Scott in Aztlan wrote in message ...
On Sun, 14 Sep 2003 05:14:48 GMT, "Craig Davidson"
wrote:

What frequency?

2.4GHz.

Sounds like a microwave oven. :-)

Or a cordless phone... Or an X-10 wireless camera...

Or an 802.11B wireless access point?

Hmm... Come to think of it, that might be a possibility, too!

Lots of people make WiFi maps by "wardriving:"
driving/walking/biking/training
around with a laptop and a program like Network Stumbler
(http://www.netstumbler.com) and a GPS receiver. A typical map looks like
this:

http://www.wificharlottetown.org/loc...charlottetownm
ap.jpg


Scott, i recommend you write software to collect the lat/lon of only the
maximum signal strength readings. then average all the latitudes (add all
the latitudes up and then divide by the number of samples). This should
result in an "average latitude" line on a map. then do the same for all the
longitudes. This "average" result should give you the location of the
transmitter if you take enough samples.

If you want to get clever then take lat/lon samples of many signal strength
readings but keep them separate. Only perform the averaging process
described above on lat/lons that had the same signal strength reading. All
of these "averages" should cluster around the transmitter location in
lat/lon. There is no need to convert lat/lon to anything else. keep it
simple. If you want to get even better results then the last step would be
to average the results of all the averages of all the signal strength
readings to pick one single most likely spot for the transmitter.

If the signals can be identified by a node name or an IP address then by all
means keep the samples sorted by access point id as you perform the
averages. if every access point is using the same default id then the job
will be a little harder.

This almost sounds like more fun than Geocaching