RO EN
On Estimating Long Range Dependence of Network Delay
Authors: Michael S. Borella
Keywords: Traffic model, Internet, time scales, periodogram.
– 0.27 Mb
Keywords: Traffic model, Internet, time scales, periodogram.
Abstract
We analyse 12 traces of round-trip Internet packet delay. We find that these traces, when viewed as time series data, often exhibit Hurst parameter estimates greater than 0.5, indicating long-range dependence. Several traces, however, are not well modelled with a constant . We discuss in detail our analytical methods and the robustness of empirical estimators of under conditions of non-negligible packet loss. We also apply a newly-developed wavelet estimator of and find that (1) it explains why periodogram-based estimators can produce inconclusive results, and (2) it produces results that can be used to estimate short range as well as long range dependence. Our overall results indicate that Internet delay is bursty across multiple time scales, which implies that end-user quality of service in the Internet is likely to be impacted by long periods of very large and/or highly variable delays.Michael S. Borella
CommWorks, A 3Com Company, 3800 Golf Rd.,
Rolling Meadows IL 60008, USA
Tel: (847) 262-3083, Fax: (847) 262-2255
E-mail:
CommWorks, A 3Com Company, 3800 Golf Rd.,
Rolling Meadows IL 60008, USA
Tel: (847) 262-3083, Fax: (847) 262-2255
E-mail:
Fulltext
– 0.27 Mb
Contents
- Foreword
- Queuing analysis with traffic models based on deterministic dynamical systems
- On the Chaotic Behavior of Variable-Bit-Rate Video and Its Application in Traffic Modeling
- A Stochastic Map Model of Phase Transition in Internet Traffic
- Sustained Rhythm and Directed Self-Locomotion under Thermodynamically Open System
- A Pseudo-Statistic Edge Detector using Fuzzy Computation
- Modeling Network Traffic in Wavelet Domain
- On Estimating Long Range Dependence of Network Delay
