These algorithms are also sensitive to loss and/or ECN. The various AQM algorithms on the middleboxes will send these signals, and the TCP stacks will respond appropriately. bvs -----Original Message----- From: rtp-congestion-bounces@alvestrand.no [mailto:rtp-congestion-bounces@alvestrand.no] On Behalf Of Jim Gettys Sent: Friday, March 23, 2012 11:58 AM To: Harald Alvestrand Cc: rtp-congestion@alvestrand.no Subject: Re: [R-C] Slides for the ICCRG session There is a problem with delay sensitive congestion control algorithms: if you have an effective AQM in place, the delays vanish, and they cease to be effective... So the premise that such an algorithm (if it is only sensitive to delay) will keep itself out of the way of other traffic seems flawed to me.... Jim On Mar 23, 2012, at 2:50 PM, Harald Alvestrand <harald@alvestrand.no> wrote:

My slides for the ICCRG session, all in one PDF - as of Friday night. Will send them to the ICCRG chairs by Sunday morning at the latest.

Comments welcome; those who wish to comment on the Google Docs document can go to

https://docs.google.com/a/google.com/presentation/d/1OfBO_UyxdFaey8R_cwd 5Vo6OIZjSy1YwxwtZyMLcM84/edit

and request access.

Harald

<IETF83ICCRGRTPCongestion.pdf> _______________________________________________ Rtp-congestion mailing list Rtp-congestion@alvestrand.no http://www.alvestrand.no/mailman/listinfo/rtp-congestion

_______________________________________________ Rtp-congestion mailing list Rtp-congestion@alvestrand.no http://www.alvestrand.no/mailman/listinfo/rtp-congestion