[BC] GPS

[Jonathan E. Hardis]

>A few comments. GPS may well have noise problems. I know that my 
>handheld unit will not work in Manhattan unless I am right by the 
>river on the harbor. I always thought it was building reflections 
>but noise and attenuation might be the real issue.

This is known as the "urban canyon" problem.

In order for GPS to work, the receiver has to have enough of a clear 
view of the sky to see four satellites.  (It solves four simultaneous 
equations for x, y, z, and t.  With a really good internal clock, 
giving t to nanoseconds, view of only three satellites is necessary.) 
When tall buildings limit the amount of sky you can see, the odds of 
it containing four satellites is small.

The GPS constellation consists nominally of 24 satellites -- 4 in 
each of 6 orbital planes.  (The system would still be within spec 
with fewer satellites, and it may contain more.  At the moment, there 
are 30 or 31 up there.)  4/24 is 1/6.  In other words, you should be 
able to see 1/6 if the entire sky -- 1/3 of the sky not counting the 
portion blocked by the earth below you.  That's tough to do at street 
level in much of Manhattan.

   - Jonathan