×

Autonomous intelligent networks. Analysis of economic outcomes and readiness of technologies

November 2021

Analytical Report (full version)

Request cost of the full version of the study: news@json.tv

Analytical Report (full version)

Autonomous intelligent networks. Analysis of economic outcomes and readiness of technologies
Autonomous intelligent networks. Analysis of economic outcomes and readiness of technologies
November 2021

Autonomous intelligent networks. Analysis of economic outcomes and readiness of technologies

November 2021

To register or Log in, to download the file

Download file

+7 926 561 09 80; news@json.tv

Write, call, if you have questions

Implementation of highly autonomous intelligent end-to-end NMS/OSS/BSS-processes by ICT-infrastructure operators and service-providers enables extremely high economic benefits.

 

Transition from typical for major telcos L0/L1 levels of autonomy and intelligence to L2/L3 ones leads to two-fold decrease of direct operational expenses on ICT-infrastructure management and services provisioning. In scale of global telecom services market in means that margin of telecom service providers may increase at $144B annually.

 

Along with decrease of direct operation costs, transition from L0/L1 to L2/L3 and further to L4/L5 levels of autonomy and intelligence enables telcos to provide next generation of telecom services called Network as a Service (NaaS), i.e. networking services been provisioned “on demand” with deterministic SLA, together with ability to keep average utilization of its ICT-resources at extremely high level – up to 90%.  Potentially it will allow to form new segment of telecom services market generating $900B of annual revenue and $585B of annual margin.

 

Currently, many operators in different ICT-domains already have the technical capabilities to provide fully software-based next-generation services within a single domain, but few of them have the ability to orchestrate E2E services, given, in particular, the fact that physical networking functions still coexist with VNFs, which, in turn, run primarily on virtual machines rather than containers. In fact, the industry is only at the beginning of defining cross-domain orchestration rules.

 

Analysis of product portfolios of telecommunications equipment and software vendors shows that even at the level of intra-domain autonomous management, their offerings do not yet cover all needs of operators arising from implementation of autonomous network concept, and at the level of cross-domain management, vendors’ solutions clearly lack functional completeness and openness.

 

Global cloud providers such as AWS, Google Cloud, Microsoft Azure, IBM Cloud, Oracle Cloud are new players in the development and implementation of the concept of autonomous networks. They not only actively stimulate telecom operators to implement innovative services and deploy next-generation network infrastructure, but also develop cloud platforms intended for operators of virtualized infrastructures to automate and intellectualize the almost complete stack of NMS/OSS/BSS processes necessary for deploying autonomous networks and providing innovative services based on them. The advantage of global cloud providers over vendors of telecom equipment and software is the possibility of pilot deployments on their own global infrastructure, which allows to quickly receive feedback and refine solutions to the level necessary for their transfer to the status of commercial operation.

 

Key definitions of autonomous intelligent network concept

 

Autonomy is the execution of a process by software applications without direct human intervention. Autonomy of a network means that all processes related to the management of its resources and services been provisioned to customers with the use of these resources are executed by software applications. According to Y.3173 framework issued by ITU-T and similar documents released by other SDOs, network autonomy and intellectualization is defined as fully automated execution of network and services management processes based on self-modified (adaptive) algorithms covering all layers of technological and business processes of an operator – form Network Management System (NMS) to Operation Support System / Business Support System (OSS/BSS). Fragmented autonomy and intellectualization of some process groups with different levels of autonomy are defined as well, along with end-to-end total autonomy. 

 

Defined by SDOs levels of autonomy are depicted on Fig. 1. The depicted classification is developed by ITU-T, other SDOs have similar classifications. ITU-T and other SDOs define three levels of autonomy for all three groups of processes (NMS, OSS, BSS): execution by humans with help from IT-systems (Human on Fig. 1), automated execution using algorithms defined by humans, which cannot be changed by software applications (Human and System), and automated execution using algorithms initially defined by human, but modified by software applications based on historical data analysis and forecasting (System). In other words, the “Human and System” level of autonomy means full automation, while the “System” level means full intellectualization of automated processes execution, i.e. complete autonomy from human intervention including self-adaptation of algorithms.

