"Thurman" <thurman@bigplanet.com> wrote in
news:vVu5k.1684$6C1.182@newsfe05.lga:
>
> When Google announced the 'My Location' service for cell phones
> without GPS, they explained they do approximation from the cell tower,
> determining direction and signal strength. There were a couple of
> things that were not explained.
>
> According to the guy that designed the user IDs determination system,
> licensed by three of the cellular companies, the tower and horn ID is
> broadcast as part of the ID packet during transmission. Assume there
> are six horns (antennas) on each tower. If the tower has been
> 'mapped', I can tell from the tower and horn ID the 'pie slice' where
> you are located. Apparently they also know the signal strength of your
> device.
There are THREE antenna panels on the selltower, not 6. If there are 6,
then two panels are connected together, in line vertically to change the
120 degree sector pattern downward to reduce interference to other cells.
The more selltowers, the further down they point the antennas. So, from
which 120 degree sector you are connected to, I can tell which of the 3
directions from the tower you are located.
>
> The CEO of Growell in 2003 claimed that Sprint reads the device signal
> strength, then increases the tower signal to compensate. He also
> claimed the other methodologies increased the device signal, thus
> decreasing battery life. I don't know if this is still the case. The
> point is, if you knew signal strength you could estimate distance from
> the tower maybe as far away as five miles BUT all kinds of radio
> issues could throw that estimate off. I have used tower tracking on a
> HTC Hermes and iPhone but it does not work on my Blackjack Version I
> not II.
Device strength is related to so many factors it isn't funny. Weather,
time of day, what vehicles are parked along what streets, what kind of
noise all the other RF sources are making near your bandwidth, Sellphone
loading itself in the area, sunspot activity that wipes out the
satellites (not a problem lately), foliage in the trees, that little girl
with the metalized helium balloon at the hot dog stand, and let's not
forget our old friend multipath propagation bouncing off every building
in 2 miles....all effect signal path attenuation, causing the tower to
change the sellphone's output transmitter power level to get reasonably
level signals at the tower on all the channels.
I'd say if you could figure it out WITHIN 5 MILES, you'd be doing great!
>
> Most, not all, devices are being tracked by two towers. The tower with
> the strongest signal is >usually< the tower serving the device. If the
> base station controller permits, mathematically you could overlay the
> pie slice of the two towers to find an approximate location. If you
> have not experienced this, they draw a blue circle of differing radius
> to alert you of the possible area. I've never seen a better location
> determination than 100 yards. Frequently it's 1000+ meters.
Without an accurate pulse ranging system, the three-sector arrays on 3
towers could, probably, draw us a triangle with you inside it 2-3 miles
on a side......"He's in there....we think."
>
> If you are in an urban area, enter street addresses, intersections,
> business names, etc. If you are in a rural area, traveling 60 mph,
> then a half mile is 30 seconds of travel. You'll see the blue dot jump
> as you are passed from one tower to another, giving an indication of
> travel.
>
> Since you are focused on the technology, research the Popular Science
> articles that will soon be provided for free in an online archive. An
> inventor filed for a patent to do 'triangulation' from two points.
> PopSci reported you can't >tri<-angulate from two points. He had to do
> a demo. He successfully demonstrated location determination from two
> mountain tops in Nevada, but the patent office was not satisfied. He
> arraigned for a test from one of the space shuttles. His test was
> aboard the shuttle that blew up at launch, not the one that came apart
> during re-entry.
Triangulation is easy with just TWO LOPs (Lines of Position). Ask the U-
boats that got sunk. But, to do this we need LINES of positions, not
SECTORS of position. The wider the lines, the more inaccurate the fix.
With an AM pocket radio, you can come within about 3-5 degrees width of a
line. With two of those fixes, you can plot a best guess in the middle
of each very narrow 3 degree beamwidth and be within a few hundred feet
of the AM transmitter's antenna at a range of 10 miles.
But the width of that line at a 3-sector sell tower is 120 degrees, not
5....a very wide area.
>
> I apologize for not being able to provide more technical info, I'm a
> designer and marketing guy. I'd suggest Wikipedia and check to see if
> the archives are online at www.popsci.com.
>
It doesn't. I've looked and looked. The reason for all the secrecy is
Sellphone Magic. We'll wish it so and it'll happen....or at least we'll
make them THINK it happens. I'm still hunting the basis of this
wonderful technology. I bet the ARMY and NAVY are most interested in it.
They've been plodding along with RDF since WW One trying to find the
enemy's transmitters. To do it without sophisticated directional arrays
and timed pulses like GPS is Cellphone Magic, indeed.....(c;
Technical Specs of iPhone 3G posted
Linear Mode
