Multi-Domain Knowledge Planes for Service Federation for 5G & Beyond Public Working Group (MDKP-PWG)

Summary

The NIST Transformational Networks and Services group, of NIST’s Communication Technology Laboratory’s Smart Connected Systems Division, has formed the Multi-Domain Knowledge Planes for Service Federation Public Working Group which will investigate growing trends on Requirements for Enablers for Cross-Industry Sectors End-to-End (E2E) Services Innovation and Delivery Agility and how to leverage dynamic resource and assets so they can be available on-demand. Potential E2E service innovators require Enablers for Automatic Discovery of the availability and proximity of various kinds of networking resources/assets (networking devices and end-devices) and infrastructures to enable them to innovate and/or deploy services over resources or assets owned by different players. Infrastructure or asset owners may belong to different industry, enterprise, government and research institutions.

The Working Group will produce the following deliverables:

1. Produce a Generic Architecture Framework Blueprint (GAFB) that articulates the role GANA Knowledge Planes (KPs) Platforms should play as Anchors for the Federation of Autonomic/Autonomous Networks (ANs): Early Draft September 15, 2023; Stable Draft December 31, 2023; Final Version February 2025
2. Describe Use Cases on the KP-to-KP Federations and on APIs for GANA KP Integrations: Early Draft October 1, 2023; Stable Draft July 31, 2024; Final Version February 2025
3. Describe Requirements for Knowledge Plane (KP) Platforms Driven Networking (including KP Governance and KP-to-KP Federations) within the GAFB: Early Draft November 2023; Stable Draft April 2024; Final Version February 2025
4. Discuss How Zero Trust Principles can be applied in GANA KP-to-KP Federations. Stable Draft September 2024; Final Version February 2025

---

Kickoff Meeting: Thursday June 22, 2023 at 5:00PM ET. Email: mdkp [at] nist.gov (mdkp[at]nist[dot]gov) for MS Teams invitation

Next Meeting: - Meeting #10.1 Monday November 25, 2024 at 10:30 ET.

Meeting Frequency: It is anticipated that there will be monthly meetings by MS Teams on Mondays at either 5:00 pm or 7:00 pm ET. A schedule of meetings and times will be supplied. Membership Participation in the WG is open to all interested parties. There are no membership fees.

---

The charter of the group is here.

Information on how to join the group can be found at the Overview page here.

MDKP Meeting materials, Agendas, Notes, Documents are found here.

 

Description

A. Problem Statement

In the 5G and Beyond Era there are interesting growing trends on Requirements for Enablers for Cross-Industry Sectors End-to-End (E2E) Services Innovation and Delivery Agility and how to leverage dynamic resource and assets availability wherever and whenever so they can be available on-demand. Potential E2E service innovators require Enablers for Automatic Discovery of the availability and proximity of various kinds of networking resources/assets (networking devices and end-devices) and infrastructures to enable them to innovate and/or deploy services over resources or assets owned by different players. Infrastructure or asset owners may belong to different industry sectors (such as Communication Service Providers (CSPs) and their Public Networks (e.g. public 5G networks) sector, Enterprises and their private networks (such as private 5G networks) sector, Government owned information and communications technology (ICT) networks sector, etc.) or may even be research institutions that seek to make use of the ICT networks for on-demand usage. The suitable assets and infrastructures could be viewed as a mix of public and private 5G networks and end devices.

Also, in this 5G and Beyond era, Telecommunications Network Operators and other Service Innovators are in search for flexibility to offer and deliver services in markets where they do not own infrastructure assets but consider leasing on-demand any available infrastructures from owners in order to deliver services in an agile manner. This new vison calls for an Ecosystem that builds upon the following dynamic capabilities required to be introduced in ICT Networks:

1. Capabilities for on-demand dynamic service composition and delivery using various assets owned by participants from across various industrial sectors that may even seek to monitize the on-demand uasge of their assets (e.g. ICT networks or cloud resources, etc.)
2. Capabilities for autonomic network(s) and connectivity formation in self-managed (self-organized) manner (without need for human intervention) in response to agile service delivery needs/requests, emergence situations and predicted situations that derive from various communication needs, movements of humans and mobility, traffic patterns, societal responses to pandemics like has been experienced with COVID-19. NOTE: The concept of “network” is generalized and may involve ICT elements, organizations, service providers, and even humans.
3. Capabilities for enabling truly connected societies in terms of resilience and survivability in the event of physical catastrophes like earthquakes, infrastructures damages, etc.
4. Capabilities of the emerging so-called Autonomic/Autonomous Networks (ANs) to federate with each other as driven by business goals and/or in response to situations that require the ANs (e.g. autonomic/autonomous ICT networks) to intelligently collaborate by way of federation to provide E2E services that serve the needs of humanity

Also noteworthy, there is a growing trend on turning factories into hosting assets that can be federated with other infrastructures and ICT assets for end-to-end agile service delivery.

