ietf
[Top] [All Lists]

Re: IETF Wireless LAN history

2000-06-29 13:20:03


"Theodore Y. Ts'o" wrote:

   Date: Thu, 29 Jun 2000 13:22:32 -0400
   From: RJ Atkinson <rja(_at_)inet(_dot_)org>

    Actually, IETF has made IEEE 802.11-DSSS the convention for wireless
   LANs at all IETF meetings for some time now.  This has been supported
   at least at Oslo, DC, Adelaide, (and will be at Pittsburgh).  It probably
   has been  supported for some significant time before Oslo, though I
   don't have first-hand remembrance of that.  In DC, Nortel was pushing
   their 802.11-FH system, but a full 802.11-DS system was up and running
   in parallel.

I'm not sure that it's been officially decided as the convention, but
it's certainly been the practice most of the time (although the Nortel
use of frequency hoppers was a certainly an exception to this rule).

I think there was one other time that frequency hoppers were used in the
past; I seem to recall dispairing that we would *ever* settle on a
standard 802.11 type when I broke down and purchased a 802.11-FH PCMCIA
card.  That way, I'd be ready no matter what a particular IETF meeting
site decided to use.

It would be good if we could standardize on one particular type, but as
I recall the last time the issue was raised, it was explained to me that
different systems worked better in different RF environments, so we
should get used to switching back and forth between DS and FH cards.
If we we could depend on it always being 802.11-DS, that might make a
number of people's lives easier.

DS appears to be better for large, flat spaces (largely 2-dimensional,
under 3 stories tall, since transcievers on the middle floor largely
cover the upper and lower).

FH is better for more spherical spaces (largely 3-dimensional).

And DS and FH do not play well together, i.e., it's much better to stay
away from concurrent overlapping installations. I had earlier measured a
BW penalty of between 1/2 to 3/4 (transferring data over only one of the
two technologies at a time, in a concurrent deployment).

The decision of what to use may depend on the local environment.

Joe



<Prev in Thread] Current Thread [Next in Thread>