NOC-SOC)collaboration approach to truly service-centric operations requires the replacement of siloed,overly manual processes and inflexible architectures with a digital enablement architecture.28 Back to Contents|Transformation approaches to realize new valuesHuawei proposes that a dig

243、ital enablement architecture should comprise the following four requirements:Converged data platform and data governance.This supports convergence of network and service assurance and enables multi-dimensional data modeling and analysis.It couples a fully converged operations data lake with graphics

244、,modeling and algorithm capabilities and related tools for data mining and analysis,enabling CSPs to derive maximum value from the data.One-stop AI solution for network operations.The digital enablement platform should provide centralized AI capabilities,tool chains,inference engines,out-of-the-box

245、AI-based services,and model training for operations processes.AI capabilities must be simple to use,requiring minimal expertise from operations personnel assigned to implement them.A DevOps platform incorporating low-code development.CSPs require improved service agility,necessitating a DevOps appro

246、ach to software integration and deployment.In addition,efficiently training operations personnel with telecoms expertise but little experience to develop digital assets requires a low-/no-code development environment.API gateway to ease API integration across autonomous networks and legacy element m

247、anagement/operational support systems(EMS/OSS).Huawei suggests that an API gateway a software application that accepts API calls from a client application and forwards them to the appropriate service is used to manage API activity and integrate an API-based approach with legacy EMS/OSS.The API gatew

248、ay platform should include the integration tools,applications and data sources that open use of its capabilities to trusted users(internal and external)within the CSP.Implementing these approaches can extend over a series of phases that best suit a CSPs competitive situation,digital transformation p

249、riorities and investment profile,to establish increasingly automated operations.Traditional and digital modes of operation can coexist and evolve over time while the CSP builds expertise and capabilities,as it steadily improves its customer experience and operational efficiency.3.2.2 TECHNOLOGY EVOL

250、UTION TO OVERCOME OPERATIONAL CHALLENGESThe previous whitepaper outlined several key challenges which CSPs must overcome if they are to transform to service-centric operations:Increasingly stringent,deterministic service performance requirements Rapid increases in network and service complexity Grow

251、ing,continuous need for domain model algorithms and knowledge The need for scenario-based development to support agile services And the need for open platforms and ecosystems to accommodate multivendor,heterogeneous IT and network environments.Huawei proposed six dimensions of technology evolution n

252、eeded to address these challenges and support service-centric digital operation.These dimensions leverage the capabilities delivered by the digital enablement platform outlined above.The five transformation challenges above map across these six dimensions of technology evolution in a one-to-many fas

253、hion.For example,deterministic network assurance addresses the challenge of deterministic service performance requirements and rapid increases in service complexity.Implementing these approaches can extend over a series of phases that best suit a CSPs competitive situation,digital transformation pri

254、orities and investment profile,to establish increasingly automated operations.29 Back to Contents|Transformation approaches to realize new values1.Deterministic Network Assurance(DNA)Service-centric DNA requires a host of automated operations capabilities if CSPs are to meet stringent end-to-end SLA

255、s required by digital services.2.Hyper Automation(HA)Hyper automation uses technologies and techniques such as robotic process automation,causal reasoning and predictive analysis to manage the rapid increase in network and service complexity.3.Model and Algorithm of Telecom(MAT)Standardization and r

256、epeatability of operational actions based on service and network conditions are critical.This requires leveraging critical telecoms expertise across cross-domain(RAN,core etc.)and cross-layer(resource,service,user etc.)network and service scenarios.4.Telecom Knowledge Platform(TKP)As telecoms-specif

257、ic models and algorithms proliferate and generate high data volumes,more sophisticated tools are needed to manage these digital assets.Mechanisms must be established to identify,compress,store,access and update the data,which must be organized into knowledge graphs.5.One Trustworthy DevOps(OTD)DevOp

