[e2e] feedback mechanisms and network modelling.

[=?gb2312?q?Jing=20Shen?=]

To me , TCP probe may be used to estimate available bandwidth and use that value for parameter setting. The difference may exist with affecting real traffic or not.
 
regards

alok <alok.dube at apara.com> wrote:
I dont not wish to "identify" the 1st congested point.
 
All I am saying is that in all possibilty, if one was to implement TCP probes and try to detect congestion, and then use this as a feedback to the source, the time taken for this feedback would be the same as time taken in a network where each router could signal back the source in case of congestion.
 
So how does TCP probe help?
 
so ECN/RED and TCP probes would give the same result?
----- Original Message ----- 
From: Jing Shen 
To: alok 
Sent: Wednesday, May 14, 2003 9:09 AM
Subject: Re: [e2e] feedback mechanisms and network modelling.


Well, I come to see what you are trying to solve. 
 
if data plane path and signaling plane path are different, feed back control equation could not be set up by overlooking delay and feedback channel properities. (maybe you have to refer to books on feed-back control).
 
Specifically,  e2e feedback does not guarantee the identification of the 1st congestion point: ECN just says" there is some place congested", RED says " packet is dropped somewhere". 
 


Jing Shen

State Key Lab of CAD&CG
ZheJiang University(YuQuan)
HangZhou, ZheJiang Province 310027
P.R.China


---------------------------------
Do You Yahoo!?
"Ïà¼û²»ÈçÁÄÌì!²»³öÃÅÒ»ÑùÃæ¶ÔÃ棡ÍøÂçÉãÏñÍ·¶Ô¶ÔÅÉËÍÖÐ~¸Ï¿ìÓÃÄãµÄÑÅ»¢µçÓÊÕʺŲÎÓë°É¡­¡­
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.postel.org/pipermail/end2end-interest/attachments/20030514/cb02cab1/attachment.html