A new adaptive sliding-based congestion control strategy for a large scale TCP/IP network with differentiated services

In this paper, a new congestion control strategy is developed for a large scale TCP/IP network with differentiated services assuming arbitrary network configuration. A bandwidth allocation controller is designed according to the Quality of Service (QoS) requirements for the three classes of traffic, and traffic regulation algorithm is introduced for congestion avoidance. Considering the uncertainties of the large scale fluid flow model, the controller is designed using an adaptive sliding mode technique, and linear matrix inequality (LMI) conditions are obtained to guarantee BIBO stability of the overall error dynamics of the network. To regulate the incoming traffic in the ordinary and best effort services, a new algorithm is presented based on the packet marking method. Finally the performance of the control strategy is verified by the simulation results.