A fuzzy congestion control algorithm for Differentiated Services

Considering the shortages of traditional RED algorithm, that are difficulty of setting parameters and lack of fairness, and the unique advantages of fuzzy control in dealing with network congestion issues, a new fuzzy congestion control algorithm for Differentiated Services named DSFCC is proposed. The algorithm adequately considers the user´s requirements of Quality of Service (QoS) of different traffic flows, and adopts the methodology of combining Sharing Index and queue length to deal with congestion problems. It uses the fuzzy relational matrix to describe the real-time status of the various types of queues and traffic flows, and intelligently controls the network congestion based the fuzzy rule. Theoretical analysis and NS simulation results show that the proposed algorithm has smoother delay characteristic, better adaptability and fairness in the case of QoS implementation. It really improves the system´s ability to deal with network congestion.