An Adaptive Scheduling Algorithm for PDD Method Based on Modelling of DiffServ Queues

Abstract

It is important to represent a mechanism to provide QoS in the Internet for applications that are sensitive to delay and loss. Our work has focused on proportional delay differentiation (PDD) architecture to improve its accuracy. The main contribution of this paper is an adaptive scheduling algorithm base on modelling of the DiffServ (differentiated services) queues. Using control theory has recently gained much recognition to achieve QoS guarantee. In this paper a linearized analytical model for DiffServ queues is proposed. This model can be used for any model based controller. The controller tries to guarantee QoS parameters in PDD architecture. Adaptive control based modelling brings out portable QoS because it automatically tunes model parameters in any platform simulation results show that the adaptive model can automatically update the model parameters and track the variation of queues in each event