logo_header
  • Topics
  • Research & Analysis
  • Features & Opinion
  • Webinars & Podcasts
  • Videos
  • Dtw

Intelligent customer experience profiles pave way for ‘perfect’ network operations and lower churn

The Autonomous customer experience index (CEI) for zero-touch 5G network - Phase III Catalyst is accelerating and improving zero-touch operations to help enable customer experience fit for the 5G era

Ryan Andrew
10 Nov 2023
Intelligent customer experience profiles pave way for ‘perfect’ network operations and lower churn

Intelligent customer experience profiles pave way for ‘perfect’ network operations and lower churn

Commercial context

Those working in 5G know that, for all its strengths, it’s a technology beset with complications. Network deployments, new use cases, virtualization, and open architectures each impose multiple challenges to network capability and operations – all of which can adversely impact customer experience. To resolve these issues, the industry’s response was to establish customer-centric 5G network environments – and then later a customer experience index (CEI) which used crowdsourcing data in conjunction with AI and machine learning to understand and score customer experience. While this made strides in understanding customer experience and remedying network issues, CSPs can still do more to stem service degradation before it occurs. The answer lies in zero-touch root cause analysis (RCA), modifications, and recovery – not just in real time, but even anticipating issues before they occur.

The solution

This is motivation behind the Autonomous customer experience index (CEI) for zero-touch 5G network - Phase III Catalyst, which uses the knowledge gained in previous project iterations to accelerate and improve zero-touch operations, and ultimately customer experience. The key focus in this phase has been how to use TM Forum’s technical guidance on closed-loop automation (TR284E), to define and test four common applications:

  • CEI diversification involves creating various CEIs each suited to applications used by the great variety of CSP customers. These are supported by AI/ML techniques, which can predict CEIs hours into the future.
  • User model generation means producing multiple ‘user type’ models (also known as ‘persona models’) based on the TM Forum asset Customer Experience Metrics IG1303. This use case demonstrates a customer experience index based specifically around the type of user connected to the network. This is used to support RCA and automate anomaly detection and issue management. This can be a trigger for orchestration and execution of corrective actions on the network layer to solve and bring the CEI back to normal level for the various different user types according to the persona model. After performing each corrective action, the status of the CEI is monitored in order to evaluate its effectiveness as part of the learning process.
  • 3D GIS representations have been built from CEI data, network coverage and other parameters at cell level. This enables analysis of CEI based on geography, service type (IM, streaming, gaming, other mobile apps) and at application level (based on the persona model). 3D GIS also makes it possible to calculate coverage routes from A to B, and therefore ensure a high CEI.
  • Automated synchronization and resolution of problems in 5G radio networks enables issues to be resolved through intelligently scheduled cell restarts. These take into account base station usage patterns and customer experience to determine the optimal preventive action window.

Application and wider value

Synchronization problems are a known root cause for customer experience degradation in 5G networks. Before total loss of service, the customer can have various other problems with phone calls and data sessions – for example handover failures, poor throughput and reduced voice quality.

The kind of breakthroughs demonstrated in this Catalyst are the staple underpinnings of autonomous networking that is transforming telecommunications as we know it. The Catalyst has created an inventory of emerging technologies capable of building networks that can diagnose, heal, optimize, and configure themselves automatically, thereby paving a route to the best possible customer experience at all times.

Central to this is the use of AI and ML which as demonstrated is a primary conduit for the provision of more sophisticated services with higher levels of quality and efficiency. ML-based algorithms have been shown to reduce the number of affected users by 30-40% on average. The reduction is especially relevant for sites with unusual traffic profiles. For example, this reduction is as high as 80-90% for sites with a night-heavy traffic profile.

Poor customer experience is a key reason for churn – if the fruits of this Catalyst are put into action, CSPs will take a major step in preventing that. On the flip side of this coin, we also know that a happy customer is likely to spend more with its CSP. This vast improvement to how detection is synchronized with resolution will ultimately enable a huge variety of budding applications that demand ultra-low latency, such as uninterrupted videoconferences, gaming sessions, high-fidelity streaming the future applications of autonomous vehicles.

According to Julia Martinez Arenas, Customer Experience Manager, Telefónica Global “there is a need to anticipate the customers’ pain points and proactively implement autonomous solutions before any degradation in services occurs. This will help ensure the best customer quality of experience while improving the network operational efficiency. Having a personalized and real time customer satisfaction index is key to fine tuning some new services in 5G such as network slicing, and allows a zero-touch automation to solve or mitigate any issues that may impact our customer's experience.”

Catalyst: Autonomous customer experience index (CEI) for zero-touch 5G network - Phase III