 

 

Along with autonomy levels definition, different “weights” are introduced depending on the phase of closed-loop autonomous management: action implementation, data collection, analysis, decision (operational decisions), and demand mapping (mid-term and strategic decisions). Combined, levels of processes autonomy and “weights” of autonomy depending on phase within a process form a kind of ladder (Fig. 1). The steps of the ladder start from Level 1 “Assisted network operations” where “Action implementation” and “Data collection” phases are automated (executed autonomously) with the use of algorithms defined by humans («Human and System»). The highest step of the ladder is Level 5 “Full intelligence”, which assumes autonomous execution of all management phases, including operational, mid-term, and strategic decisions making - network upgrade, for instance, with use of self-modified algorithms.

 

Using a comparison of autonomy and intellectualization levels offered by ITU-T, TM Forum, and other SDOs and combining it with analysis of network autonomy concept implementation cases, one can state that starting from L2/L3 the autonomy and intellectualization cover all NMS and OSS processes, and partially cover BSS-stack – at least charging, billing and self-service portal. L4/L5 levels of network autonomy and intellectualization cover the whole stack of NMS/OSS/BSS-processes, including those that during all the history of telecoms always been executed by humans without use of any automation except Exel – strategic market trends analysis, development of new services design and development of strategic plans for network upgrade and extension. It is important to mention that all levels of autonomy starting from L1 “Human and System” assumes end-to-end automation of proper group of processes, so the difference between levels is not in the degree of automation – all considered levels assume full automation, but in the degree of autonomy and intelligence of fully automated processes execution.

 

Context for the autonomous intelligent network concept

 

The autonomous network concept is considered by SDOs exceptionally in the context of next-generation networks, based on software-defined control and data planes separation (SDN), network functions virtualization (NFV), intra- and inter (cross) domains orchestration (MANO), and artificial intelligence (AI).

 

Main goals for autonomous intelligent network concept implementation based on SDN, NFV/MANO, and AI technologies are formulated by SDOs as following:

 

- to enable fully automated on-demand provisioning of services with deeply customized SLA, which consists of a broad range of QoS metrics defined and managed by software applications, while a range of managed/customized metrics may exceed networking ones (throughput, latency, jitter, packet loss) – these are high-level metrics such as cybersecurity level, availability and reliability level, etc.;


- to noticeably cut operational costs related to personnel directly engaged in the execution of network management and services provisioning processes, with a focus on complicated customized services with managed/customized SLA;


- to improve energy consumption efficiency and utilization of network resources.

 

 

These three main goals mean that the autonomous network concept is focused on different aspects of provisioning next-generation network services, i.e. services that possess essential characteristics of cloud services - Network as a Service, NaaS. In contrast to NaaS, “traditional” network services do not possess some or all essential characteristics of cloud service: self-service via customer portal, rapid elasticity, and/or managed SLA. These are network-dependent voice services, “best-effort” mobile broadband access services, leased channels and VPN services with no elasticity and self-service, fixed broadband access services with no SLA and statically limited throughput. 

 

We define “traditional” telcos as telcos with a prevailing share of “traditional” network services in their revenue structure, and next-generation telcos as telcos with a prevailing share of NaaS and other cloud services. Traditional telcos are still characterized by a low level of NMS/OSS/BSS-processes automation, while next-generation telcos have achieved a higher level of automation to meet all essential characteristics of the cloud model for their commercially available dynamic software-defined services. Comparison of some financial KPIs of traditional and next-generation telcos allow to quantitatively assess improvements of economic indicators due to shifting from lower level of network autonomy (i.e. level of NMS/OSS/BSS-processes automation) to higher one.

 

Results of economic outcome assessment

 

Results of our research shows that the volume of economic outcome due to the implementation of autonomous networks concept is directly dependent on the level of autonomy and intelligence achieved by an operator – the higher is the level of autonomy and intelligence within groups of processes and the broader is the coverage throughout NMS/OSS/BSS stack of processes the higher is the volume of economic outcome.

 

Implementation of highly autonomous intelligent end-to-end NMS/OSS/BSS-processes by ICT-infrastructure operators and service-providers enables extremely high economic benefits.

 

Transition from typical for major telcos L0/L1 levels of autonomy and intelligence to L2/L3 ones leads to two-fold decrease of direct operational expenses on ICT-infrastructure management and services provisioning. In scale of global telecom services market in means that margin of telecom service providers may increase at $144B annually.

