Scott in Aztlan wrote.

all you have is a radio receiver with an
omnidirectional antenna which gives you a signal strength indication, the
assumption being that the RSSI is roughly proportional to the line-of-sight
distance between the transmiter and the receiver). The idea is to write a
computer program to take these position and RSSI values and prodce an
estimate
of the transmitter's location.


Dale DePriest wrote in message ...

It is an interesting idea theoretically. You could plot circles on a map
from the relative signal strength value and centered on the GPS
location, then increase the circle sizes proportionally until they
intersect and this will provide a rough idea of the location. With more
locations it would be better and better, throwing out some of the
circles that don't fit. It will probably work if you live in one of the
plains states and have a very sensitive RF meter and lots of driving time.

Dale

One problem is that when you get close to the transmitter the signal
strength meter hits the maximum and ceases to provide useful information. A
method for attenuating the signal is useful for this problem.

I like the intersecting circle idea. Remember to use the inverse square law
for the circle sizes. The power level of the signal will decrease by the
square of the distance from the transmitter. Its not linear.

A similar concept to the intersecting circles theory is that if you collect
a set of data along a straight line (by driving down a straight road) then
the transmitter is most likely on a perpendicular line crossing the road at
the point where you received the maximum signal. You can work around the
"maxing out the signal strength meter" problem by noting the two points on
the line where the signal shows maximum and then draw the perpendicular line
half way in-between.
This idea assumes the transmitter has an omni directional antenna as well.