An Intelligent Predictive Model for Call Drop Management

Many factors added to the poor quality of services rendered by Global System for Mobile Communication (GSM) operators,one of the factors is call drop.Call drop is also one of the most important key performance indicators (KPI) in measuring customer's satisfaction. The data obtained from UCL repository was trained with a cortical learning approach to determine the threshold value. The value was compare with value obtained during drive test to determine the quality of network performance. In order to achieve this, six indicators for network performance was used such as network performance is poor and there is a call drop, is excellent, there is no call drop or is fair, call drop may occur. Again the system was able to determine the quality of network performance in the area of study with predictions for call drops using the aforementioned indicators. The results also as explained has proved a novel approach and the prediction will encourages customers' confidence in the use of different GSM network and will also increase the business potentials of the operators.