258、s,and the need to make development easier and more efficient for operations personnel,will allow more customized and agile development of digital assets.Low-code/no-code development is essential if CSPs are to realize their agility,cost efficiency and value creation business goals.6.Ecosystem Enable

259、ment(EE)Ecosystem Enablement opens the digital operations platform to accommodate multivendor,heterogeneous IT and network environments and support expansion of CSP businesses into a broader set of services and use cases,some of which might be offered by partners.It supports integration and intercon

260、nection of a CSPs existing OSS/BSS and network resources into the digital operations automation framework and architecture.3.2.3 FROM AS-IS PRESENT METHOD OF OPERATIONS TO FUTURE METHOD OF OPERATIONSOur whitepaper last year described the transformation path for NOC-SOC collaboration from as-is netwo

261、rk-centric present operations,evolving towards to-be solutions for service-centric operations as a future operations phase.The table on the next page highlights the key features in different phases and solutions.Ecosystem Enablement opens the digital operations platform to accommodate multivendor,he

262、terogeneous IT and network environments and support expansion of CSP businesses into a broader set of services and use cases,some of which might be offered by partners.30 Back to Contents|Transformation approaches to realize new values3.3 KEY EVOLUTIONS SINCE WHITEPAPER 1.0:NEW VALUE,NEW TECHNOLOGY,

263、NEW APPLICATIONS3.3.1 CSPS ARE MISSING KEY OPERATIONS CAPABILITIESCSPs existing operations are very network-centric,focused on network availability and performance(network KPIs).When outages and/or degradation are detected,closed-loop actions typically focus on network-level recovery and/or optimiza

264、tion.Whilst high network availability and performance are foundations for good service quality,this does not always equate to good service experience from a user perspective.Since the release of our whitepaper last year,Huaweis experience in supporting CSPs undergoing service-centric transformation

265、has highlighted three important areas of capability missing from most organizations:1.Impact analysis.CSPs need near real-time,accurate service and user impact analysis whenever network anomalies(outages and/or degradation)are detected.2.Agile workaround solutions.When anomalies are detected,CSPs ne

266、ed to be able to run agile,automated workaround solutions to restore/maintain service and minimize user impact,in parallel with network-level recovery and/or optimization.In summary,the full closed-loop tasks should:Minimize the negative impact of anomalies by restoring normal service operation impl

267、ementing a service restoration workaround as quickly as possible In parallel execute remediation actions by resolving network operation root cause problems and/or errors Typically,fallback the workaround for service restoration after network recovery and resolution.3.Monitoring.Whilst CSPs are addin

268、g more service and user-level monitoring to detect anomalies beyond network level,they also need to have root cause analysis and cross-layer correlation capabilities.PHASEAs-is solutionNOC-SOC collaboration:1.Align NOC&SOC operations2.Align Complaint subflow,Service subflow,Performance subflow&Fault

269、 subflow together into full value stream of“Monitor and Handle Anomaly Event”of Service&Network Assurance domain 3.Align the subflows by Ticket4.Select limited operation tasks to build automation1.Easy to start&low investment2.Help operations realize why&what for service-centric transformation1.The

270、key problems of silo data&silo anomaly detections often in service&network incident correlation are not solved yet2.Missing the new future-ready digital enablement platform for continuous transformation journeySOLUTIONKEY FEATURESPROSCONSPRESENT METHOD OF OPERATIONSTransition solutionIntroduce the n

271、ew “To-be digital enablement platform”first:1.Low-/no-code design time2.Multi-dimensional data modeling and analysis.3.Event-based management for service&network anomaly correlation4.Automated closed-loop capabilities1.Future-ready platform is in place2.Operation team could start to exercise the new

272、 way of working&new skills requirements1.Much more capex investment to do complicated architectural planning and/or proper partner selection for future-ready platformTo-be solutionAfter future-ready platform is in place,move on to a complex,systematic transformation programspan technologies,processe

