Ali Broumandan - University of Tehran
Nasser Yazdani

Proportional Delay Differentiation (PDD) model is an important approach for achieve QoS in Internet. It aims to maintain prespecified packet queueing-delay ratios between different classes of traffic at each hop. Our work has focused on output queue delay proportion fairly among traffic classes of different priorities in Proportional Differentiated Services architecture to improve its accuracy. In this paper, we propose a novel scheduling method based on Hybrid Proportional Delay algorithm to adjust output queue delay proportion fairly among traffic classes to meet the desired delay ratios. HPD is combination of WTP and PAD algorithms. In this work Genetic Algorithm (GA) is used to obtain set of variable control parameters in WTP algorithm and we revised priority criteria in PAD scheduler. Simulation results show that, in comparison with HPD, OHPD can provide more accurate results in short and long time-scale.

Quality of Service, proportional DiffServ, Genetic Algorithm,Packet scheduling