GÉANT Whitepaper May 26, 2021

Image Added(pdf)

5G

5G OAV Architecture Analysis

5G is a mobile technology that is addressing, by design,abroad range of requirements: very high data rates, very low latency, low energy consumption, high scalability, improved connectivity and reliability, and improved security [FGPPP18], [FGPPP20]. The 5G specifications focus on serving mobile operator needs in terms of extreme mobile broadband services and include several features to support vertical industries in terms of enablers for an industrial Internet of Things (IoT)and Ultra Reliable Low Latency Communications(URLLC)[FGPPP20]. Within the 5G system, end-to-end (E2E) network slicing, service-based architecture, Software-Defined Networking (SDN) and Network Functions Virtualisation (NFV) are seen as the fundamental pillars of the 5G architecture.

EOSC

EOSC OAV Architecture Analysis

The European Open Science Cloud (EOSC) initiative [EOS20] is part of the EuropeanCommission’s effortto increase the competitive digital economy in Europe. The main idea of the initiative is to support Open Science, which in turn will enable a more agile approach to innovation.  The main goal of EOSC is to provide researchers from different research fields with the means necessary not only to conduct research, but alsotomake their research outcomes open to the wide community. Thus, in essence, the end goal is to create a cloud-based virtual environment that will provide a number of open services that can be used to store, manage and analyse research data.Six different actions are being implemented in parallel to achieve this goal, described in the EOSC implementation roadmap [ETS20]. One of the action lines focuses on the architecture of the virtual environment,whichshould achieve a federated, seamless view of all available research infrastructures and services built on top.

Image Added

EOSCtechnicalarchitecture–interactionsbetween
thematic and common services

ETSI GANA

ETSI GANA OAV Architecture Analysis

One of the main requirements for future architectures and digital service providers is to be able to transform all manual management of network and services into Automated and Autonomic Management & Control (AMC) so that the system dynamically responds to the changing environment. This behaviour is also known as autonomics and is the main focus of the Generic Autonomic Network Architecture (GANA) [GANA16]. In other words, the GANA reference model provides a way to build autonomic architecture using a generic AMC framework that can integrate with any type of network management architecture.

The ETSI Network Technologies (NTECH) Autonomic network engineering for the self-managing Future Internet (AFI)workinggroup leads the AMC standardisation efforts that will transform the traditional networking environment into a smart and intelligent network with a number of self-* features including self-healing, self-configuration, self-optimisation and others.

Image Added

 AFI GANA reference model

ETSI OSM

ETSI OSM OAV Architecture Analysis

ETSI Open Source MANO(OSM)[EOSM20] is a management and orchestration solution that complies withtheETSI NFV architectural framework. While it offers its own Virtual Infrastructure Manager (VIM),it also supports many third-party VIMssuch asOpenStack, VMware vCloud and VIO, Amazon Web Services, Azure and Google Cloud and,since release seven,also Kubernetes. This last integration allowsthedeployment ofmore than 20,000 pre-existing production-ready Kubernetes applications, with no need of any translation or repackaging, which are added to the steadily increasing number of Virtual Network Functions (VNFs)commercially or freely available.

Image Added

 ETSI NFV reference architectural framework

ETSI ZSM

ETSI ZSM OAV Architecture Analysis

TheEuropean Telecommunications Standards Institute(ETSI)Zero-touch network and Service Management(ZSM)working group has the aim of defining a universal framework to enable end-to-endzero-touchserviceautomationinvolvingdifferentmanagementdomains.The proposed framework[EZSM]focuses on closed-loop service automation based on monitoring and high-quantity data collection, to which machine learning and artificial intelligence proceduresapply. The end-to-end automation procedures considered by ETSI ZSM comprise all operational processes and tasks, including delivery, deployment, configuration, lifecycle management, assurance, optimisation and finalisation. It aims to provide 5G E2E network slicing and management capacity, ideally with 100% automation or zero human intervention,and aims to be general enough to integrate current and future networks and services.

