Fuzzy clustering of aggregated flows for traffic control in ATM networks

In this paper we consider the problem of modelling the stochastic flows of traffic in ATM networks. The modelling approach is based on fuzzy set partitioning of the space of feasible flows describing the traffic occurring in virtual connections. The goal of this model is to provide the means for describing the smoothing mechanism of the statistical variations of traffic patterns around the nominal prescribed values. Some typical features from a class of Gaussian multistep smoothing procedures are used in this context. Computer simulation results for ATM asymmetrical flows consisting of 300 and 1000 bits/packet entities flowing in capacitated links are included