[ previous ] [ next ] [ threads ]
 
 From:  Jim Thompson <jim at netgate dot com>
 To:  Douglas Stringer <dougstringer at mac dot com>
 Cc:  Justin Wilson <j2sw at mtin dot net>, m0n0wall at lists dot m0n0 dot ch
 Subject:  Re: [m0n0wall] Wireless Questions
 Date:  Sun, 13 Nov 2005 19:31:51 -1000
Douglas Stringer wrote:

> Yeah.. I actually got several indoor B's back in 2003 and have them  
> at my shop and have deployed several outdoor B's in NYC (through  
> Tanis) ---and had to keep them all on one channel per switch, btw :).
>
> Anyway, what causes the interference if you're using non-overlapping  
> channels on a WRAP with 2 wireless cards with m0n0?... and how is it  
> noticed? Lower data rate?
> doug.

I really think this is off-topic for m0n0wall (just as its been 
off-topic several times on the Soekris list.)

Essentially, the receiver in the radio(s) "sees" some signal from 
off-channel.   This is seen in one of two ways:

A is receiving a packet (from C) and B decides to transmit.
1) If the signal level from C his high enough, then B won't transmit 
(since its baseband will set CCA, which inhibits transmits
    for the duration of the incoming packet.)
2) If the signal level is below the CCA threshold, but somewhat above 
the sensitivity threshold then, if the packet is on-channel
(or nearly so, say ch3 on ch1) then the second packet is seen as 
"interference".  If not, then the second packet is seen as 'noise'.
In either case, the sensitivity level is lowered, at the least (this 
reduces range).  If the signal level is high enough, then the "channel 
vector' is changed, and the receiver at A can't decode the (rest of) the 
incoming packet from B.
3) If the signal level is below the sensitivity threshold, but above the 
noise floor, then the effective noise floor is raised.

In situation 3, range and/or modulation rate is reduced.  (By 'range is 
reduced' I mean that you'll find things work better (but not perfecctly) 
if you decrease the distance between 'A' and 'B'.)

In situation 2, the modulation rate is typically reduced as 'A' no 
longer can ACK the incoming frame.  In the degenerate case, no 
communication can take place at all.
In situation 1, throughput is reduced, as 'B' and 'A' won't transmit 
while either has an incoming frame.  This is actually the best of the 
three conditions, and explains why its better to operate co-channel than 
in a situation where '2' or '3' is happening.

The subject is covered somewhat differently here; 
http://www.netgate.com/zz_faq.php#67

And er, here:
http://www.smallworks.com/archives/00000087.htm
http://www.smallworks.com/archives/00000068.htm
http://lists.soekris.com/pipermail/soekris-tech/2004-September/021771.html
http://lists.soekris.com/pipermail/soekris-tech/2005-July/023799.html
http://lists.bawug.org/pipermail/wireless/2004-May/029616.html
http://lists.shmoo.com/pipermail/hostap/2004-December/008752.html
[...]

But thats about enough self-love, eh?  :-)

jim