Study of Fuzzy Logic-based Controller for Diff-Serv Bandwidth Broking

Abstract

Combining both voice and data on the same network infrastructure results in need for an advanced network which is to be simple and scalable.  This resulted in new approach for Ethernet where Differentiated Service (Diff-Serv) is introduced. This is taken into consideration as well as QoS as a way of providing class of service to end users. Therefore, optimizing available bandwidth efficacy is one of the goals of this work which is centered on investigating the impact of crucial factors on performance of implementing a fuzzy logic controller. These factors can be external or internal and will be manipulated by Fuzzy Logic controller that will work as bandwidth broker to give each user his optimal Code Point (CP). In this work, the CP will not only consider packet loss rate as external factor  to check the congestion, but it will also consider the internal factors which are a combination of both service-level agreement (SLA) and the type of application being used. This CP will be marked in the transmitted packets, and then the router will check that and will treat it as agreed between user and administrator.