TRAFFIC MODELING AND DESIGN METHODOLOGIES FOR BROAD-BAND NETWORKS

Designers of modern high-speed complex networks, such as frame relay ( FR) and ATM, face three principal areas of difficulty: the need to ens ure accurate and efficient traffic modelling; the large number of comb inations of control parameter settings that is available; and the rari ty of some events which contribute to Quality of Service measures, res ulting in a high cost of simulation. This paper shows how designers ca n evaluate and design such networks without making oversimplified assu mptions. We review some of the relevant techniques and show how they c an be used separately or in combination. We use the transform-expand-s ample (TES) modelling technique to synthesize traffic based on measure d autocorrelation and marginal probability distribution of the real tr affic traces. For longer time-dependent models we use self-similar met hods. For evaluating extremely low overflow probabilities for queues d riven by bursty traffic, we show how we have applied importance sampli ng techniques to reduce simulation run times by several orders of magn itude. To cope with the large numbers of parameters, we use factorial experimental design to identify the most significant parameters and th en to find near-optimal values for them by means of mean field anneali ng. Throughout the paper we give examples based on variable-bit-rate v ideo traffic (MPEG-2 and H.261) in FR and ATM networks.