Re: Pointer to ATM performance work?

[salo () msc edu (Tim Salo)]

> Date: Tue, 9 May 95 11:43:58 EDT
> From: Richard Colella  <colella () snad ncsl nist gov>
> To: nanog () merit edu
> Subject: Pointer to ATM performance work?
> Cc: colella () nist gov
>
> I seem to remember discussion at one of the meetings of some ATM
> performance work that looked at the benefits of discarding whole TCP
> packets when a cell was dropped.  Can someone give me a pointer to
> this?

I recommend the following, for a start:

A. Romanow and R. Oskouy, "A performance Enhancement for Packetized ABR
and VBR+ Data," ATM Forum/94-0295, March, 1994.

        [A short proposal that ATM switches implement EPD]

A. Romanow, "TCP over ATM: Some Performance Results," ATM Forum/93-784,
July, 1993, available via anonymous FTP from: 
playground.sun.com:/pub/tcp_atm/tcp_forum.7_93.ps

        [Primarily background.  Says: "We also plan to explore RED..."]

A. Romanow, Packet Discard Strategies for Controlling ATM 
Congestion, ATM Forum/94-0107, January, 1994.

        ["... Viewgraphs will be distributed at the meeting."]

A. Romanow and S. Floyd, "The Dynamics of TCP Traffic over ATM
Networks," available via anonymous FTP from:
playground.sun.com:/pub/tcp_atm/tcpatm.jsac.ps
see also .../tcpatm_extended.00.ps

        [Explores Partial Packet Discard and Early Packet Discard.]

S. Floyd and V. Jacobson, "Randor Early Detection Gateways for 
Congestion Avoidance," IEEE/ATM Trans. on Networking, vol.1 no.4,
pp. 29-41, Sept. 1993.

        [I think this is the original RED paper.]

You might also look at:

S. Floyd, "TCP and Explicit Congestion Notification," available via
anonymous FTP from:
ftp.ee.lbl.gov:papers/tcp_ecn.4.ps.Z

Once you have read these, I have a list of more general papers about
TCP performance over ATM...

-tjs