 

Along with decrease of direct operation cost, transition from L0/L1 to L2/L3 and further to L4/L5 levels of autonomy and intelligence enables telcos to provide next generation of telecom services called Network as a Service (NaaS), i.e. networking services been provisioned “on demand” with deterministic SLA, together with ability to keep average utilization of its ICT-resources at extremely high level – up to 90%.  Potentially it will allow to form new segment of telecom services market generating $900B of annual revenue and $585B of annual margin.

 

The pioneering launch of NaaS took place five years ago, but real breakthrough of NaaS commercialization has started since 2020 when leading cloud providers elaborated clear strategies of cooperation with telcos in provisioning of converged IT/telco services of distributed multi/hybrid clouds (including edge ones) where NaaS acts as important part of the offering. Rapid growth of distributed clouds services market may fully turn the potential of NaaS into teclos’ revenue and margin as early as 2030.

 

Commercialization of new generation telecom services (NaaS) and autonomous SDN&NFV networks goes hand in hand with their standardization, which, in particular, resulted in standardization of new business-models and roles for telcos and other ICT-providers: providers of applications, including industrial ones, providers of end-to-end virtual ICT slices and providers of virtual resources of networking and computing domains. Implementation of autonomy and intelligence to NMS/OSS/BSS-processes differs greatly depending on these business-roles.

 

The most matured and successful in implementation of autonomy and intelligence to their NMS/OSS/BSS-processes are the leading providers of cloud computing and converged computing-networking services, which operate own hyperscale datacenters and backbone networks – in terms of new business-roles they are acting as operators of some infrastructure domains (datacenters, backbone networks) and providers of applications. Some niche players, which may be classified as next generation telcos, – Akamai, Zscaler and others, focused on provisioning of highly valued cloud telecom services such as SD-WANaaS, SECaaS, CDNaaS shall be considered among most matured ICT-providers as well. They are far ahead of traditional telcos and may be considered as candidates #1 for end-to-end network slices providers.

 

Traditional telcos' future business role is to become operators of peripheral infrastructure domains: edge computing, access and transport networks. Traditional telcos are the less matured than ICT-providers and still do not leverage full potential of autonomous networks. The obvious lag in NMS/OSS/BSS autonomy level between leaders of cloud services market and traditional telcos is caused by necessity to apply “brownfield” scenario – the need to totally revise their existing product portfolio and business-model, including organization structure and corporate culture, which is extremely difficult for corporate structures with 50-100 thousand of employees. So the traditional telcos need a clear transformation strategy. 

 

Nevertheless, as a part of integrated offering from leading cloud providers, some traditional telcos have recently succeeded in launching next generation telecom connectivity services for distributed hybrid clouds: bandwidth on demand (BoD) on backbone networks, networks slices and MEC on 5G networks. It marks a turning point for traditional telcos in their transformation towards autonomous networks and domain providers of NaaS. It is worth to mention that possibility for economically effective provisioning of NaaS is emerging already at L2 level of network autonomy.

 

Geographically the breakthrough in provisioning of NaaS based on autonomous NMS/OSS/BSS-processes took place in North America, where up to 60% of global cloud services market is concentrated. In regions characterized by less developed cloud markets traditional telcos are far less active in development and implementation of autonomous NMS/OSS/BSS-processes. This negative attitude to autonomous networks must be changed. Otherwise the constantly decreasing consumption of network-dependent voice services and decreasing margin on data services (broadband access, private lines and IP MPLS VPN) will completely ruin telcos’ potential to return investments into networks expansion and upgrade, which, in its turn, will lead to degradation of services quality and availability.

 

Taking into account that de-facto full-scale transformation towards autonomous networks has already  launched by the most advanced traditional telcos in North America and followed by telcos in Western Europe, China, South Korea, Japan and some other countries, conservative approach based on slow and fragmented improvement of NMS-stack with mostly no influence on OSS/BSS-processes, which is still widely used by telcos, shall be considered as a “proof of concept” – a first step to show technological and economical feasibility of autonomous networks and services. However, from strategical point of view there is no alternative to total replacement of existing NMS/OSS/BSS stack for fully autonomous intellectual ones. Otherwise economical impact provided by fragmented improvements of NMS-processes will be less than 0,5% of telcos’ revenue, while effects of radical decrease of direct operation expenses and possibility to provide new services keeping high level of ICT-infrastructure utilization cannot be achieved. 

