Heterogeneous network traffic possesses diverse statistical properties whic
h include complex temporal correlation and non-Gaussian distributions. A ch
allenge to modeling heterogeneous traffic is to develop a traffic model whi
ch can accurately characterize these statistical properties, which is compu
tationally efficient and which is feasible for analysis. This work develops
wavelet traffic models for tackling these issues. In specific, we model th
e wavelet coefficients rather than the original traffic. Our approach is mo
tivated by a discovery that although heterogeneous network traffic has the
complicated short- and long-range temporal dependence, the corresponding wa
velet coefficients are all "short-range" dependent. Therefore, a simple wav
elet model may be able to accurately characterize complex network traffic.
We first investigate what short-range dependence is important among wavelet
coefficients. We then develop the simplest wavelet model, i.e., the indepe
ndent wavelet model for Gaussian traffic. We define and evaluate the (avera
ge) autocorrelation function and the buffer loss probability of the indepen
dent wavelet model for Fractional Gaussian Noise (FGN) traffic. This assess
es the performance of the independent wavelet model, and the use of which f
or analysis. We also develop (low-order) Markov wavelet models to capture a
dditional dependence among wavelet coefficients. We show that an independen
t wavelet model is sufficiently accurate, and a Markov wavelet model only i
mproves the performance marginally. We further extend the wavelet models to
non-Gaussian traffic through developing a novel time-scale shaping algorit
hm. The algorithm is tested using real network traffic and shown to outperf
orm FARIMA in both efficiency and accuracy. Specifically, the wavelet model
s are parsimonious, and have the computation complexity O(N) in developing
a model from a training sequence of length N. and O (AI) in generating a sy
nthetic traffic trace of length M.