273、s,platforms and people:1.New DevOps processes&data governance etc.2.Comprehensive people upskilling program3.Evolve from rule-based automation to AI-based&LLM-based automation1.Transform to highly integrated service-centric operations2.Align with industrial vision such as autonomous operation1.Very

274、complicated transformation program and failure rate is highFUTURE METHOD OF OPERATIONS31 Back to Contents|Transformation approaches to realize new valuesNot all the capabilities proposed above exist in operations and OSS systems currently deployed,so it will require significant capex investment to a

275、cquire all the functions necessary for service-centric transformation.Huawei proposes that CSPs review the following three elements when considering this investment:New value.While there is a high degree of consensus on the concept and necessity of service-centric transformation,it is important to i

276、dentify the new business and operation value that service-centric operations can bring in the future before proceeding with investment.New technology.The challenges related to isolated network and service assurance are mainly caused by siloed data and siloed anomaly detections in as-is operations,so

277、 new technology features are needed to solve these problems.New applications.New technologies must be adaptable to new scenario-based applications and achieve practical results in order to deliver return on investment.Figure 17:Key automation capabilities missing from CSPsFigure 18:Factors for consi

278、deration in service-centric transformation32 Back to Contents|Transformation approaches to realize new values3.3.2 NEW VALUETraditional network operations are siloed without a holistic view.Performance measurements of operations are concentrated on a set of network-centric metrics such as network av

279、ailability,MTTR,onsite duration and so on so business outcomes are limited mainly to operational efficiency percentage improvement.Service-centric operations provide a holistic view(cross-domain and cross-layer)of all processes relating to user experience,network quality and operation efficiency obj

280、ectives,so business outcomes bring in new values such as assuring more revenue,enabling service innovation and improving user experience.Identifying these new values,and designing new service-centric metrics,are mandatory in delivering return on investment,as discussed in section 2.3.3.3 NEW TECHNOL

281、OGYOur previous whitepaper illustrated six dimensions of technological requirements for service-centric transformation.In order to drive operational change to the next level,Huawei proposes that the following new technology evolutions should also have more significant roles:Digital Twin Network(DTN)

282、.Integrating and correlating cross-domain and cross-layer data was established as a mandatory need in our previous whitepaper,as a converged data platform.Here we further overlay the capabilities of the topology restoration,data correlation and analysis,network status replay and simulation on top of

283、 the converged data platform.This can now be further evolved into a DTN,for troubleshooting,root cause analysis,simulation and so on.Generative AI(GenAI).Almost all industries are researching how to apply LLM-based GenAI technology to transform existing business models.GenAI will deeply change CSP o

284、perations models,and AI agents based on GenAI will be the key enabling technology for AN levels 4 and 5.Expected Demand Not Served(EDNS).To add service and user impact analysis,as well as workaround solutions to restore/keep alive/minimize service and user impact,CSPs need to build algorithms to sim

285、ulate and measure the impact of network incidents.EDNS has been introduced by the MAMA project to enable members to define new standards to measure and manage digital transformation initiatives that target autonomous operations with the focus on business outcomes.We explore these topics further in s

286、ection 3.4.3.3.4 NEW APPLICATIONBased on deployment of the new technologies mentioned above,new applications are being leveraged by CSPs to enable new capabilities and solutions for implementing service-centric operations.Section 3.5 explores this topic in detail.3.4 UTILIZING HUAWEI ASSETS:NEW PLAT

287、FORM,NEW TECHNOLOGY(GENAI,DTN,EDNS)3.4.1 THE TO-BE DIGITAL ENABLEMENT PLATFORM AND NEW PROGRESSTo recap,in the previous whitepaper Huawei proposed the key characteristics of the new to-be digital enablement platform designed specifically for Service&Network Assurance(S&NA)domain as follows:1.Low-/no

288、-code design time:easy development of auto APP/APIs.2.Converged data platform(run-time)multi-dimensional data modeling&analysis3.Anomaly-to-incident-to-resolution(run-time):a)Event-based management multi-dimensional event correlation of silo anomaly detections b)Closed-loop automation framework easy