B. Need to leverage Emerging Standards on Autonomic/Autonomous Networking (ANs) to design and standardize the desired architectural framework blueprint

This new vision calls for standards-based Intelligent Platforms that drive the Autonomic Management and Control (AMC) of Networks and Services to possess the capability to federate with each other to enable E2E services innovation and delivery agility across multi-domains that may need to be involved in the process. Such federation capability among such platforms ensures that when there is a need for on-demand network and connectivity formations the intelligent platforms react proactively and reactively to create network services and connectivity across network domains and provide for the service and security assurance of the services in an End-to-End fashion. Such intelligent platforms are part of the growing trend of Autonomic/Autonomous Networks (ANs) that are characterized by Self-* features such as Auto-Discovery of Information and Context pertaining to Environment of AN’s operation, Self-Configuration, Self-Diagnosis, Self-Repair, Self-Optimization, Self-Defense, Self-Protection, Self-Healing, Self-Aware, etc. In the industry and standardization communities such Intelligent Platforms have been defined and characterized as Knowledge Plane (KP) Platforms. As of today, the KP platform concept has been formally defined and standardized by ETSI in ETSI TS 103 195-2 on Generic Autonomic Networking Architecture (GANA): An Architectural Reference Model for Autonomic Networking, Cognitive Networking and Self-Management of Networks and Services [1]. The ETSI GANA KP is defined in ETSI TS 103 195-2 as:

           a construct that exhibits cognitive capabilities and behaviours in performing the management and control of networks and services, and operates on network-wide views (including knowledge continuously gathered about the state and behaviours of network elements/functions) to dynamically program (configure) network resources and parameters, and adaptively and autonomically make changes to the network and services compositions and configurations in order to meet the objectives (e.g. business objectives) desired for the whole network in the best optimal way.

A GANA KP Platform is designed using various kinds of algorithms for its intelligence, such as Artificial Intelligence (AI) models, Machine Learning Algorithms (ML) that are employed by the so-called GANA Decision-making-Elements (DEs) of the KP Platform. The DEs also implement Complex Event Processing (CEP) for Autonomic Management & Control of the Network and Services under the responsibility of the KP as described in [1][4][10] and in ETSI TR 103 747. ETSI TS 103 195-2 defines various types of GANA KP DEs that may be implemented in a particular type of a KP platform depending on types of Managed Entities (MEs) of the underlying network infrastructure: such as Auto-Discovery & Auto-Configuration-DE; Routing-Management-DE; Mobility-Management-DE; QoS-Management-DE; Security-Management-DE; Forwarding-Management-DE; Fault-Management-DE; Performance- Management-DE; Resilience & Survivability-DE. In line with the emerging paradigm of End-to-End Network Disaggregation within the network infrastructure layer and in the management and control layer, [1][6][10] and ETSI TR 103 747 discuss various types of GANA KP Platforms, such GANA KP for Radio Access Network (RAN), GANA KP for Multi-Access Network; GANA KP for Multi-Layer SDN programmable Transport Network; GANA KP for Core Network; GANA for a Data Center (DC), etc. The GANA KP DEs run as software in the Knowledge Plane Platform and drive the objectives for self-* operations (self-adaptation, self-optimization, self-monitoring, etc.)  of the AMC for the network and services by (re)-configuring Managed Entities (MEs) through various means.

The ETSI GANA Reference Model [1] and its definition of the GANA Knowledge Plane (KP) concept is the de-facto standardized holistic framework for Multi-Layer Autonomic Management & Control (AMC), i.e. Multi-Layer Autonomics (hierarchical closed control-loops) and Multi-Layer AI/ML models for AMC. The ETSI GANA embraces and fuses various models for designing autonomic systems and networks to provide guidance on approaches that can be taken in implementing GANA DEs. The GANA Model embraces models such as the USA’s NSF-funded 4D Architecture project [2], Knowledge Plane for the Internet [3], IBM’s MAPE model [5], and other models that emerged over the history on research on Autonomic Computing and Autonomic/Autonomous Networking (ANs), as discussed in ETSI TS 103 195-2 and other GANA related documents.

