How to navigate the journey to 5G network slicing and autonomous operations
Introducing QoS into the mobile arena
The race towards 5G monetization is setting the stage for service providers to leverage the full power of network slicing. With its ability to support diverse services, each with specific performance requirements on a common network platform – network slicing is a powerful 5G enabler that brings compelling commercial opportunities.
Of course, network slicing is not new. Slicing on 4G fixed networks has already been implemented by some operators – Vodafone and Swisscom for example. But while most of those implementations are based on a ‘one-size-fits-all’, ‘best-effort’ delivery mode with no guaranteed quality of service (QoS), it’s a totally different ball game with 5G.
The introduction of 5G network slicing end to end across the mobile network, from access through transport to the core, brings quality of service into the limelight. 5G technology allows network slicing to create a truly differentiated quality of service for enterprises as well as consumers – each with specific use case requirements, such as online gaming, autonomous cars, industry 4.0, telemedicine and many more.
“We are not building infrastructure for the customer, we are building it for society. How can we build a network which will not fail?”
That's according to Telefonica’s CTIO, Enrique Blanco at the 2019 5G Core Conference in Madrid. Enrique and other service provider leaders know that in order to enable the 5G revolution and deliver their customers’ quality and performance demands, network failures cannot be tolerated, irrelevant of severity, location or impact.
To effectively live up to this challenge, we’re seeing solution architects increasingly promoting an autonomous vision for 5G network operations, based on a ‘closed-loop approach’ to monitoring and assuring the performance and availability of network functions that underpin each network slice. The expectation is that with this approach, the majority of problems will self-heal.
But achieving this level of fully autonomous network slice operations means first overcoming some existing operational challenges:
- Not all problems can be mitigated by simply scaling/re-provisioning network resources for the network slice
- Rules (or policies) for how the network resources supporting the slice will dynamically ‘breathe’ (scale/optimize/share) are very complex, and it will take time to develop trust in the autonomous mode of working
- Today, most service outages are typically due not to a single failure, but to multiple co-existing events, such as outages, rerouting and non-optimal configurations
- The process of troubleshooting end-to-end service issues requires operations to navigate increasingly complex configurations/events/counts that span many systems
5G network slicing demands a new approach to network operations
The most fundamental aspect of future network slice use cases (aside from slice creation and orchestration), is full-service lifecycle management with real-time closed-loop operations, taking into consideration the following elements:
- A network slice is about the service that it serves rather than the network elements supporting it. It’s therefore imperative to consider the business SLA as the primary aspect of network slice operations in terms of capacity planning, redundancy design and resource prioritization
- Successful operation of a network slice must be handled by a cross-domain solution that spans the RAN, transport and core networks, as well as multi-network vendors, to support the E2E service
- To realize and support critical services in real time with zero failure (or at least 7 9s availability), it is mandatory to leverage machine learning in some of the network slice use cases. This will enable the processing of billions of events, counters and topology states to extract meaningfully actionable insights and next-best optimization/resolution recommendations.
Ultimately, emerging enterprise use cases and industry verticals like automated agriculture, mining and manufacturing, delivery drones, and telemedicine represent the holy grail for 5G network slicing. But since it will take time for these capabilities to fully mature, it makes sense to start the network slicing journey now with a strategy that creates quick wins and time to value. Beginning with slices that support operational efficiency and consumer-focused services (like enhanced mobile broadband for gaming) will help achieve those objectives.
Discover how you can unleash the potential of 5G with business-focused, automated and adaptive network slice lifecycle management.