John,
I remember bringing up this issue very briefly at the BOF. The effect of
link layer retransmission can have degrading effect on TCP performance. In
the cellular wireless context, we generally have a link layer ARQ mechanism
active over the radio link. The remote TCP sender may timeout while local
retransmission is going on over the radio link. Actually, the timer settings
for link layer retransmission mechanism need to be done very carefully in
order to avoid conflicting with TCP timer. Also, TCP's flow control
mechanism can be adversely affected by any flow control going on locally
over the radio link. The main point is that TCP and the wireless link layer
may have several duplicate functions, and its important for a
new-link-layer-ARQ designer or even a TCP-over-link-layer implementor to
make sure that these are not conflicting.
Thanks,
Sanjoy Sen
Advanced Wireless Network Technology
Nortel Networks
972-685-8274 / Fax : 972-684-3744 / sanjoy@nortelnetworks.com
<mailto:sanjoy@nortelnetworks.com>
-----Original Message-----
From: John Border [SMTP:border@hns.com]
Sent: Tuesday, January 19, 1999 5:13 PM
To: mallman@lerc.nasa.gov; pilc@lerc.nasa.gov
Cc: Vern Paxson; Aaron Falk; Dawkins, Spencer
[RICH1:2011-M:EXCH]
Subject: Re: PILC: prioritization
Was there any discussion about the effects of links which do
link layer
retransmissions on high layer protocols? These could be looked at
as links
with varying delay where the variation can be relatively large.
John Border
Hughes Network Systems
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 09:12:19 EST