289、-to-orchestrate auto APIs for end-to-end automations.Service-centric operations provide a holistic view(cross-domain and cross-layer)of all processes relating to user experience,network quality and operation efficiency objectives,so business outcomes bring in new values such as assuring more revenue

290、,enabling service innovation and improving user experience.33 Back to Contents|Transformation approaches to realize new valuesBased on latest developments,Huawei proposes further detail to the practical design of the to-be digital enablement platform,as shown in the graphic:DecisionExecutionAnalysis

291、AwarenessDataIntegration with other system(REST,API,Data,etc.)Autonomous NetworkIncidentMAENCEResource DataAlarmDial-testTopology dataService DataKPIComplaintService deteriorationVIP user experienceService&Network Incident&Automated Closed-loop FrameworkKnowledge RecyclingDataTTRuleService&Network I

292、ncident Correlation Network Topology50+scenariosTOP3 vendorsService Impact AnalysisCross domains/layersAI Algorithm groupIntelligent CorrelationCross domains/layersAI Algorithm groupFeature Matching&Event GenerationWorkaround Solution DecisionAI-basedService RestorationSimulationAssisted ExecutionRo

293、ot Cause Analysis Rule-based AI-basedRecovery Solution DecisionRule-based AI-basedClosed-loop ExecutionOutage recoveryDegrade optimizationRisk preventionNetwork Recovery AutomationService Restoration WorkaroundManual handling&executionClosingData CollectionAnomaly AwarenessLatest practical design of

294、 the“To-be digital enablement platform”for Service&Network Assurance(S&NA)domainFigure 19:Huaweis latest design of the to-be digital enablement platformThe new digital enablement platform is designed to be future-ready for an autonomous service and network assurance domain.It collects multi-dimensio

295、nal data from both network and service domains,collaborates with autonomous network single domain events,and identifies abnormal events with user and service impact by applying anomaly detections,intelligent correlation,intelligent RCA,service impact analysis,matching features,and providing parent/c

296、hild correlation of events.A correlated event supports two closed-loop streams in parallel:(1)network-level recovery;and(2)service-level restoration,which can be executed synchronously to achieve agile service restoration,reduce service losses,and enable service-centric operation transformation and

297、evolution.It also implements knowledge recycling capability for operations data,rules,work orders and other data during the processing of event awareness,analysis,decision and execution.AI methods are used to automatically extract event processing-related models,with incremental methods used to enha

298、nce event processing intelligence.The new platform also defines an integrated framework for processing event awareness,analysis,decision and execution.It incorporates low-/no-code design time and undergoes lots of pre-integration.The objective is to enable the platform to quickly integrate automatio

299、n capabilities from different suppliers in addition to utilizing self-developed automation capabilities and achieve autonomous operations for various service and network assurance scenarios.3.4.2 NETWORK-CENTRIC OPERATIONS TO SERVICE-CENTRIC OPERATIONS REQUIRE INTRODUCTION OF 3 NEW CORE TECHNICAL FU

300、NCTIONSEssentially,the six-dimensional model proposed in the previous whitepaper described only the capabilities required for future O&M systems but did not discuss how these capabilities can be organically integrated into a complete system architecture,or identify the key technical functions needed

301、 in that system.Huawei proposes that the six dimensional capabilities of the system will ultimately be organically integrated into the following three new core technical functions:34 Back to Contents|Transformation approaches to realize new values1.One view to details.Integrate status information(pa

302、st,present,future)of all network layers,service layers and user layers through the Digital Twin Network,to achieve a fully graphical interface for viewing/drilling down into details.2.One question to root cause.GenAI encapsulates and integrates all algorithms and model capabilities in the telecoms f

303、ield(including network element and network management capabilities),and provides assistant,copilot and agent style human-machine interaction.3.One click to closed loop.The system is embedded with service-oriented algorithms(e.g.Expected Demand Not Served EDNS algorithm),connecting multidimensional d

