Utilization of Machine Learning Regression Models for Predicting Coverage in Celular Network

With the rapid advancement of telecommunication technology, particularly the rollout of 5G networks, predicting signal coverage accurately has become essential for optimizing base station placement and assessing network efficiency. This study aims to develop and evaluate machine learning regression models to predict signal coverage in both urban and suburban areas. We explore several regression algorithms, including Linear Regression, Ridge Regression, Lasso Regression, Decision Tree Regression, Random Forest Regression, Gradient Boosting Regression, Support Vector Regression, K-Nearest Neighbors Regression, and XGBoost Regression. Data for this study were collected through drive tests using G-Net Track Pro and cellular operator data, which included parameters such as signal strength, transmitter and receiver heights, the distance between transmitter and receiver, frequency, altitude, elevation angle, tilting offset angle, and horizontal and vertical distances from the boresight. The performance of these models was assessed using metrics like Mean Absolute Error (MAE), Mean Square Error (MSE), Root Mean Squared Error (RMSE), and R-squared <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\mathrm{R}^{2})$</tex>. Among the tested models, Random Forest Regression demonstrated the highest accuracy, with an MSE of 9.56, MAE of 2.08, RMSE of 3.09, and an <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{R}^{2}$</tex> of 0.85. These results highlight the potential of Random Forest Regression as a reliable tool for predicting 5G signal coverage in various environments.