An Intelligence Approach to Congestion Control in Differentiated Services Networks

Once a network is set, as data traffic increases, there are two approaches to preserve Quality of Service (QoS) in the network. One is increasing total bandwidth witch is not always feasible. The other is Traffic Engineering (TE). Dynamic bandwidth allocation and Active Queue Management (AQM) are among those methods, used commonly for TE in communication networks. In this paper a fluid flow model for M/M/1 queues has been considered in the bottleneck switches of Differentiated-Services networks. In this paper a novel PID controller for a fluid flow model for M/M/1 queues has been designed and fuzzy logic strategy has been used to tune the parameters of PID controller. Simulation results reveal the efficiency of proposed controller.

Keywords:

M/M/1 queue , Congestion Control , Fuzzy PID Controller , Diff-Serv