Characterization and simulation of the LAN traffi c by MMPP model

This paper presents a method to characterize and simulate the traffi c of any Local Area Network (LAN) which displays a self-similar behavior. We used the queuing model D-MMPP/D/1 to generate a discrete stream and to analyze its behavior when passing trough the queue using different buffer capaciti...

Full description

Bibliographic Details
Main Authors: Jaime Alberto Moreno Mogollón, Jhon Jairo Padilla Aguilar, Vladimir Escobar Ordóñez, Andrés Fernando Correo Montesino
Format: Article
Language:English
Published: Universidad de Antioquia 2014-03-01
Series:Revista Facultad de Ingeniería Universidad de Antioquia
Subjects:
Online Access:https://revistas.udea.edu.co/index.php/ingenieria/article/view/18810
Description
Summary:This paper presents a method to characterize and simulate the traffi c of any Local Area Network (LAN) which displays a self-similar behavior. We used the queuing model D-MMPP/D/1 to generate a discrete stream and to analyze its behavior when passing trough the queue using different buffer capacities. This method has been tested by other authors under controlled-laboratory conditions for incoming traffi c only. We demonstrate that this methodology can be used for LAN networks under real and present conditions for incoming and outcoming traffi c. We used the traffi c captured on Internet input/output links from Universidad Pontifícia Bolivariana Seccional Bucaramanga´s LAN. Additionally, this methodology solves the problem of modeling self-similar traffi c due to its capacity of generating a simulated stream approximated to the measured stream. The exactitude of the model is shown in the quantile (Q-Q plot) between the measured stream and the one generated by D-MMPP. In our study, the model and algorithms were initially validated using streams provided by Bellcore laboratories. Furthermore, we performed some adjustments in order to obtain a simulated stream in terms of bits per second (bps) as opposed to packets per second in order to perform a link’s capacity simulation in bps.
ISSN:0120-6230
2422-2844