Modeling and Development of a Novel Quality of Service Prediction Model for Global System for Mobile Communications Network using Artificial Neural Networks

Abstract

Quality of service (QoS) performance evaluation is an essential indicator in determining the efficiency of services rendered by an industry. Comparison of some key performance indicators with standard threshold values has been a major approach for determining QoS of Global System for Mobile Communication (GSM) in Nigeria. This comparative approach, which usually involves human involvement is prone to error. Thus, an automatic artificial neural networks (ANN) predictive QoS model was developed in the study presented in this paper. In carrying out the study, five key performance indicators (KPIs) data were collected form the GSM operator used. The collected KPIs parameters were used to develop a mathematical model that was transformed into the proposed automatic QoS predicted model using ANN. The developed QoS prediction model when evaluated was found to be accuracy and performed favorably well when compared with the manual approach being used by the Nigerian Communications Commission. The developed automatic QoS prediction model for this study is thus suggested as a better replacement for the current manual method based on its accuracy and non-human involvement in predicting QoS of GSM network investigated.