 

In this regard it should be noticed that new business-models are positively dependent from each other: if an operator increases its efficiency it positively impacts efficiency of other operators and providers in case they are selling or consuming resources to/from each other. It means that leading providers of cloud services are interested in radical improvement of traditional telcos efficiency and in their ability to provide MEC and NaaS in access and transport networks domains. In practice this positive dependence is observed as development and implementation by cloud providers of strategic cooperation plans for telcos aimed on intensification of autonomous networks deployments by traditional telcos and their ability to provide NaaS which is important part of distributed cloud services, while the most advanced telcos are synchronizing their strategies with those of cloud providers.

 

Readiness of technologies and standards

 

Autonomous network initiatives are being developed by several standards development organizations (SDOs) such as ITU-T (principles and classification of levels of autonomy), TM Forum Network Automation Initiative (NMS / OSS / BSS processes of autonomous networks), ETSI ZSM (focuses on automation network operations), ETSI ENI (focusing on AI for network automation), 3GPP (wireless network automation), and GSMA (exploring network automation use cases).

 

Note, with some differences in specialization, all SDOs agree that the concept of autonomous networks is necessary for the implementation of innovative services focused on industrial applications and is inherently associated with a new network architecture based on the principles of virtualization and software-defined control.

 

Along with SDO, the largest vendors of telecommunications and computing equipment, as well as vendors of NMS/OSS/BSS-processes automation tools, are actively involved in forming the conceptual vision of autonomous networks.

 

Global cloud providers such as AWS, Google Cloud, Microsoft Azure, IBM Cloud, Oracle Cloud are new players in the development and implementation of the concept of autonomous networks. They not only actively stimulate telecom operators to implement innovative services and deploy next-generation network infrastructure, but also develop cloud platforms intended for operators of virtualized infrastructures to automate and intellectualize the almost complete stack of NMS/OSS/BSS processes necessary for deploying autonomous networks and providing innovative services based on them. The advantage of global cloud providers over vendors of telecom equipment and software is the possibility of pilot deployments on their own global infrastructure, which allows to quickly receive feedback and refine solutions to the level necessary for their transfer to the status of commercial operation.

 

The vendors’ solutions are the components required by operators to move from the traditional approach ("manual management") to the provision of end-to-end network services and prepare domains for programmatic management by a coordinating authority (orchestrator) or an external system. This will allow domain operators and end-to-end layer operators to implement consistent software management using end-to-end service orchestrators and MANO platforms to manage virtual resources in domains such as RAN, IP backhaul, IP and DWDM transport, VNF and MEC.

 

In this operating model, the high-level descriptions of cross-domain services (templates) are transformed into technical requirements, which are then implemented in the form of chains of service functions, individually configured for each individual user.

 

Currently, many domain operators already have the technical capabilities of fully software-based next-generation services within a single domain, but few of them have the ability to orchestrate E2E services, given, in particular, the fact that physical networking functions still coexist with VNFs, which, in turn, run primarily on virtual machines rather than containers. In fact, the industry is only at the beginning of defining cross-domain orchestration rules.

 

Analysis of product portfolios of telecommunications equipment and software vendors shows that even at the level of intra-domain autonomous management, their offerings do not yet cover all the needs of operators arising from implementation of autonomous network concept, and at the level of cross-domain management, vendor solutions clearly lack functional completeness and openness. At the same time, all the considered vendors of solutions for autonomous networks are already using technologies of cognitive thinking, machine learning and artificial intelligence capabilities to proactively identify and eliminate network events and ensure network security.

 

Methodology for assessing the maturity level of management systems used by operators of ICT infrastructure and providers of ICT services

 

The methodologies developed by standardizing organizations for assessing the level of autonomy and intellectualization of networks and services (Fig. 1) make it possible to estimate the level of maturity of control systems used by operators through an internal audit. However, these methodologies do not make it possible to assess the potential economic effect of the transition to a higher level of maturity, and also do not allow evaluating operators “from the outside” based on publicly available data. The methodology developed by J’son&Partners Consulting allows solving these problems using the classification of the levels of autonomy and intellectualization of networks and services proposed by standardizing organizations.

 

The J'son&Partners Consulting methodology is based on objectively assessed indicators describing the financial and economic effects of the implementation of autonomous intelligent network and service management.

