Short-Term Traffic Volume Prediction In Umts Networks: Validation of Kalman Filter-Based Model.

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Accurate traffic volume prediction in Universal Mobile Telecommunication System (UMTS) networks has become increasingly important because of its vital role in determining the Quality of Service (QoS) received by subscribers on these networks. This study explores traffic volume prediction and, adapts and validates the Kalman filter-based short-term traffic volume prediction model for UMTS networks. In this study, we adapt and validate the Kalman filter-based traffic volume prediction model which is used more in transportation engineering. The model was adapted based on two key assumptions that make it possible for us to characterize the short-term traffic volume patterns for UMTS networks to suit the Kalman filter algorithm. The model so adapted was carefully fine-tuned and implemented in MATLAB. The model was then validated with traffic volume data collected from a live 3G network using the graphical and r2 (coefficient of determination) approaches to model validation. The results indicate that the model performs very well as the predicted traffic volumes compare very closely with the observed traffic volumes on the graphs. The r2 approach resulted in r2 values in the range of 0.87 to 0.99 which compare very well with the observed traffic volumes. A little was done on sensitivity analysis of the model parameters, and this has been recommended for future research. The result obtained in this study brings out the fact that, the Kalman filter algorithm is very useful in predicting short-term traffic volumes for UMTS networks.
A Thesis submitted to the Department of Electrical and Electronics Engineering, Kwame Nkrumah University of Science and Technology in partial fulfilment of the requirements for the degree of Master of Science,