Speed of 802.11n network with pre-N and N clients?

How well does an 'n" network perform with two or more clients when one
of then has a "b" or a "b" interface?

--
a d y k e s @ p a n i x . c o m
Don't blame me. I voted for Gore. A Proud signature since 2001

(E-Mail Removed) (Al Dykes) hath wroth:

>How well does an 'n" network perform with two or more clients when one
>of then has a "b" or a "b" interface?

Which "N" are you talkiN about? Draft-N? Airgo Real-MIMO? Ruckus
Wireless beam forming? They all respond differently.

Draft-N has 3 modes:
Legacy 802.11a/b/g only
Mixed Mode 802.11n and 802.11a/b/g
Green Field 802.11n only
There are also some mode variations that depend on 20/40MHz channel
bandwidth that are being debated. See Page 3 of the article below.
Backwards compatibility of Pre-N/Draft-N/Whatever-N. (3 Pages)
<http://www.wirelessnetdesignline.com/howto/199700564;jsessionid=45L0315B2GUNQQSNDLRSKH0CJUNN2 JVN>

In mixed mode, just one 802.11a/b/g connection will drastically slow
the Draft-N connected devices. The problem is that the various
articles do not benchmark the effect.

Performance will also vary by manufacturers implementation. Measuring
mixed modes is also subject to quite a bit of creativity and
variations in procedures. For example, do the 1 or 2 connected
802.11a/b/g clients move any traffic during the tests? Also, most of
the multi-channel MIMO receivers can listen for 802.11a/g (not b)
receive on one antenna/receiver, while listening for Draft-N on the
others antennas/receivers.

There's also the numerous combinations of modes. For example, the
Pre-N access point is going to have quite a different performance in
Mixed Mode, with an 802.11a/b/g connection, than in Green Field Mode,
where the 802.11a/b/g signal constitutes a source of interference and
does not connect to the access point.

I'll make it easy on myself and abstain from predicting performance.

Benchmark of Belkin Draft-N access point:
<http://www.smallnetbuilder.com/content/view/28732/96/>
Mixed mode seems to be about 20Mbits/sec for 802.11g and 43Mbits/sec
for Draft-N only. See:
<http://www.smallnetbuilder.com/content/view/28732/96/1/9/>
However thruput is also seriously affected by encryption mode
(probably a bug).

A later review of the Buffalo Nfiniti Dual-Band Somewhat Draft-N
Router:
<http://www.smallnetbuilder.com/content/view/29885/96/>
They use a few tricks like bonding 5.7 and 2.4GHz channels to double
the thruput. Why trash just the 2.4GHz band with 40MHz wide noise,
when you can also take out the 5.7GHz band? Sigh. Anyway, about
117Mbits/sec running flat out with every feature enabled and with no
legacy devices around. Unfortunately, they didn't test Mixed mode.

Whether you can buy one is another question:
"Update: CSIRO shuts down Buffalo WLAN shipments"
<http://www.smallnetbuilder.com/content/view/30070/97/>
<http://www.engadget.com/2006/11/16/csiro-wins-landmark-wlan-lawsuit-against-buffalo-more-to-come/>
<http://www.csiro.au/news/WLANInjunction.html>
Australia's national research consortium holds a patent they claim
Buffalo has infringed upon. So, they sue everyone in sight.
<http://www.google.com/patents?vid=USPAT5487069>
Kinda looks like the Australians claim they own most of 802.11a/g. So,
all I have to do is watch what the standards committees are doing,
patent it before they nail down the standard, and wait 10 years for
market value to be large enough to make it worth suing. What a great
business model for government sponsored organization.

--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558