 

Due to the presence of different business roles and, as a consequence, differences in approaches to the optimized management of networks and services, to assess the level of maturity two types of ICT operators should be distinguished:

 

Domain operators:


- telecom infrastructure, with a division into operators of access networks and backbone networks (the latter are automated faster);


- computing infrastructure - cloud IaaS / PaaS providers;


- Operators of end-to-end network layers - providers of OTT telecom services with high added value (SD-WAN, SECaaS, CDNaaS, UCaaS, etc.).

 

The set of indicators linking economic benefits to maturity levels consists of two blocks – financial-economic and technical-economic. The comparison of operators of two different types - end-to-end layers and domains, is possible only for the block of financial and economic indicators, and it is impossible for the block of technical and economic indicators. Thus, all operators of end-to-end layers have implemented virtualization technologies and software management, and the load level depends not so much on themselves, but on the ability of domain operators to provide domain resources "on demand", in this case, the load level of the layer operators exceeds 80%.

 

The financial-economic block of indicators:

 

- The level of operating costs for network management and service provision (excluding depreciation) relative to revenue. The target level is below 10% for L2 / L3 and below 5% for L4 / L5. The assessment method is the analysis of public financial statements.


- Share of cloud services (NaaS) in operators' revenue structure. The target level of services corresponding to the characteristics of the cloud is below 10% for L0 / L1, above 50% for L2 / L3 and above 80% for L4 / L5 for domain operators, 100% starting from L2 level for cross-domain service providers. The assessment method is the analysis of public financial statements.

 

The technical-economic block of indicators:

 

- Implementation of all five cloud principles for at least one type of telecom services provided - clearly characterizes the level of automation of NMS/OSS/BSS processes. Assessment method - analysis of public technical documentation.


- Availability of new generation ICT infrastructure elements, the operation of which is impossible without a high level of automation of NMS/OSS/BSS processes (example: 5G RAN / Core SA). Assessment method - analysis of public technical documentation and official reports.


- The level of the average load of resources (network, computing), separately for "best effort" and managed services, the level of load of network resources - separately for access networks and transport / backbone networks (for the backbone it is higher). The target level is above 30% for L2 / L3 and above 60% for L4 / L5, due to:


- Opportunities to proactively manage demand using digital models of ICT infrastructure and price elasticity models implemented using artificial intelligence technologies (now only the most advanced IaaS / PaaS providers have) - demand mapping in Fig. 1.


- Possibilities of cross-domain orchestration, that is, automatic interaction with other operators / providers (formed for VPC Edge and BoD services).

 


Analysis method - interviewing the operators.

 

The indicator of the level of operating costs for network management and provision of services relative to revenue has the greatest weight, since it is an integral indicator that characterizes the financial result of the automation of NMS/OSS/BSS processes. The indicator of the share of revenue from cloud services in the total revenue of the operator also has a large weight due to the significant influence of this indicator on the dynamics of the operator's revenue. In the technical and economic block, the indicators of the availability of a new generation infrastructure - 5G Stand Alone and MEC, and the level of resource utilization are significant. 

 

 

 

  

 

 

The detailed results are represented in full version of the report:


«Autonomous intelligent networks: analysis of economic outcomes and readiness of technologies»


List of content


1.DEFINITIONS AND METHODOLOGY


2.CONCEPTS FOR LIFECYCLE AUTOMATION OF NETWORKS AND SERVICES – KEY APPROACHES, TECHNOLOGIES AND ECONOMIC PRINCIPLES, HISTORY OF DEVELOPMENT, AND PERSPECTIVES


2.1.DEFINITIONS OF AUTOMATED AND AUTONOMOUS MANAGEMENT, LEVELS OF AUTONOMY AND INTELLIGENCE
2.2.ECONOMIC GOALS, KEY TECHNOLOGICAL AND ECONOMIC PRINCIPLES
2.2.1.Decrease of labor intensity and per unite cost of NMS/OSS/BSS processes execution
2.2.2.Ability to provide composite cloud telecom services
2.2.3.Improvement of average utilization of resources used for “on demand” services provisioning with managed SLA

 

