What if you could predict the quality of a cellular radio link?

Link Quality Prediction: delivering better service performance and improved network efficiency with predictive network and service characterisation

With the Standalone 5G New Radio specifications agreed in 2018 we are edging ever closer to the 5G vision of limitless connectivity. 

At the same time, the fundamentals of the telecoms network remain challenging. Consumer demand for mobile data continues to grow whilst willingness to pay remains static in most markets leaving operators in the position of delivering more data, for lower revenues. 

Higher value services, such as mission critical industrial or commercial applications, have been difficult for telecom operators to access since existing networks operate largely on a best-efforts basis.

Against this background, approaches that that help utilise network resources more efficiently will allow mobile operators to get a better return on their network investment either through a reduction in network costs and/or an improved quality of service which results in a commercial return (reduced subscriber churn for example).

Emerging machine learning technologies offer opportunities for a step change in cellular performance

We recently conducted a study on behalf of Intel and SK Telecom to identify use cases for a new machine learning technology which does exactly this.

Intel’s Link Quality Prediction (LQP) protocol draws on multiple sources of historical and real-time network and non-network data to forecast the quality of any given RAN link. The network or service can then dynamically respond to optimise the end-user’s quality of experience and the network performance.
 
The LQP protocol was developed by Intel and is currently undergoing trials with SK Telecom who are training the LQP prediction algorithms to conduct automated network management.
This week we have published the findings of our study on the use cases for LQP, and the ways in which it could benefit operators.

Our report can be downloaded at www.cambridgeconsultants.com/linkqualityprediction

This is an example of how we're able to combine the experience of our dedicated technology strategy consulting team with deep engineering and sector expertise to provide actionable insights to our clients.

I am at MWC this week with a stand at Hall 7, 7B21. 

Please email me if you would like to meet our team at the event – we will have experts attending to provide insight across technology strategy, ultra-reliable wireless design and service innovation.
 

Author
Michal Gabrielczyk