A doubly stochastic Poisson model for self-similar traffic

This paper presents a data traffic model capable of describing the long-range dependence and self-similar burstiness structure found in measurement studies of packet data traffic. The model introduced is based on doubly stochastic Poisson processes. The intensity of arrivals is modeled as a continuous stochastic process. This process satisfies most of the properties found in the measurement studies, namely long-range dependence and self-similarity. The generality and simplicity of this model makes it attractive in data traffic modeling.