304、ata correlation,service impact analysis,service restoration and network recovery flow,integrating with GenAI and DTN capabilities,and integrating event awareness,analysis,decision and execution to minimize the user and service impact of incidents.3.4.3 THE CONNECTION BETWEEN THE SIX DIMENSIONS MODEL

305、 AND 3 CORE TECHNOLOGICAL FEATURESThe following figure shows the mapping relationship between the six dimension capabilities and three core technology features:EDNSa)Enhance wireless traffic loss reductionb)Explore transmission loss reduction etc.DTN(Logical Brian)a)Basic layerb)State layerc)Cross-l

306、ayer correlation d)Layer rendering etc.GenAI(Creative Brian)a)Domain vector knowledgeb)Domain prompt engineering c)Domain API assetsd)Service enablement toolchain etc.Deterministic Network Assurance(DNA)Model&Algorithm of Telecom(MAT)Data&AI+PeopleOne Trustworthy DevOps(OTD)Ecosystem Enablement(EE)H

307、yper-Automation(HA)Telecom Knowledge Platform(TKP)Further integrate&group the 6 dimensional capabilities into 3 core technological features 1.Digital Twin Network.DTN represents the logical brain of operations.It integrates cross-layer and cross-domain data,and comprises past and present states,as w

308、ell as inferring potential future states.The Deterministic Network Assurance(DNA)in the six dimensional model has a clear demand for the presentation,simulation and prediction of states in each layer of the DTN system.Radio Resource Management(RRM)simulation technology,network simulation technology,

309、future user and service simulation technology in the Models&Algorithms of Telecom(MAT),and the state and trend prediction algorithms of network services,will also map to DTN features,so can be organically integrated into a large system.2.Generative AI.GenAI represents the creative brain of operation

310、s.It will become the entry point for all system kernel capabilities in the future.Therefore,it needs to understand Telecom Knowledge Platform(TKP)as well as MAT,and achieve true logical reasoning based on self-programming,connecting various scenarios into end-to-end operation flows with Hyper-Automa

311、tion(HA)capabilities.3.Expected Demand Not Served.EDNS will be a collection of specialized algorithms for subdividing service dimensions in the telecoms field,including but not limited to single domain and cross-domain service impact measurement algorithms,optimal path optimization algorithms,and un

312、ique execution algorithms for service restoration and/or network recovery in various scenarios.DNA and MAT in the six dimensions model will map to this technology feature.Figure 20:Mapping relationship between the six dimensions model and three core technology features35 Back to Contents|Transformat

313、ion approaches to realize new valuesThe following sections further elaborate on these three core technological features.3.4.4 DETAILED DISCUSSION ON 3 CORE TECHNOLOGICAL FEATURES:DTN,GENAI,EDNSA.Digital Twin Network(DTN)A digital twin network is a network system that creates virtual twins of physica

314、l network entities in a digital manner and can interact and map with physical network entities in real time.DTN is designed for invisible and intangible telecoms infrastructure and network services,providing precise modeling,mapping and perception mapping physical and virtual network twins in a visu

315、al form.It also supports higher-order characteristics such as root cause derivation,historical replay,future prediction,and closed-loop control at a small cost,ultimately achieving the goal of symbiotic evolution with continuous testing(CT)networks.Introducing DTN technology can assist in solving pr

316、oblems such as low efficiency of O&M work order flow,low reliability of dedicated lines,difficulty in defining cross-domain incidents,and invisible root cause derivation in telecoms networks.A.1 Technical descriptionBased on the unique twin model of the CT network,coupled with key root technologies

317、such as distributed parallel data acquisition,graph calculation and data replay,the DTN system in the CT field will have the following key capabilities:1.Cross-domain logical topology construction from physical topology of network service path to cross-domain and cross-layer topology of service;from

318、 single-domain topology orchestration and splicing into a cross-domain full-view graph2.Real time calculation and replay of network status overlaying alarms and golden indicators onto network elements,connection links and service paths;compute network status based on network connections and service

