1 / 16

By Manish Jain and Constantinos Dovrolis 2003

By Manish Jain and Constantinos Dovrolis 2003. End-to-End Available Bandwidth: Measurement Methodology, Dynamics, and Relation With TCP Throughput. Presented by Caroline Williams. Purpose.

Download Presentation

By Manish Jain and Constantinos Dovrolis 2003

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. By Manish Jain and Constantinos Dovrolis 2003 End-to-End Available Bandwidth: Measurement Methodology, Dynamics, and Relation With TCP Throughput Presented by Caroline Williams Rosa Williams

  2. Purpose The authors are not satisfied with the current definition of available bandwidth nor the tools to measure available bandwidth. The authors propose: • A concise available bandwidth definition • A methodology to measure available bandwidth • A tool that implements the methodology Rosa Williams

  3. Motivation • Available bandwidth is an important metric for: • Congestion control • Streaming applications • Quality-of-service verification • Server selection • Overlay networks • As such, the definition should be agreed upon, the measurements accurate and nonintrusive. Rosa Williams

  4. DefinitionEnd-to-end Avail-bw Rosa Williams

  5. MethodologySelf-Loading Periodic Streams (SLoPS) • A stream consists of K packets of size L, sent at constant rate R • One-way delays (OWD) of successive packets at RCV show an increasing trend when R > A • A is converged upon through an iterative algorithm at RCV. RCV notifies SND of new R. The “algorithm will converge to a range [Rmin, Rmax] that includes A.” Rosa Williams

  6. ImplementationPathload • Process SND generates fleets of timestamped packet streams for R • Process RCV determines the OWD trend for the fleet. Then, adjusts Rmin or Rmax according to the SLoPS algorithm. A new R (halfway between Rmin and Rmax) is fed back to SND. • Continue theabove two steps until Rmax – Rmin£a user defined resolution • [Rmin, Rmax] can be calculated in less than 15 seconds using default parameters Rosa Williams

  7. Verification (NS Simulation) Rosa Williams

  8. Verification (Experimental) Rosa Williams

  9. Dynamics: Variability and Load Conditions Rosa Williams

  10. DynamicsVariability and Statistical Multiplexing Rosa Williams

  11. Dynamics: Effect of the Stream Length Rosa Williams

  12. DynamicsEffect of the Fleet Length Rosa Williams

  13. TCP and Avail-BW Rosa Williams

  14. IntrusivenessPathload Rosa Williams

  15. Conclusion • Available bandwidth is elusive • Jain and Dovrolis have provided a methodology that reports a range of rates that includes avail-bw • Their tool is nonintrusive and reliable in a “wide range of load conditions and path configurations”. Rosa Williams

  16. References • Information Sciences Institute. ns-2. http://nsnam.isi.edu/nsnam/index.php/Main_Page. October 21, 2007. • M. Jain, C. Dovrolis. End-to-end available bandwidth: measurement methodology, dynamics, and relation with TCP throughput. IEEE/ACM Trans. Netw. 11(4): 537-549 (2003) • T. Oetiker. MRTG - The Multi Router Traffic Grapher. http://oss.oetiker.ch/mrtg/. October 21, 2007. Rosa Williams

More Related