1 / 8

Self-Similar Wide Area Network Traffic

Self-Similar Wide Area Network Traffic. Carey Williamson. University of Calgary. Introduction. A recent paper has established the presence of network traffic self-similarity in wide area Internet traffic as well “Wide Area Traffic: The Failure of Poisson Modeling”

Download Presentation

Self-Similar Wide Area Network Traffic

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Self-Similar Wide Area Network Traffic Carey Williamson University of Calgary

  2. Introduction • A recent paper has established the presence of network traffic self-similarity in wide area Internet traffic as well • “Wide Area Traffic: The Failure of Poisson Modeling” • Authors: Vern Paxson and Sally Floyd • ACM SIGCOMM’94 • Extended version available by ftp

  3. Introduction (Cont’d) • Original intent: show that self-similarity is not present in WAN traffic • Failed! • Self-similarity IS present in WAN traffic • Identified where it appears and where it does not • Identifies limitations of Poisson models

  4. Main Contributions • Identified presence of self-similarity property in Internet traffic • Defined methodology for testing for the presence of self-similarity, and for testing the goodness of Poisson models • Identified importance of “heavy tails” • Proposed explanations/models for SS • Proposed complete model for telnet

  5. Measurement Study • Detailed measurement study of very lengthy Internet packet traces, with high resolution timer, and lots of storage space • Traces range from 1 hour to 30 days in duration • Millions of TCP packets and connections • Several different sites

  6. Data Analysis • Detailed statistical analysis: • connection interarrivals, per application analysis, packet level, connection level, tests for Poisson-ness, models, evaluation, ... • Very rigourous: confidence intervals, sophisticated statistical tests, sound methodology, ... • A wonderful paper to read

  7. Main Results • Connection arrivals for telnet appear to be Poisson, but... • Packet arrivals are definitely not Poisson • Connection arrivals for ftp and other applications do not appear to be Poisson • Traffic exhibits long range dependence and other aspects of self-similarity

  8. Conclusions • Self-similarity is present in aggregate WAN Internet traffic • Poisson models (or Markovian models of any sort) do not capture reality at all (except possibly for telnet connection arrivals) • Important to consider self-similar traffic

More Related