From: rootmoose at telia.com (Ake Nordin)
Subject: WiFi question

(with a nod to Esmond Kane)
At 17:50 2004-11-19, Paul Schmehl thusly scribed:
> --On Thursday, November 18, 2004 09:32:27 AM -0600 Paul Schmehl <pauls@...allas.edu> wrote:
> 
> > --On Wednesday, November 17, 2004 12:41:44 PM -0500 "Lachniet, Mark"
> > <mlachniet@...uoianet.com> wrote:
> > 
> > > Could also be RF interference.  One of my coworkers tracked down a
> > > particularly interesting problem with motion sensor lights.  

<quoted text trimmed down a bit, some lines broken in the process...>

> After forwarding this to our wireless expert, he responded
> with this (which he has authorized me to forward to the
> list.)
>  
> I find it hard to believe that this is possible.  2.4Ghz is
> the 9th harmonic.  By the time you get to the 4th harmonic
> of a signal, even in very very noisy radiators, the strength
> of the harmonic component of the signal is extremely minute.

Says what? Not every distortion mechanism give monotonically
falling spectral intensity. Device resonance may tilt that
spectrum substantially. If the stuff is cheap enough, it's
antenna may be a vital part of that resonator (i.e. far
better tuned at 2.4GHz than at 240MHz...)

> And, given the fact that one of those sensors (which most
> likely does *not* truly operate in the 240MHz portion of the
> spectrum) will have a very low output (Part 15 device), the
> 10th harmonic of that signal will be undetectible as it will
> be at or below the level of background noise.

Low output it may be, but received power is inversely
proportional to distance squared in ideal (freefield)
conditions. The AP inside the same building (room?) is
possibly quite close to the detector. Then consider the
irregularities of radio propagation inside buildings, and
the possibilities of various structures that can act as
waveguides...

> Finally, if a device managed to get past all of the
> improbabilities above, the chances of it *accidentally*
> creating a signal that looked like an 802.11 beacon packet,
> complete with preamble, header, etc is so off the charts as
> to be laughable.

This (the preamble especially) is what _should_ eliminate
the motion sensors from the list. I'm out on this one (too
lazy to do the math), but is the 802.11b air interface that
resilient (does it really require that much redundancy)? It
should be, but that would also be some lost (usable)
bandwidth.

> > One other thing...  If that device truly was operating
> at 240MHz, then the first harmonic would be 480MHz.  I'm
> pretty sure that frequency lies in the public service bands
> (ie fire/police).  If not, its very close.  Given that and
> the fact that the first harmonic would be much stronger than
> the 9th harmonic, I'm pretty sure someone in those bands
> would have complained loudly to the FCC as they don't take
> intereference issues in those bands lightly.

Sorry. 

1) The building will contain very much of that energy
(which never was very much on a metropolitan scale, FCC Part
15 and all that).

2) The noise characteristics as received by those services
would be intermittent, very bursty and come from many
different directions all over the city. No easy clues telling
what to complain about there.

3) I don't know about US emergency communication radios, but
typical European systems (before Terrestrial Trunked Radio)
are so bad anyway that this contributed noise hardly would
be noticed.


-- 
  .
 /Ake Nordin       +46704-660199       rootmoose@...ia.com
 Duston Sickler: "There are only 10 types of people in the
 world, those who understand binary and those who don't."

Powered by blists - more mailing lists