Real-Time Mobile Data Processing Analysis to Track LTE Signal Quality in Mobile Environments

Abstract

The proliferation of high data rate applications on mobile devices necessitates continuous enhancements to LTE networks for optimal performance and user satisfaction. This dissertation comprehensively investigates LTE signal strength and quality metrics, focusing on RSRP, RSRQ, SNR, and ASU, from data recording to advanced signal quality prediction using deep learning. Data was meticulously collected with the LTE-Signal-Logger application across various Algerian regions, capturing LTE signal dynamics under different mobility conditions, speeds, and times. The data was processed and stored in a MySQL database via a Python backend server for efficient retrieval and analysis. Visualization and exploration of the data revealed significant factors influencing signal strength and quality, including geographic location, time of day, and user mobility. Advanced deep learning models with hyperparameter tuning were employed to predict LTE signal quality accurately. The study’s results provide valuable insights into LTE signal behavior in real-world scenarios, demonstrating the potential of deep learning techniques to optimize network performance and improve user experience, contributing practical solutions for network operators and engineers.