Modeling networks and active queues management with stochastic bounds and histograms
Abstract
We present an extension of a methodology we have introduced recently. The main property we used or proved is the monotonicity of many networking elements. Furthermore, we use as input traces the measurements performed on real networks. Assuming the stationarity of flows, we obtain histograms (distributions) for the arrivals. Unfortunately, these distributions have a large number of values and the numerical analysis is extremely time-consuming. Using the stochastic bounds and the monotonicity of the networking elements, we show how we can obtain, in a very efficient manner, guarantees on performance measures. Here, we present two extensions: the merge element which combine several flows into one, and some Active Queue Management (AQM) mechanisms. The first extension allows to study networks with a feed-forward topology, while the second extension lets us to consider AQM in an unified manner