Performance Evaluation of Empirical Path Loss Models for Some GSM Networks in a Tropical Urban Environment: A Case Study of Zaria City

Abstract

An accurate path loss prediction is essential for efficient wireless network planning, particularly in tropical urban environments where terrain and atmospheric conditions significantly influence signal propagation.  This study investigates the performance of four widely used empirical path loss models (Ericsson, COST-231, Plane Earth, and ECC-33) in Zaria City, Nigeria, representing a typical tropical urban environment.  The problem addressed in this research is the inaccuracy of conventional path loss models when applied to tropical urban environments without required local validation, leading to inefficient network planning and poor quality of service.  Field measurements of Received Signal Strength (RSS) were collected through GPS-aided drive tests using a spectrum analyzer and a digital weather station to simultaneously capture environmental parameters such as temperature, humidity and pressure.  The collected data were used to compare measured and predicted path loss values using statistical tools such as Root Mean Square Error (RMSE) and correlation analysis.  The results revealed that the COST-231 model achieved the lowest RMSE values (3.36 dB for MTN and 3.83 dB for 9 - Mobile), indicating the highest prediction accuracy in dense urban conditions.  The Ericsson model showed moderate performance, while ECC-33 and Plane Earth models exhibited significant deviations and poor suitability.  The findings are limited to measurements collected during the late dry season and should therefore be interpreted within that climatic context.  The study also provides practical guidance to mobile network operators, researchers, and planners interested in improving the performance of GSM networks in tropical cities.