5G Management and Orchestration Architecture Framework
Jan Groenendijk1 and Zou Lan2
1 Rapporteur of 3GPP SA5, Ericsson, Ireland
2Rapporteur of 3GPP SA5, Huawei, China
E-mail: jan.groenendijk@ericsson.com; zoulan@huawei.com
Received 04 March 2019; Accepted 20 May 2019
The 5G network is expected to be service oriented with the premise that it will support various services of vastly different requirements on configuration including automation, performance, accounting and network status reporting. In support of this premise 3GPP agreed to adopt a service based management architecture framework in Rel-15, providing amongst other aspects the flexibility for integration with other architecture framework(s) to jointly fulfil the 3GPP service and network management requirements.
Keywords: 5G, Management, Orchestration, Management Service, Management Function, Service Consumer, Service Producer, Service Interface, Management Data Analytics, Life Cycle Management.
The 3GPP management architecture framework developed by working group SA5 adopts a service-oriented approach, which is described as interaction between a Management Service (MnS) consumer and an MnS producer. For example, an MnS consumer can request operations from an MnS producer on fault supervision service, performance assurance service, provisioning service and notification service, etc. An MnS offers management capabilities. These management capabilities are accessed by MnS consumers via standardized service interfaces, depicted in Figure 1, composed of individually specified MnS components.
The basic elements of an MnS include a group of management operations and/or notifications agnostic of managed entities (Management service component type A), management information represented by an information model of managed entities (Management service component type B), and performance information of the managed entity and fault information of the managed entity (Management service component type C). The advantage of separating the basic elements into management service components is to allow the management service components to develop independently according the needs of the network technology managed, and the automation requirements from the network (and service) operations.
An MnS may contain the combination of the following:
(1) management service component type A and management service component type B
(2) management service component type A, management service component type B and management service component type C.
The instances of MnSs carry information about specified management service components in the metadata attributes. Figure 2 illustrates examples of management service instances with various management service components of type A, type B and type C (more details on entities in Figure 2 can be found in [2] and [3]):
Deployment (or the use) of management services is flexible and they may reside in different management layers. For example, a network provisioning service may reside at the network and network slice management layer, and a subnetwork provisioning service may reside at the subnetwork and network slice subnet management layer. SA5 recognizes the need for automation of management by introducing new management functions such as a Communication Service Management Function (CSMF), Network Slice Management Function (NSMF) and a Network Slice Subnet Management Function (NSSMF) to provide an appropriate abstraction level for automation. Having identified the need or existence of a CSMF in Rel-15 further work on management service aspects regarding this function is being carried out in Rel-16 timeframe.
In the service-based management architecture framework a Management Function (MnF) plays the role of either Management (MnS) Service producer or MnS consumer. An MnS produced by an MnF may have multiple consumers. The MnF may consume multiple MnSs from one or multiple MnS producers. An example of an MnF playing both roles (MnS producer and MnS consumer) is illustrated in the Figure 3.
Management functions may interact by consuming management services produced by other management functions. Figure 4 below illustrates multiple scenarios:
The interactions between the management service producer and a management service consumer follows one of the two following paradigms:
Figure 7 shows an example of a deployment scenario for management of a mobile network including network slicing.
In this deployment scenario:
A Management Data Analytics Service (MDAS) provides management data analytics to support network management and orchestration. MDAS can be deployed at different levels, for example, at domain level (e.g., RAN, CN, NSSI) or in a centralized manner (e.g., in a PLMN level). A domain-level MDAS provides domain specific analytics, e.g., resource usage prediction in a CN or failure prediction in an NSSI, etc. A centralized MDAS can provide end-to-end or cross-domain analytics service, e.g., resource usage or failure prediction in an NSI, optimal CN node placement for ensuring lowest latency in the connected RAN, etc. Figure 8 illustrates an example of deployment model of the MDAS:
In this deployment scenario, see Figure 9:
In Figure 10, the EGMF produces exposure governance management capabil-ity that operators can apply on MnF 1 MnS for exposing different derivation ofMnFIMnS to:
The road to 5G lies ahead of us and we are moving along it swiftly while leaving behind the service oriented 5G management specifications in 3GPP Release 15. The journey will not end here; the Release 15 specifications provide building blocks for standardisation of new management services, extended information models and new measurements, that enable network and service operations to be automated allowing the service provider or network operator to respond faster to their customers’ requirements and the requirements of the network evolution (or changing network technologies).
The journey of specifying service-based interfaces for management and charging of the evolving 3GPP 5G eco-system will continue into Release 16, including, but is not limited to new management features such as for, example communication service management, management automation, intent driven management, the integration of ONAP and 3GPP 5G management framework.
[1] 3GPP TS 28.533: “Management and orchestration; Architecture framework”. Rel. 15.1.0.
[2] 3GPP TS 28.531: “Management and orchestration; Provisioning”. Rel. 15.1.0.
[3] 3GPP TS 28.541: “Management and orchestration; 5G Network Resource Model (NRM); Stage 2 and stage 3”. Rel. 15.1.0.
Jan Groenendijk is a Rapporteur of 3GPP SA5 and works in Ericsson Solution Area OSS, Product Development Unit OSS in Athlone, Ireland. He has received his engineering degree in Electrical and Electronics Engineering from the Technical University Delft in the Netherlands. He has worked for KPN Research and KPN Network Services for ten years, after which he joined Ericsson Expertise Ireland in 1998.
He has been working as a standardization delegate representing Ericsson since 2015 in NGMN and from 2016 in 3GPP SA5.
Zou Lan is Huawei wireless OSS standard prime delegate. She has received master degree in Computer Science from Xi’an Jiao Tong University, China before she joined Huawei. She has strong technical background in network management and extensive experience in standards. She has been actively participating in standardization and research activities such as 5G network and network slicing management, self-organizing network, network virtualization, fixed and mobile converged management. Zou Lan is an active player both in 3GPP SA5 and ETSI ZSM. She has held numerous rapporteurships in 3GPP SA5 since 2008.
Journal of ICT, Vol. 7_2, 81–92. River Publishers
doi: 10.13052/jicts2245–800X.721
This is an Open Access publication. © 2019 the Author(s). All rights reserved.
1 Service Oriented Management Architecture Framework
1.3 Management Functions (MnF) and Management Services (MnS)
1.4 Interaction Paradigm Between MnS Producer and MnS Consumer
2 Examples of Management Service Deployment Scenarios
2.1 Deployment Scenario for Network and Network Slice
2.2 Deployment Example for Management Data Analytics Services (MDAS)