A flow-based model for Internet backbone traffic

We model traffic on an uncongested backbone link of an IP network using Poisson Shot-noise process and M/G/$\infty$ queue. We validate the model by simulation. We analyze the model accuracy with real traffic traces collected on the Sprint IP backbone network. We show that despite its simplicity, our model provides a good approximation of the real traffic observed on OC-3 links. This model is also very easy to use and requires few simple parameters to be input.