Network slices utilize Network Functions (NFs) and resources across the various network segments from Access to Core (and in many cases also deployed on a public cloud); therefore, managing the lifecycle of network slice instances poses a major operational challenge for the mobile operator. An end-to-end perspective is crucial to meet the needs of diverse services, use cases, and business models. The 5G network and telco cloud infrastructure layer hosts the physical and virtual resources that are necessary to create network slice subnets. These include both virtualization software and hardware, which is comprised of memory, computing, storage, and networking resources.

Figure 2: 3GPP Network Slice Management Functions.
Source: ETSI GR NFV-REL 010 V3.1.1 (2019-06)

Network Slice Lifecycle Management

Managing the lifecycle of network slices requires the modeling, deployment, instantiation, and ongoing maintenance and health checking of numerous NFs that are deployed in multiple and diverse virtual and physical environments. Furthermore, to gain maximum operational efficiency, NFs are instantiated and scaled in an automated, zero-touch operations mode. This enables 5G slices to deliver their promise of dynamic flexibility, agility and scalability, which in turn will allow provisioning the network slices, to enable them to satisfy various use cases and customer needs.

The CSMF is responsible for translating the communication service-related requirements to network slice-related requirements. The NSMF is responsible for the management and orchestration of network slice instances, as well as the translation of the end-to-end slice requirements for the various subnet requirements, and the NSSMF is responsible for the management and orchestration of network slice subnet instances.

5G network slicing management will be based on SDN and NFV technologies and principles, which enable dynamic programmability and control. Mobile operators will adopt new methodologies that will enable them to automate network slice life cycle management that is based on the modeling of resources and services. The automation of network slice operations will be based on creating and designing service models and operational policies in advance. These will enable closed loop operations that are based on fault and performance monitoring, as well as providing the ability to utilize advanced analytics and machine learning, to maintain the service intent. Furthermore, this mode of operation will optimize the consumption of network and cloud resources, while ensuring the attainment of service and slice KPIs.