TITLE: Machine learning models for prediction of mobile network user throughput in the area of trunk road and motorway sections
AUTHORS: Zoran Ćurguz, Milorad Banjanin, Mirko Stojčić
ABSTRACT: User throughput in the telecommunications network stands out as one of the key performance indicators. Today, telecommunications service providers have the task of providing a reliable and secure connection for users in all geographical locations and at all times, and adequate network throughput to meet the growing need for streaming services. These requirements primarily apply to areas around important roads, such as motorways and trunk roads. The main goal of the research is to create models based on machine learning techniques for predicting the average user throughput in the M:tel network, in a geo-area that includes the section of Motorway “9th January” (M9J), Banja Luka-Doboj, between the Johovac node and the town of Prnjavor, and the area of the M17 trunk road section, between the Johovac node and the town of Doboj. Predictive models were created on the IBM SPSS Modeler software platform, and a comparative method was used to compare and select the models that show the highest prediction accuracy. The results have shown that k-Nearest Neighbors (k-NN)-based models have the highest prediction accuracy for both sections, with the model created for the trunk road section having significantly better performance.
KEYWORDS: Average user throughput, Predictive models, Machine learning techniques, k-NN
PAGES: 27-35
DOI: 10.59478/ATCT.2022.5