WCDMA Mean User Throughput Prediction Using Linear Regression Algorithm | IEEE Conference Publication | IEEE Xplore

WCDMA Mean User Throughput Prediction Using Linear Regression Algorithm

Publisher: IEEE

Abstract:

The services that can be offered by a telecommunication operator can be evaluated based on many Key Performance indicators (KPIs). The mean throughput perceived by the us...View more

Abstract:

The services that can be offered by a telecommunication operator can be evaluated based on many Key Performance indicators (KPIs). The mean throughput perceived by the user is one of the most important KPIs an operator has to monitor and this KPI depends on the total traffic carried by the network. So the operators should have a clear idea about the current mean user throughput that it can provide and the estimation of the mean user throughput in the coming months to plan the services and the offers that can be suggested to the clients. In this paper, we use a training data set that contains the daily evolution of the two KPIs (mean user throughput and total traffic carried over the network) for a mobile operator during one year and we implement the machine learning linear regression technique to prove that the mean user throughput can be estimated by a linear function of the total traffic carried over the network using gradient descent algorithm. Finally, we discuss several case studies to illustrate the potential of the optimum linear function in enabling the adjustment of network services, from the dimensioning perspective.
Date of Conference: 19-21 February 2019
Date Added to IEEE Xplore: 11 April 2019
ISBN Information:

ISSN Information:

Publisher: IEEE
Conference Location: Paris, France

References

References is not available for this document.