Image Added

 ETSI ZSM reference architecture

GVM

GVM OAV Architecture Analysis

The Generalized Virtualization Model (GVM) describes an architecture that supports managing the lifecycle of different network topologies as virtual network slices over a physical infrastructure in a fully automated fashion. These virtual networks are environments where abstracted virtualised objects,called resources, can be defined, instantiated and arranged to create application-specific insulated networks and services. This process is performed via agraphicaluserinterface (GUI) and without the need for human intervention from network operationsor engineering teams.

Image Added

 Virtualisation, management and user control layers in GVM

MEF LSO

MEF LSO OAV Architecture Analysis

Metro EthernetForum Lifecycle Service Orchestration(MEF LSO) [MSOS16] is a layered abstraction architecture defined to build the systems for coordinated management and control across all network domains responsible for delivering an end-to-end connectivity service. LSO provides interoperable automation of operations over the entire lifecycle of Layer 2 and Layer 3 connectivity services,as shown inFigureB.2. This includes design, fulfilment, control, testing, problem management, quality management, billing & usage, security, analytics and policy capabilities. The architecture specifies high-level functional requirements, the functional management entities and the management interface reference points (logical points of interaction) between them.

Image Added

 LSO reference architecture

Open Baton

Open Baton OAV Architecture Analysis

Open Baton [OB20] is an open source platform developed by Fraunhofer FOKUS andtheTechnical University ofBerlin that enables implementation of an NFV environment based on the ETSI NFV MANO architecture framework [ETSINFVSPEC].

Open Baton uses an agile development process to provide a customisable network service orchestration framework based on the MANO NFVO. The framework supports multiple sites that provide heterogeneous Network Function Virtualisation infrastructure. Using the generic Virtual Network Function Manager (VNFM) the architecture can manage a large set of different VNFs.

Image Added

 Open Baton architecture release 5

ONAP

ONAP OAV Architecture Analysis

The Open Network Automation Platform (ONAP) [ONAP20] arose after the combination of two Network Function Virtualisation (NFV) orchestration and management initiatives: Enhanced Control, Orchestration, Management & Policy(ECOMP), led by AT&T, and Open Orchestrator(Open-O), led by China Mobile.

Image Added

 ONAP architecture

SENSE

SENSE OAV Architecture Analysis

SDN for End-to-end Networked Science at the Exascale (SENSE) [SEN20], [MON18]was developed in a research project with the goal to provide a network architecture that will facilitate the rapid deployment of smart network services in support of science applications. The main contributors to this project are ESnet/Lawrence Berkeley National Laboratory, California Institute of Technology, Fermi National Accelerator Laboratory, Argonne National Laboratory and the University of Maryland College Park.
The architecture was developed with next-generation big data/exascale science workflowsin mind that will require intelligent network services in cloud computing, AI and machine learning. The idea is to support scientists and their science workflows with a system that will automatically build virtual guaranteed networks over large-scale distributed storage and computing infrastructures without any human intervention and without the static non-interactive network infrastructures in place today.

Image Added

 SENSE architecture

TALENT

TALENT OAV Architecture Analysis

Terrestrial Satellite Resource Coordinator (TALENT) is a coordination tool which provides end-to-end services over satellite domain, radio system, cloud and Mobile Edge Computing (MEC) resources [TAL19]. It provides a single and easy-to-use point of interaction for all stakeholders involved in the ecosystem, such as terrestrial and satellite operators as well as different 5G verticals. TALENT is completely inline with 3GPP and ETSI definitions, extending them towards satellite systems.The original idea of TALENT is based on the definition of hierarchical and distributed orchestration [HDO17], where an overarching orchestrator is able to manage and coordinate several independent domains (satellite, radio and cloud). It is assumed that a domain manager (DM) exists in each domain that can work with the resources of this domain. In this sense TALENT becomes a light, scalable and efficient solution, agnostic to elementsof domains coming from various vendors.