Prediction of Mobile Network Performance Using Supervised Machine Learning Models

Mobile network management and drive tests provide services that give a clear insight into the quality of mobile network coverage and other wireless networks including identifying areas of poor signal quality and identification of black spots. This research investigates the potential of supervised machine learning models to predict mobile network performance, focusing on Coventry University and Coventry City Center, United Kingdom. The measured download speed represented as Downlink bit rate (DL_bit rate) at these locations are used to obtain user performance data using Lebera mobile virtual network operator SIM card in order to assist with the visualization of the mobile network performance maps. The study explores the application of various machine learning models, including Random Forest, Linear Regression, Gradient Boosting, Support Vector Regression (SVR), and K-Nearest Neighbors (KNN), to forecast mobile network throughput speed (DL_Throughput). Among the models, the Random Forest Regression (RFR) model demonstrated the highest accuracy, with a Mean Absolute Error (MAE) of 6.79 and an R-squared (R²) value of 0.496, indicating its effectiveness in capturing network performance variability. The findings highlight the considerable promise of machine learning in optimizing mobile networks, enabling telecommunications providers to automate network management, enhance resource allocation, and improve the Quality of user Experience (QoE). The application of predictive analytics in this context offers significant advantages, including increased network efficiency, enhanced user satisfaction, and reduced operational costs. This study underscores the importance of integrating AI-driven predictive models into mobile network optimization strategies to meet the evolving demands of urban environments.