A quality of service assessment technique for large-scale management of multimedia flows

Abstract

This paper presents the concept and preliminary experiments of a system for assessing on the Quality of Service of multimedia flows. The goal is to devise a mechanism that allows a service provider to take action whenever poor quality of service is detected in the delivery of multimedia flows. Such procedure is fully automatic since it is based on a goodness-of-fit test between source and destination packet interarrival histograms. If the null hypothesis of the test is accepted the flow is marked as in good standing, otherwise it is marked as anomalous and the network management system should take action in response. The proposed technique is analyzed in terms of hardware complexity and bandwidth consumption. The results show this technique is feasible and easily deployable at a minimum hardware and bandwidth expense.