Modeling and evaluation of congestion control for different classes of network traffic

Abstract

SUMMARYTraffic congestion has become a crucial problem in wireless networks due to its detrimental effects on the end-to-end Quality-of-Service (QoS) of mobile users. This issue has posed various challenges to developing effective congestion control mechanisms and devising new performance evaluation tools. Various congestion control schemes have been proposed to support differentiated QoS. Among these schemes, Active Queue Management (AQM) has been recognized as an effective mechanism to manage downlink buffers in wireless networks. Most existing studies on AQM are based on the use of averaged queue length and the related performance analysis is carried out via software simulation. This paper contributes to the development of an analytical performance model for a finite capacity queuing system subject to AQM and multiple traffic classes based on the instantaneous queue length. Expressions for the aggregated and marginal performance metrics including the mean queuing length, packet loss probability, mean waiting time, system throughput, response time, and utilization are analytically derived. The validity of this new model is demonstrated by comparing the analytical results against those obtained from simulation experiments. The analytical model is then used to investigate the effects of varying thresholds on the aggregated and marginal performance metrics under different combinations of arriving rates for multiple traffic classes.

Special Issue: Performance Analysis and Enhancements of Wireless Networks

“…In the third paper [3], Wang et al develop an analytical performance model for traffic congestion control in wireless networks. Authors analyse a finite capacity queuing system under an active queue management (AQM) mechanism based on the instantaneous number of packets from multiple traffic classes in the queue.…”