Quaoar hath wroth:

Nice analysis! One point: the wireless signal is constant, but the
household microwave signal is semi-discreet in time.

You have it backwards. When belching RF, the microwave oven power
envelope looks like a half wave rectified waveform at 120Hz (twice the
power line frequency) at "full power". The average power is about 70%
of the peak power. However, you're correct that most people don't run
a microwave oven continuously all day long. (Exception... Microwave
plastic injection molding pre-heater and other industrial microwave
ovens). I think (guess) that FDA/CDRH 1030.10 specifies RMS (heating)
or average power, not peak power.

802.11g has an even lower average duty cycle. FCC 15.247 specs are
written to prevent any one spread spectrum wireless device from
hogging all the air time. That means that it not only doesn't sit on
one frequency, but also turns off the power for some time to allow
other devices on the same RF channel to function. I don't recall the
exact numbers and am too burned out tonite to dig through the specs.
However, I did find:
http://www.elliottlabs.com/documents/OFDM.pdf
which tries to analyze the typical transmit duty cycle for 802.11g.
The authors guess is about 10% for typical traffic, which I guess is a
good typical value. Like the microwave oven, most users are not
belching wireless data continuously. (Exception... streaming wireless
video).

If the microwave is in use one hour per day [ the 1 hour per day use of
a microwave oven is in my experience far greater than most people use
for the cup of tea, soup, etc.], the equivalent energy exposure of the
two sources of radiation are (using your numbers):

Household microwave oven: 1 hour * 5 mw/sq cm = 3600sec * 5mw/sq cm =
1800 mw-sec/sq cm.

OK, I'll use that. However, note that the 5mw/sq-cm is probably
average power.

Wireless transmitter: 24 hr * 0.16 mw/sq cm = 24 * 3600 sec/hr * 0.16
mw/sq cm = 13,825 mw-sec/sq cm.

Nope. You forgot to apply the duty cycle of 10% transmission time.
That reduces it to 1383 mw-sec/sq-cm. However, you're also assuming
that this user spends 24 hours in front of the wireless access point.
That's possible for confirmed programmers, hackers, and fanatical
gamers, but methinks the average user will see a much shorter exposure
time. If we assume a workplace model, I would guess 8 hours exposure
which would reduce the exposure rate to 461 mw-sec/sq-cm, or about 1/4
that of the microwave oven.

Similarly, the 10% average duty cycle varies by the type of user. The
fanatical file sharing addict will probably approach 100% duty cycle,
while the light weight mail and web page surfer, will be close to 10%.

The total energy exposure from the wireless is 7.68 times the microwave
exposure on an integrated basis. This is admittedly a quick estimate,
since almost no one is located directly to either source in practice.
Yet, it is not conclusive that the wireless is trivial compared to the
microwave.

The problem with this type of analysis is that the various assumptions
that have been made in order to generate a single "typical" value have
such a wide range of potential errors, that the resultant
conglomerated calculated values are almost worthless. The main
problem is that the typical wireless user may not also be a typical
microwave oven user. In addition, we've ignored the distance from the
RF emitter, which has a huge effect on exposure (inverse square law).

I suspect that I might be the worst case user. My Verizon XV6700 cell
phone has an 802.11b radio inside. I often leave it on doing the
WiFiFoFum data collection. Even at 10% or less duty cycle, the
proximity of the RF emitter on my belt dramatically increases my
exposure.

The correct way to do this would be some manner of dosimeter, similar
to an ionizing radiation dosimeter. It takes into consideration duty
cycle, signal strength, distance from emitter, and such. It just
accumulates the total RF power exposure. I could probably design and
build one, but I don't think that selling to the paranoid market is
going to be a accepted as a winning business model.


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