Stochastic Approach for Modeling Multi-Class Congestion Control Mechanisms Based on RED in TCP/IP Networks

In this paper, a discrete-time stochastic queueing model for the performance analysis of congestion control mechanism based on Random Early Detection (RED) methodology with bursty and correlated traffic by using a two-state Markov-Modulated Bernoulli arrival process (MMBP-2) as the traffic source is presented. A new RED analytical framework is proposed and a two-dimensional discrete-time Markov chain is introduced to model two traffic classes (each dimension corresponding to a traffic class with its own RED parameters) by using this active queue management algorithm which takes into account the reduction of incoming traffic arrival rate due to packets dropped probabilistically with the drop probability increasing linearly with system contents. The analytic evaluation of the queue considered could be of interest for modelling the performance of the RED algorithm with two traffic classes in TCP/IP networks. The mean system occupancy, mean packet delay, probability of packet loss, and system throughput are computed from the analytical model for dropping policy, which are functions of the thresholds and maximum drop probability. The results clearly demonstrate how different threshold settings can provide different tradeoffs between loss probability and delay to suit different service requirements. Also the effects of input parameters on a selected set of performance metrics are studied