Wavelet-based models for network traffic

We propose a novel set of wavelet-based stochastic models for self-similar network traffic with non-Gaussian behavior. We show that these models are sufficiently accurate and parsimonious, and have very low computational complexity in analysis and synthesis.