The paper describes a new class of stochastic processes, which generalizes the concept of self-similarity. This class of processes is suitable for traffic modeling since it permits to take into account the presence of several traffic components with different statistical properties and defines proper conditions for the convergence of the aggregated traffic to suitable limit processes. The main result of the paper is that under adequate normalization conditions the aggregation of heterogeneous flows converges to the sum of two independent processes (α-Stable Levy Motion and Fractional Brownian Motion), able to take into account the self-similar nature of broadband traffic as well is its bursty (i.e., non Gaussian) marginal distribution
Modeling of Network Traffics
PAGANO, MICHELE
2010-01-01
Abstract
The paper describes a new class of stochastic processes, which generalizes the concept of self-similarity. This class of processes is suitable for traffic modeling since it permits to take into account the presence of several traffic components with different statistical properties and defines proper conditions for the convergence of the aggregated traffic to suitable limit processes. The main result of the paper is that under adequate normalization conditions the aggregation of heterogeneous flows converges to the sum of two independent processes (α-Stable Levy Motion and Fractional Brownian Motion), able to take into account the self-similar nature of broadband traffic as well is its bursty (i.e., non Gaussian) marginal distributionI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
