The introduction of 5G will create a more dynamic network environment, enabling an incredible assortment of new services and business models across new markets.  To address the different needs and requirements of all these new services comes the creation of individual network ‘slices’ that act as mini-networks, delivering a specific mix of features to meet the needs of each service.  These network slides will need to be closely managed to ensure the level of service being delivered is what is required based on the context of the service being provided-  KPIs they need to meet, and the level or type of connectivity required.   This ‘contextual service assurance’ will give communications service providers the ability to understand and therefore manage the impact of one service type over another according to what is being provided. Contextual conditions could include the importance of the service, its likely duration, performance requirements, resource availability, contractual KPIs that need to be met, and prevailing network conditions.

At TEOCO, we believe Contextual Service Assurance will be a key component for successfully delivering 5G services, and for this reason we have partnered with Heavy Reading on a whitepaper and upcoming webinar that share the name:  Contextual Service Assurance in 5G: New Requirements, New Opportunities. In both, we take a deep dive on the topic of contextual service assurance and why it is critical for 5G success. Here is a sneak peek at some of the areas that are addressed:

The implications of new RAN capabilities
5G will stand or fall on the radio network. The 5G air interface is highly configurable in order to scale across spectrum bands and deployment scenarios. This will allow for services to span multiple network domains, including radio, transport and core networks. However, understanding how these domains map to the underlying infrastructure and what the service and network context is will be critical to effective service assurance. How can CSPs obtain the contextual information required to support the delivery of new services, such as unlicensed or shared spectrum services, while upholding their associated SLAs and retaining the network economics?

Virtualization and Cloud need greater insight into underlying platform performance
Decoupling applications from hardware has major implications for fault, configuration, accounting, performance, security and service assurance generally.  This impact is multiplied when VNFs are deployed on shared in a multi-vendor cloud infrastructure. This is due to the responsibility to monitor the performance of the infrastructure as it scales out virtual network functions (VNFs) according to demand (or to recover in the event of failure) falls to the NFVI. So, how do CSPs get the contextual information that is required to provide insight into the performance of the underlying technology platforms?

Network slicing needs context to manage service experience to resource use
Network slicing, or the ability to support diverse services on a common network platform, with each slice optimized to the specific functional requirements of a customer or application, is deemed as one of the more commercially promising features of 5G. While we have gone into depth on Network Slicing in one of our earlier blogs ‘We’re going to need a bigger slice’, understanding the interplay between slices competing for resources and the service experience is at the heart of contextual service assurance in 5G. How do CSPs correlate service experience with the underlying resource use to ensure a slice is not consuming more than it needs, or that sufficient capacity is available for scale-out if the service reaches a capacity or performance threshold?

Closing the Loop
Automation depends on feedback loops between the service experience, network configuration and resource utilization. One major challenge for orchestrators to automate the process of making changes to network configuration (to actuate) is how to make them aware of the state of the end-to-end network service and what role each domain is contributing to a satisfactory or unsatisfactory performance. How do CSPs provide contextual information to the domain orchestrator via a closed-loop relationship that allows for continuous optimization of services and resources?

Machine Learning use cases continue to grow
Artificial Intelligence and machine learning will play a key role in the networks of tomorrow, but early progress is already being made by some CSPs.  Root cause analysis, network optimization and security are just a few of the areas where it is beginning to play a role. Will natural language processing and AI- enabled optimization come next?