An Analytical Model for Priority-Based AQM in the Presence of Heterogeneous Network Traffic

Active queue management (AQM) coupled with differentiated scheduling mechanism is an important and promising scheme for congestion control and quality-of-service (QoS) guarantee in communication networks. This paper develops a new analytical performance model for priority-based AQM subject to heterogeneous bursty traffic modelled by Markov-modulated Poisson processes (MMPP) and non-bursty traffic modelled by Poisson processes. Individual thresholds for each traffic class and pre-emptive resume (PR) priority scheduling mechanism are adopted to control traffic injection rate and support differential QoS. The model derives the expression of the aggregate and marginal performance metrics of the priority-based AQM system. The credibility of the model is demonstrated by comparing the analytic results with those obtained through extensive simulation experiments. The proposed model is used to investigate the impact of the squared coefficient of variation (SCV) of bursty traffic on the performance of AQM.