Modeling internet backbone traffic at the flow level

Abstract

Abstract-Our goal is to design a traffic model for noncongested Internet backbone links, which is simple enough to be used in network operation, while being as general as possible. The proposed solution is to model the traffic at the flow level by a Poisson shotnoise process. In our model, a flow is a generic notion that must be able to capture the characteristics of any kind of data stream. We analyze the accuracy of the model with real traffic traces collected on the Sprint Internet protocol (IP) backbone network. Despite its simplicity, our model provides a good approximation of the real traffic observed in the backbone and of its variation. Finally, we discuss the application of our model to network design and dimensioning.

The multi-source model for dimensioning data networks

“…Internet traffic is most often modeled at flow level 2 , assuming some ideal bandwidth sharing between ongoing flows [15,3,2,1,19,27,17]. Modeling traffic at packet level proves too complex and is hardly effective, given that users typically perceive quality of service at flow level [12].…”

Towards Bandwidth Estimation Using Flow-Level Measurements

“…Most notably, the authors of [10] estimate statistics of the traffic rate from flow characteristics. The underlying model requires that the network links are not congested and an adaptation to the considered traffic is needed.…”

Traffic characterization for traffic engineering purposes: analysis of Funet data

“…In a recent study, Gunnar et al [3] confirm the validity of the mean-variance relationship (1) and give values c = 1.5 and 1.6 based on data traces from a global operator's backbone with ∆ = 300 s.…”

“…During daytime 10-20 % of TCP traffic is 1 For details about Finnish university network (Funet), see http://www.csc.fi/suomi/funet/verkko.html.en HTTP. There exists also considerable amount of peer-to-peer traffic.…”

“…Single flows coming from Mbps-size access networks do not find the backbone link as a bottleneck. According to recent studies [1], flow arrivals in such a link constitute a Poisson process, which in a longer time scale is surely time-varying but in a shorter time scale can be considered as time-homogeneous. On the other hand, it is widely accepted that the flow sizes follow a heavy-tailed distribution, such as Pareto with shape parameter α ∈ (1, 2).…”