DOI: 10.36347/sjet.2018.v06i10.003

Pages: 293-305

Call blocking is the inability of mobile users to gain access to the network. Despite the strong signals displayed on the subscriber’s phone, several attempts to make successful calls usually end in failures. This research work was focused on an adaptive traffic modeling for reducing call blocking probability in a wireless mobile communication system. It was aimed at developing a good traffic model that will enable the service providers to utilize the allocated channels as efficiently as possible so as to reduce call blocking probability. Etisalat Nigeria mobile network was used as the study field. Firstly, the traffic pattern of the network was determined. This was achieved by using the performance and monitoring tool to measure the system traffic volume. Then, the channel utilization of the network was obtained by measuring the busy hour attempted call, successful call during busy hour, busy hour blocked call, available traffic channels using the key performance indicator. Erlang B model was used to evaluate the system performance using the data obtained. This was done by varying the traffic load with increasing and decreasing number of channels. An adaptive traffic model using artificial neural network was then developed to minimize the call blocking probability. The developed model was simulated using artificial neural network fitting tool. Using a developed algorithm, the model was compared with the fixed channel allocation and dynamic channel allocation. The results obtained showed that the artificial neural network maintained the lowest call blocking probability. This model has an improved performance over the existing system.