Multiscale Predictability of Network Traffic

Distributed applications use predictions of network traffic to sustain their performance by adapting their behavior. The timescale of interest is application-dependent and thus it is natural to ask how predictability depends on the resolution, or degree of smoothing, of the network traffic signal. To help answer this question we empirically study the onestep-ahead predictability, measured by the ratio of mean squared error to signal variance, of network traffic at different resolutions. A one-step-ahead prediction at a low resolution is a prediction of the average behavior over a long interval. We apply a wide range of linear time series models to a large number of packet traces, generating different resolution views of the traces through two methods: the simple binning approach used by several extant network measurement tools, and by wavelet-based approximations. The wavelet-based approach is a natural way to provide multiscale prediction to applications. We find that predictability seems to be highly situational in practice---it varies widely from trace to trace. Unexpectedly, predictability does not always increase as the signal is smoothed. Half of the time there is a “sweet spot” at which the ratio is minimized and predictability is clearly best. We conclude by describing plans for an online waveletbased prediction system. Effort sponsored by the National Science Foundation under Grants ANI-0093221, ACI0112891, and EIA-0130869. The NLANR PMA traces are provided to the community by the National Laboratory for Applied Network Research under NSF Cooperative Agreement ANI-9807479. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation (NSF).}