3.AVAILABLE TECHNOLOGIES AND SOLUTIONS FOR CREATING AUTONOMOUS NETWORK AND SERVICE MANAGEMENT SYSTEMS USING ARTIFICIAL INTELLIGENCE
3.1.COMMON VISION
3.2.TM FORUM VISION
3.2.1.Autonomous networks
3.2.2.Evolution of the OSS / BSS reference architecture towards autonomous networks
3.2.3.TM Forum Catalyst project
3.3.ETSI VISION
3.4.ITU-T VISION
3.5.3GPP & GSMA VISION
3.6.VISION OF TELECOMMUNICATION EQUIPMENT AND SOFTWARE VENDORS
3.6.1.Telecommunication equipment vendors
3.6.1.1.Cisco Systems
3.6.1.2.Aruba Networks (HPE)
3.6.1.3.Juniper
3.6.1.4.Extreme Networks
3.6.1.5.Nokia
3.6.1.6.Ericsson
3.6.1.7.Alcatel-Lucent
3.6.1.8.Ciena
3.6.1.9.Huawei
3.6.2.NMS / OSS / BSS software vendors
3.6.2.1.Amdocs
3.6.2.2.Netcracker (NEC)
3.6.2.3.Oracle
3.6.2.4.Ericsson

 

4.METHODOLOGY FOR ASSESSING THE MATURITY LEVEL OF MANAGEMENT SYSTEMS USED BY OPERATORS OF ICT INFRASTRUCTURE AND PROVIDERS OF ICT SERVICES

 

5.IMPLEMENTED PROJECTS AND TYPICAL USE-CASES
5.1.NMS / OSS / BSS AUTOMATION AND INTELLECTUALIZATION PROJECTS AIMED TO IMPLEMENT NEW SERVICES
5.1.1.AWS Wavelength (AWS in partnership with 5G networks operators)
5.1.2.Global Mobile Edge Cloud (Google in partnership with 5G networks operators)
5.1.3.Azure Edge Zones with Carrier (Microsoft Azure in partnership with 5G networks operators)
5.1.4.IBM Cloud Satellite
5.2.NMS/OSS/BSS FRAGMENTARY AUTOMATION AND INTELLECTUALIZATION PROJECTS NOT ASSOCIATED WITH THE IMPLEMENTATION OF NEW SERVICES
5.2.1.Optimization of utilization and energy consumption of RAN, optimization of long-term planning of network development
5.2.2.Technical support optimization and management of wired networks

 

6.EXECUTIVE SUMMARY

 

