[mat-wg] Bandwidth measurements

  • To: Henk Uijterwaal henk@localhost, mat-wg@localhost, Тимур Султанов tursul@localhost
  • From: Andrei Sukhov amskh@localhost
  • Date: Tue, 13 Jul 2010 23:38:46 +0400
  • Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yandex.ru; s=mail; t=1279049927; bh=PFZU095UGeXVaN+NPcjYr5vD65UbSWS4OoFfKDAO6qs=; h=From:To:Subject:MIME-Version:Message-Id:Date:Content-Type; b=CRque5+GUuOqr3Dlcjkv3qTxhSjtB8Tb4nX8lK6LbSpAUM64CQVZDBfitGckO7lTz xVbdbFDHidCB2ixxCjvq+NSGWq3/7ZR7pSFD0Qe+yKlg2nxtSJcbMjK82QnnykjHK1 YfBpUCECM44LNp3MWaSXbTrPYc5rbiIdQ+LsBz24=

Hi Henk,

It is again me, Andrei Sukhov.

Earlier our team tries to add mechanism of bandwidth measurements to RIPE Test Box data. We found a method how to calculate an available bandwidth from delays of different sizes of packets. Our report has been submitted in Berlin and Lisbon RIPE meetings.

After Lisbon we have made additional experiments and compare our results with well-known pathload utility. The area of applicability of our method will be also found.

All our results are included in paper that is attached to this letter. I would like to ask how our method may be used in RIPE Test Box for calculation of available bandwidth. Probably we should to make a report on IETF session (in Beijing ?).

Please, give us an advice.

Sincerely,  Andrei Sukhov

Attachment: Paper4IETF.pdf
Description: Adobe PDF document