An admission control approach for multifractal network traffic flows using effective envelopes

Abstract

In this paper, we present a novel admission control scheme that takes into account the multifractal properties of network traffic flows. To this end, we mathematically determine the variance of the aggregated traffic processes as well as their wavelet energy across time scales, considering a multiplicative cascade based multifractal model for the network traffic traces. In addition, we derive an analytical expression for the effective bandwidth estimation of the multifractal modeled traffic process. We use the proposed effective bandwidth equation to determine the effective envelope for multifractal traffic flows. The development on the multifractal traffic effective envelopment allows us to design an admission control scheme capable of efficiently guaranteeing quality of service in terms of effective envelope violation and delay probabilities for multifractal traffic flows. Through simulations, we compare the admissible number of flows in connections under different types of scheduling algorithms as well as different traffic models, such as the monofractal fBm and the proposed multifractal approach.