Effect of Data Traffic Patterns on QoS Parameters

Abstract

In the design and management of data communication networks the appropriate choice of traffic model that would reflect the exact pattern of the users’ traffic has always been an issue of study. In the literatures different models have been presented. This work presents the comparison of several data traffic models (most popularly used in the literatures) on the basis of their influence on the quality of service (QoS) parameters. It is intended that this comparison would place at the disposal of researchers the data traffic models that would assist in making effective worse-case or best-case analysis of data communication systems. Object oriented computer simulation modeling technique provided the general platform necessary for such level of comparison. In this work MATLAB simulink simevent simulation environment was employed. The results present exponential distribution as the traffic model to be employed for worst-case analysis while Gamma distribution is for best-case communication systems analysis.