The GANA Knowledge (KP) Platform concept has been adopted in the various International Standards SDOs/Fora such as NGMN [6], TM Forum [7], BroadBand Forum (BBF) [12] [16], IEEE [8][9], ITU-T [15], 3GPP [13], and other. A recent study [4] has established that Knowledge Planes (KPs) Platforms should play the role of Anchors for Federation of Autonomic/Autonomous Networks (ANs) for Enabling E2E Services Innovation and Delivery Agility Across Industry Sectors (as representing Multi-Domains) in 5G & Beyond Era. What is important to note is that the ETSI GANA KP concept can be realized/implemented in part or wholly in various forms. For example, the GANA KP for the traditional Radio Access Network (RAN) has been implemented as the so-called Centralized Self-Organizing Network (C-SON), while for the emerging Disaggregated RAN based on the O-RAN Alliance Architecture the GANA KP is in part implementable as the so-called Radio Intelligence Controller (RIC) Platform and its xApps or rApps (see [17]), whereas the GANA KP for the Core Network of a Mobile Network can be implemented as an Umbrella Platform that integrates with the 3GPP NWDAF/NWDAS and  3GPP MDAF/MDAS Functions via their Northbound Interfaces to utilize and dynamically program the analytics services of those functions/services for the KP’s autonomic operations over the core network.

---

REFERENCES

1. ETSI TS 103 195-2 (published by ETSI in May 2018): Autonomic network engineering for the self-managing Future Internet (AFI); Generic Autonomic Network Architecture, Part 2: An Architectural Reference Model for Autonomic Networking, Cognitive Networking and Self-Management
2. USA NSF sponsored Research under ITR Awards ANI-0085920, ANI-0331653, and ANI-0114014: Albert Greenberg, et al: A clean slate 4D approach to network control and management: In ACM SIGCOMM Computer Communication Review Homepage archive Volume 35 Issue 5, October 2005, Pages 41-54
3. David D. Clark, Craig Partridge, and J. Christopher Ramming. A knowledge plane for the Internet. In SIGCOMM, pages 3–10, 2003.
4. Ranganai Chaparadza; Muslim Elkotob; Benoit Radier; Tayeb Ben Meriem; Taesang Choi; Robert B. Bohn; Abdella Battou; Tao Zhang, "The Era of Knowledge Plane (KP) Platforms Driven Networking—Anchor for Federation of Autonomic/Autonomous Networks (ANs) Across Industry Sectors," 2022 13th International Conference on Information and Communication Technology Convergence (ICTC), 2022, pp. 950-955, doi: 10.1109/ICTC55196.2022.9952650.
5. IBM White paper: An architectural blueprint for autonomic computing: June 2005
6. Autonomic Networking, AMC and ETSI GANA in 5G End-to-End Architecture Framework by NGMN® Alliance: "P1-Requirements and Architecture: NGMN® 5G E2E Architecture Framework v3.0.8": https://www.ngmn.org/publications/5g-end-to-end-architecture-framework-v3-0-8.html  
7. TM Forum ODA Intelligence Management GB1022 adoption of ETSI GANA Knowledge Plane Concept and GANA design principles
8.  IEEE NGR (International Network Generations Roadmap) 2022 EDITION; IEEE INGR (International Network Generations Roadmap)/Future Networks, Standardization Building Blocks (SBB) Roadmap Chapter
9.  IEEE NGR (International Network Generations Roadmap) 2022 EDITION;            IEEE INGR (International Network Generations Roadmap)/Future Networks, Systems Optimization Roadmap Chapter
10. White Paper No. 4 of the ETSI 5G POC: ETSI Gana as Multi-Layer Artificial Intelligence (AI) Framework for Implementing AI Models for Autonomic Management and Control (AMC) of Networks and Systems; and Intent-Based Networking (IBN) via GANA Knowledge Planes (KPs): https://intwiki.etsi.org/index.php?title=Accepted_PoC_proposals 
11. ETSI TR 103 747 (published by ETSI in 2021): Federated GANA Knowledge Planes (KPs) for Multi-Domain Autonomic Management & Control (AMC) of Slices in the NGMN® 5G End-to-End Architecture Framework
12. ETSI GANA autonomics onto BroadBand Forum (BBF) architectures (ETSI TR 103 473 V1.1.2)
13. ETSI GANA autonomics onto 3GPP Backhaul and EPC Core Architectures (ETSI TR 103 404)
14. ETSI GANA autonomics onto Heterogeneous Wireless Access Technologies using Cognitive Algorithms (ETSI TR 103 626)
15. ETSI GANA autonomics in ITU-T IMT2020 Architectures (ITU-T Y.3324)
16. ETSI GANA Model Adoption in BroadBand Forum (BBF)’s AIM Framework in TR-436 Access & Home Network O&M Automation/Intelligence
17. O-RAN Alliance Architecture Specification with Non-Real Time RIC Platform and rApps, and Near-Real Time RIC Platform and xApps concepts is available at   https://www.o-ran.org/