Network Slicing and the Intent-Based Network
February 15, 2022 - Julia Hogarty
The goal is to create an intent-based network with operating models for each network slice, defined based on specific connectivity needs.
Network slicing has been a concept and ambition trumpeted by 5G puritans for a few years now – An end-state nirvana demonstrating the fully realised potential of standalone 5G in all its glory. But what are we talking about when we say, ‘network slicing’? Network slicing, in its simplest terms, is the transformation of the network into a set of logical, virtualised network instances which share a common underlying infrastructure. Each of these unique ‘networks’ is designed to serve a defined business purpose, boasting its own dedicated network resources, and assigned specific performance targets. This would allow for new efficiencies to be achieved as operators are enabled to deliver precisely the right connectivity for a precise business outcome or use case. No waste.
Utilising Network Functions Virtualisation (NFV), Software-Defined Networking (SDN) and automation, operators are empowered to rapidly create new network slices to support specific groups of users, services, applications, or industry verticals. Network slicing will be an enabling evolution of 5G to allow service providers to deliver a plethora of new services with vastly differing requirements to an array of different types of users. Operators will be able to deliver on latency, throughput, and reliability requirements for specific services in a way that simply wouldn’t have been possible with previous technology generations. Moreover, different latency, throughput and reliability standards can be guaranteed for each network slice, depending on specific service demands. In theory, it will be the unlocking of 5G as far as what a cloud-based network is able to facilitate.
Arguably, the goal is to create an intent-driven or intent-based network, where the operating model for each network slice is automated and defined based on the specific connectivity requirements of each ‘slice objective’. The role of Network Data Analytics Function (NWDAF) will be critical in making this possible. As the 5G analytics function, NWDAF will be responsible for monitoring the service-level agreement (SLA) for each network slice as defined based on the specified intent of the slice. This lays the groundwork for network slice automation where an AI capability is informed by the NWDAF of the status of each of the network functions on a given slice. In turn, the network slice is independently able to self-adapt and auto-align with SLA requirements and overall business intent by means of predictive analytics and root cause isolation.
Automation will become essential for network slicing to work as operators will face the challenge of hosting hundreds, if not thousands, of network slice instances. With each slice requiring its own custom needs to be continually monitored and maintained, it would be simply impossible to operate manually. Instead, we will be looking at end-to-end automation performing zero-touch lifecycle management dynamically as variables such as service demand and service requirements change over time. It may have felt like a ways off yet, but network slicing is coming into view and is set to open up an array of new industry applications and revenue opportunities.