Evaluating the Impact of Management Traffic and Self-Similarity on Wireless Network Performance

This article makes a taxonomic study of wireless traffic using a model of self-similarity based on traffic components at the link, network and transport layer. We evaluate applications performance using two different simulation scenarios in which the traffic is defined according to the statistical parameters observed in real traffic. The simulations evaluate network performance using different compositions of wireless management traffic and several degrees of self-similarity. The results show a better performance for delay and loss when certain management traffic components are removed, even for the cases when the traffic is self-similar. The results may help to identify unwanted management traffic components for certain network scenarios.