multiple WAPs - physical proximity

I was at a friends small office, and noticed their WAPs in the back room.

For some reason, they had acquired a couple of Linksys b/g WAPs
and had "stacked" them on top of each other..... they fit so nice
So the antennas were almost touching each other.

I was wondering, if you were going to run 2 WAPs in very close proximity,
how "should" you physically arrange them ?

My RF knowledgebase only has mental entries for 2m & 450 repeaters.
--
----------------------------------
"If everything seems to be going well,
you have obviously overlooked something." - Steven Wright

On Sun, 13 Apr 2008 16:36:14 GMT, "P.Schuman"
<(E-Mail Removed)> wrote:

>I was at a friends small office, and noticed their WAPs in the back room.
>
>For some reason, they had acquired a couple of Linksys b/g WAPs
>and had "stacked" them on top of each other..... they fit so nice
>So the antennas were almost touching each other.
>
>I was wondering, if you were going to run 2 WAPs in very close proximity,
>how "should" you physically arrange them ?

Bad idea.

802.11g and all spread spectrum transmissions do not stay within a
nice neat spectral mask. In theory, spread spectrum transmissions
spray RF well outside the official occupied bandwidth. For example,
if you have an access point running on Ch 6, you'll have the bulk of
the RF energy between Ch 4 thru Ch 8, but here will still be some RF
all the way out to Ch 1 and Ch 11. The amounts involved are not huge,
but at the very close proximity of the boxes (and presumably the
antennas), self-interference between radios is going to be a problem.

Even if the system uses a mess of external antennas and coax cables,
there will be case leakage problems. The WAP54G and WRT54G boxes (the
only ones that stack neatly) are not internally shielded. They will
radiate through the case to each other quite nicely.

The band pass filter at the antenna is useless because it only
prevents out of band radiation. It merrily passes everything that's
in band.

The effects of such interference is not huge. It mostly shows up as a
high packet loss or a loss in thruput. Having the connection speed
slow down is also an indication of problems. The easiest test is to
use ping:

Start -> run -> cmd <enter>
ping -t ip_address_of_access_point

You should see a *consistent* 2-3msec latency. If it varies all over
the place, and/or shows timeouts, you've got a problem. However,
that's for just one access point. Generate some traffic to one of the
other access points, and see how it affects the ping results. If they
were fine with just the initial setup, but where adding traffic on the
2nd access point cause the latency to climb, then congratulations, you
have self-interference.

>My RF knowledgebase only has mental entries for 2m & 450 repeaters.

I spent yesterday getting IRLP running to our club repeaters. Some
bright person (me) suggest that this would be a great time to move
racks around. My back and wallet both hurt. Tomorrow we do EchoIRLP.
<http://www.LearnByDestroying.com/k6bj/IRLP/>

--
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