Study of fuzzy logic-based controller for diff-serv bandwidth broking

Combining transmission of both voice and data on the same network infrastructure was a challenging task which required an advanced network architecture which needs to be scalable and simple. This resulted in a new approach for Ethernet systems where Differentiated Service (Diff-Serv) was introduced to respond to this need and in addition, improve Quality of service (QoS) as a way of providing class of services for different end users. Therefore, optimizing the use of the network available bandwidth is a primgoal of this work which is focused on investigating the impact of implementing a fuzzy logic controller of the crucial factors affecting the performance of the network. These external or internal factors will be controlled by the developed Fuzzy Logic system that will act as bandwidth broker (BB) to provide each user with his optimal Code Point (CP). Within the scope of this work, the CP will not only involve packet loss rate as an external factor to check the network congestion, but in addition, it will consider the internal factors; which are a combination of both service-level agreement (SLA) and the type of user application used, as external factors as well. The Code Point will be marked in the transmitted packets, where it will be possible to the router to check it, and will treat it as per the agreement between the user and the administrator.