Simulation Based Comparative Performance Analysis of Qos Traffic Scheduling Using Fair and Delay Adaptive Scheduler (fdas) Versus Wfq and Edf

Abdelwahab M. Elnaka,Qusay H. Mahmoud,Xining Li
DOI: https://doi.org/10.1109/ccnc.2016.7444910
2016-01-01
Abstract:Several objectives are aimed when scheduling traffic for transmission on heterogeneous networks. Traffic flows might belong and required to satisfy different set of requirements for a diverse set of applications, end users and networks. Schedulers have several goals to achieve in order to meet the needs of all stakeholders involved. Some schedulers are built to ensure fairness in bandwidth allocation, others try to prioritize certain flows and increase their share in the network resources, typically available bandwidth, and last but not least some are built to meet the stringent and time critical delay constraint of certain flows. Weighted fair queuing (WFQ) is a very well known scheduling algorithm that fairly allocates bandwidth between a set of flows contending for available output bandwidth while taking their weights into its scheduling calculation. Earliest deadline first (EDF), on the other hand, is a scheduling mechanism that favors flow packets that are about to expire and gives them preferential treatment and priority to be transmitted first. Neither WFQ nor EDF achieves both fairness and deadline goals simultaneously. For this reason, we introduced our fair and delay adaptive scheduler (FDAS) to bridge the gap and balance between fairness and delay requirements of incoming traffic flows. In this paper, we present a detailed simulation comparative study to manifest the difference between our proposed scheduler and the two popular schedulers and how it outperforms both to relatively achieve both bandwidth allocation fairness and traffic delay bound objectives simultaneously.
What problem does this paper attempt to address?