Assurance built for the 5G era: delivering high-quality services in the next generation

Free perspective

Communications service providers (CSPs) are rolling out 5G networks that will enable a wide array of use cases and innovative new services, many of which will support enterprises’ critical business functions and needs. CSPs must therefore ensure that the quality of their networks and the services provisioned over them remain high and that they meet increased expectations. Assurance tools for 5G networks and services must be purpose-built with new and diverse monitoring solutions that can gather data from a variety of sources. Tools such as containerised probes, OpenTracing and network telemetry can provide highly granular and real-time data in containerised 5G environments. These monitoring solutions must also collect service data, which can be correlated with network data to gain a holistic and end-to-end view of the network from the customer’s perspective, thereby enabling efficient troubleshooting and advanced automation capabilities.

Assurance solutions for 5G networks and services will need to have embedded machine learning/artificial intelligence (ML/AI). ML/AI will be needed to analyse, in real time, the reams of data streaming in from the network and it must be embedded at every layer in the assurance stack. This will enable the algorithms to be much more targeted towards the specific application and provide more robust insights and predictions of service quality. It will also create efficiencies in data processing and adaptation compared to systems that use external ML/AI. A library of ML/AI models will be required by network and service monitoring solutions to support different scenarios and services. This means that network engineers can select the most appropriate model for a particular use case. These models must be modular to enable simple replacements and updates to model components as they gain experience.