List of figures


Fig. 1. Digitalization of an enterprise
Fig. 2. Levels of autonomy and intellectualization for network and services management
Fig. 3. Volume, structure, and dynamics of the global market of traditional telecom services during 2016 – 2020 and forecast for 2021-2025, bln. USD
Fig. 4. Volume, structure, and dynamics of global public cloud services market during 2016 – 2020 and forecast for 2021-2022, bln. USD
Fig. 5. Global SDN network deployed by Google and its utilization level
Fig. 6. Level of labor expenses on personnel directly executing NMS/OSS/BSS-processes compared to revenue, L0-L1 levels of autonomy (traditional telcos) and L2-L3 levels of autonomy(SECaaS, SDWANaaS and CDNaaS providers), %, 2020
Fig. 7. Level of R&D expenses compared to revenue, L0-L1 levels of autonomy (traditional telcos) and L2-L3 levels of autonomy (SECaaS, SDWANaaS and CDNaaS providers), %, 2020
Fig. 8. EV/EBITDA ratio and shares profitability by verticals, 2005-2018
Fig. 9. Capitalization dynamics of major traditional telcos and digital service-providers in 2015 and 2020, bln. USD
Fig. 10. Cost of banking transaction by way of its execution, 2011
Fig. 11. Distribution of an operators’ share (%) of broadband B2B subscribers between ranges of ARPU (rubles per month), 2017
Fig. 12. Possible distribution of an operators’ share (%) of broadband B2B subscribers between ranges of ARPU (rubles per month), Telco 1.0 (traditional telecom business-model) and Telco 2.0 (new telco business model based on network slicing)
Fig. 13. Estimation of OPEX increase for customized services provisioned using network slicing compared to universal “best effort” services been provisioned using single non-sliced network
Fig. 14. Sliced network OPEX comparison with non-sliced network for different levels of automation, %
Fig 15. Total expenses on sliced network compared to non-sliced network, %
Fig. 16. Implementation of design phase of complex service by ASDC module of ECOMP platform
Fig. 17. Functional blocs of ECOMP implementing BSS and OSS processes
Fig. 18. Multi-tier architecture of web applications
Fig. 19. Global WAN traffic structure by sources of its generation, Zb, 2021
Fig. 20. Global CDN, web-security, SD-WAN and SD-WAN security cloud markets, 2017 – 2023, mln. USD
Fig. 21. The concept of traditional and next generation services provisioning with use of single next generation NMS/OSS/BSS platform for autonomous intellectual management of all services and infrastructure elements, and combined network infrastructure
Fig. 22. Potential volume of global BoD and Managed NFV services consumption for hybrid clouds deployment during 2019-2025, mln. USD
Fig. 23. Potential volume of international B2O segments of BoD and Managed NFV consumption during 2019-2025, mln. USD
Fig. 24. Potential annual volume of global NaaS and distributed cloud computing with MEC, bln. USD, probably 2030
Fig. 25. Hybrid multiclouds: ways to provide connectivity between instances of distributed datacenters and hosted instances
Fig. 26. Weekly dynamics of traffic of a certain group of applications (i.e. traffic of logically isolated network slice) within an edge node
Fig. 27. Structure of traffic between nodes of distributed hybrid system of datacenters, consisting of edge and core datacenters
Fig. 28. Degradation of multiplication efficiency in case of static resources provisioning to slices (number of slices = 16) compared to those of non-sliced network, by level of network nodes
Fig. 29. Improvement of multiplexing efficiency in case of dynamic resources allocation to network slices (number of slices = 16) compared to those of non-sliced network, by level of network nodes
Fig. 30. Definition of end-to-end slice
Fig. 31. Principles of interaction between operators of domains, cross-domain (E2E) slice operator and industrial consumer of E2E slice
Fig. 32. E2E network slice and cross-domain orchestration defined by 3GPP Release #16
Fig. 33. Cross-domain orchestration (optimization) logic for autonomous vehicles E2E slice, which use resources supplied by ICT-infrastructure domains of 5G RAN, MEC, SDN-based transport and long-haul networks and core datacenters
Fig. 34. TM Forum Vision of Autonomous Networks
Fig. 35. The framework for building autonomous networks
Fig. 36. Business Process Framework top-level categories
Fig. 37. Top-level ODA functional architecture
Fig. 38. Functions of the Party Management block in ODA with second-level processes described in eTOM
Fig. 39. Functions of the Intelligence Management block in the ODA with the second-level processes described in eTOM
Fig. 40. Functions of Core Commerce Management and Production blocks in ODA with second-level processes described in eTOM
Fig. 41. Cognitive network management system architecture
Fig. 42. ZSM architecture
Fig. 43. ML architecture for autonomous networks
Fig. 44. Oracle Communications Service and Network Orchestration Functional Architecture
Fig. 45. Functional Management System Architecture for AWS Wavelength
Fig. 46. Closed Autonomous Control System Loop for AWS Wavelength
Fig. 47. Software-defined management of network and compute infrastructure AWS Wavelength, current look of the service - reconstruction based on public data
Fig. 48. Software-defined management of network and compute infrastructure AWS Wavelength, current look of the service - reconstruction based on public data
Fig. 49. The ability to combine virtualized utilities for different network tier instances
Fig. 50. Functional scheme of the Google Anthos platform used to automate NMS/OSS/BSS processes in distributed computing services
Fig. 51. The logical scheme for choosing a configuration for placing and scaling distributed elements of an application
Fig. 52. Features of the MS Azure for operators platform
Fig. 53. IBM Cloud Satellite components, their connectivity and responsibilities for managing them
Fig. 54. Networked IBM Cloud Satellite Distributed Components Using IBM Direct Link Connect 2.0.
Fig. 55. Assessment of the global structure of energy consumption of ICT equipment, fact for 2013, forecast for 2025
Fig. 56. Dynamics of peak power consumption of a typical base station of 2G-3G, LTE and 5G standards
Fig. 57. Optimization of RAN energy consumption as a complex task requiring the use of autonomous network technologies
Fig. 58. Coordination of load peaks on the RAN and the schedule of charge-discharge cycles of the base station batteries


List of tables


Table 1. Criteria for ranking the level of maturity of management systems for NMS / OSS / BSS-processes of operators, ranges of criteria values, their weights and compliance with the levels of autonomy and intelligence L0 / L5 according to ITU-T
Table 2. Example of ranking calculation for domain operators
Table 3. Example of ranking calculation for operators of end-to-end network layers

RU
EN