In order to solve the practical problem in the fine time-scale traffic measurement of large-scale networks

such as Internet

the paper introduced a non-stationary traffic train (NTT) model

which aims to directly model the coarse time-scale traffic samples.NTT model is the superposition of a large number of active traffic train segments

each of which can be described with a fractional Gaussian noise process

while train arrivals form a non-homogeneous Poisson process and train durations are exponential.The application of NTT model in the actual measurement and analysis of CERNET backbone traffic verifies the validation and effectiveness of NTT model

which confirms that NTT model can be used in the study of network behavior based on various time-scale traffic measurements.