How can a Satellite's position/orbit be calculated using only range measurements from ground stations? This task is often done in a process known as Satellite Laser Ranging (SLR). SLR stations (of known coordinates) track satellites, recording range measurements to the satellite at known times. I would like to know how the orbit / new position is calculated (I have no intention of doing the calculation myself). 
Note: 
the stations are globally distributed
multiple measurements are required in the orbit calculation
a starting "state" (position/velocity) is required in the orbit calculation
Thanks for any replies
If this question is better asked somewhere else, let me know
 A: At it's simplest level, satellite tracking is a 3D version of triangulation.  Suppose one were able to have 3 observers at the vertices of a triangle on the ground all simultaneously and instantaneously measure the distance from themselves to the satellite.  Then one could use the known coordinates of the 3 observers along with the measurements they each made to find the unique possible location for the satellite at the time the measurements were made.  Only basic geometry would be required, and one could repeat the measurements at regular intervals to track the satellite.  But, alas, this picture is too easy, and this isn't how satellite tracking works at all.
A much more accurate (though still simplified) picture is this:
Our satellite cruises along, sending out a broad radio signal to the effect of "Satellite 123 here, at the click, the time by my clock is 12:47:00.0000 (click)" ................. "Satellite 123 here, at the click, the time by my clock is 12:47:01.0000 (click)" .............etc.
Scattered about on the ground are stations which listen to this chatter from the various satellites, and whenever they receive a click, record the satellite's number, the time by the satellite's clock, and the time by the station's clock.  All of this data from all the stations gets sent along to a central processing computer, which has a massive task on its hands.
One might think that the computer could simply use the time stamp differences between the satellite clock and the station clock for each click to compute the transmission time and thus the distance from the station for that click. At that point, the computer would just need to have 3 stations in range, and it could solve the same geometry problem as above to locate the satellite at the time of each click.
But the relativistic time dilation caused by the earth's gravitational field is actually so much stronger at the earth's surface than in the satellite's orbit that it is not possible to synchronize the satellite's clock with those on the ground well enough to allow accurate position measurements.  So the computer has to do lots of extra calculations to correct for relativistic effects.
