[e2e] Is a control theoretic approach sound?

[Ibrahim Matta]

Hi Saad, the problem is NOT with control theory, but sometimes its
application!

If I understand your concern correctly, yes, control theory can show you
that end-to-end flow control with per-flow queueing is more stable than
E2E flow control with aggregate queueing. 

In addition to Keshav's work you point to, see Scott Shenker's SIGCOMM
1990 paper titled "A Theoretical Analysis of Feedback Flow Control".

Also, you can find a simple control-theoretic model and its
(straightforward Z-transform) analysis at:

Ibrahim Matta and Liang Guo. Differentiated Predictive Fair Service for
TCP Flows. In Proceedings of ICNP'2000: The 8th IEEE International
Conference on Network Protocols, Osaka, Japan, October 2000. 
http://www.cs.bu.edu/faculty/matta/Papers/icnp00.ps

The Technical Report has the full analysis:
Technical Report BU-CS-2000-012, Boston University, Computer Science
Department, Boston, MA 02215, May 2000.
http://www.cs.bu.edu/techreports/ps/2000-012-diffserv-tcp.ps


Best regards, Ibrahim


--
Ibrahim Matta, Associate Professor 
Dept. of Computer Science, 111 Cummington Street
Boston University, Boston, MA 02215, USA
Email: matta at cs.bu.edu, Tel: (617) 358-1062, Fax: (617) 353-6457
http://www.cs.bu.edu/fac/matta



> -----Original Message-----
> From: Saad Biaz [mailto:biazsaa at eng.auburn.edu]
> Sent: Monday, July 28, 2003 3:47 PM
> To: end2end-interest at postel.org
> Subject: [e2e] Is a control theoretic approach sound?
> 
> Dear all:
> 
> I wonder if the control theoretic approach to congestion
> control is not fundamentally flawed for best-effort networks (no flow
> based fair queueing).
>
[text snipped]
> Recall that Keshav's control theoretical approach is based on the fact
> that there is a round robin packet scheduling at the routers. In this
> case, the actions of one sender have consequences directly targetted
to
> this sender.
> 
> Any feedback will be most appreciated. Is there something I am
missing?