319、paths3.3D twin visualization 3D presentation based on physical location4.Large capacity data management capability 1 billion point edge diagram of optical cable and equipment;1 million services path diagram of dedicated lines.A.2 CT network specific twin modelIn addition to defining layered models s

320、uch as network element resources,network topology,service paths,services and users,the CT network twin model also needs to overlay state time sequence data and generate dynamic views.Layer NameDefinitionUser layer1.Dedicated line SLA(Gold,Silver,Bronze etc.),QoS,Satisfaction etc.2.ToH User package l

321、evel,complaint info etc.Service layerService:CSPs offered“services”to user in the model of ToC,ToH,ToBToC:Voice,SMS,Data etc.ToH:IPTV,Home Broadband etc.ToB:2B dedicated line,5G 2B etc.Service path layerService Path:The e2e path of network nodes&link information passed between the given service acce

322、ss starting&ending points,the planned link sequence in the network protocol,and the master-slave relationshipNetwork topology layerPhysical Connectivity:Physical connection(fiber optic,network cable)from network element A to network element BLogical Link:The logical connection between two network el

323、ements(such as L2,L3 links,transmission links,tunnels etc.)Network element resource layerResource&status information of a single network element.Network element:Functional units&base stations in the networkLogical resource:Cell,carrier frequency,IP,single board etc.Connection inter network element:A

324、AU-BBU connection,OTN inter board connection,panel backboard connection relationshipPhysical spaceEnvironment info:Building,data center,physical site,street etc.ViewAlarm spatiotemporal aggregation viewAlarm root cause viewLatency viewTraffic ViewOverlay alarmOverlay latency,traffic,KPI etc.Figure 2

325、1:CT network twin model layers36 Back to Contents|Transformation approaches to realize new valuesThe challenges faced by modeling include difficulty in defining stable hierarchical models,complex association rules between multiple layers,reasonable allocation of IDs for multi-dimensional data,and th

326、e distribution,storage and unified use of multi-dimensional data.For example,the definition and implementation of UIDs for various resource objects,alarms and timing data in communication systems are different,making it difficult to directly associate data.It is necessary to perform ID mapping based

327、 on name,ID,location information and so on,combined with business rules,machine learning,graph algorithms,and other algorithms,in order to map various UIDs to a unified ID.Through this unified ID,data from various data silos can be associated,achieving data integration to ensure accurate and compreh

328、ensive user and service analysis.A.3 Distributed parallel graph computingDistributed parallel graph computing is an effective method for processing large-scale graph datasets.It uses a distributed computing framework to execute graph algorithms in parallel on a cluster of multiple computers or serve

329、rs.The graph data structure consists of vertices and edges.As the data size continues to grow,processing graph data on a single machine becomes increasingly difficult.Distributed parallel processing speeds up and can handle ultra large-scale graphs.Distributed graph computing is used for alarm and o

330、ther status data,as well as for resource association and aggregation calculations,and supports second level subgraph queries to support real-time demarcation calculations for events and incidents.Distributed parallel graph computing,which involves processing and analyzing large-scale graph data in d

331、istributed systems,is a complex and challenging task.The technical challenges it faces mainly include dynamic graph processing.In scenarios where the graph structure changes over time,how to efficiently handle dynamic updates of the graph,such as adding and deleting nodes and edges,while maintaining

332、 real-time and accurate computation,is a complex problem.The storage and indexing of large-scale graphs,efficiently storing and indexing large-scale graph data to support fast queries and traversals,is a fundamental but highly challenging problem.A.4 Data replayData replay is a technique that allows

333、 users to reprocess and analyze previously collected or recorded data,simulating the environment and conditions in which raw data is generated or received.It supports storage of multidimensional data such as historical resources,states and time series,and records the process of resource changes,and slice storage of multidimensional data,including topology,state,and time series data,based on time w