Identification and Modeling of Dominant Causes of Congestion In A Cellular Network
Akut E. K B.Eng · IOSR Journal of Electronics and Communication Engineering · 2012
In this paper, teletraffic data were collected from Airtel Nigeria Ltd. for a period of 18 months.Frequencies of occurrence of congestion in the network were derived from the teletraffic data using the Nigerian Communications Commission (NCC) benchmark.The dominant causes of congestion in the network were identified from the data and their value of occurrence was seen to be random in nature.A Poisson probability distribution model was therefore implemented for the prediction of congestion in the network.The identified four dominant causes of congestion and their percentage relative weights based on frequency of occurrence were: (i) X 1 : Faulty System (50.63%); (ii) X 2 : Faulty Trunk(s) (29.11%); (iii) X 3 : Power Failure (16.46%) and (iv) X 4 : Cable Cut (03.80%).MATLAB (M-file Data Acquisition Toolbox) programming environment was used to write a software program in C++ language that simulated the four dominant causes of congestion.Pearson Product Moment Correlation Coefficient formula was used to test the relationship between the measured and simulated congestion.The value r = 0.7487 was the correlation coefficient obtained between the measured and simulated results.This shows that the model is reasonably reliable.