On 2007-12-02, Larry <noone@home.com> wrote:
> George <george@nospam.invalid> wrote in
> news:bvadnf8PPeVrpc7anZ2dnUVZ_uKdnZ2d@comcast.com:
>
>> Without repeating the spammers URL they use triangulation (the
> speed of
>> light is a constant so you can deduce how far away you are from
> a radio
>> transmitter etc).

Unfortunately the speed of light isn't really constant (only
the speed of light through a vacuum is), it can vary considerably
with the media the EMR is passing through. If the speed of light
really were a constant there'd be no refraction. I get the idea
though.

>> It gives much coarser results than GPS. It is
>> at least
>> something because say you bought a recent Blackberry from VZW
> which
>> has a real GPS built in that is disabled at VZWs request.
>>
>>
>
> Impossible. Pipedream. The signal from the tower rarely comes
> directly to your phone so it could be "timed" by anything.

That's not quite true. The signal from the tower usually
propagates directly to the phone, it is just that the phone
also usually receives additional, delayed copies of the signal
from reflections. Usually the directly propagated signal is
the strongest (the reflections travel further, and are additionally
attenuated by the reflection itself), but sometimes an obstruction
on the direct signal path will cause it to be attenuated as much
or more than a less-obstructed reflection.

> SELLphones operate at the bottom of the microwave band and bounce
> all over the place with "multipath propagation", bouncing off
> buildings, cars, terrain, bridges, and every electric wire over
> the street. This multipath is what used to cause your old AMPS
> phone to fade in and out when you were in a poor signal area.
> The direct signal either was aided by, or cancelled by, the
> reflected signals, all of them. If they cancelled too much, you
> lost the tower and got noise. Move 3 3/4" (1/4 wave on 800 Mhz)
> and you found a "peak" signal where the reflected waves were in-
> phase with the main signal.

This is true for narrowband signals, like AMPS, but much less true
for very wide signals, like 1.25 MHz wide CDMA (let alone 5 MHz wide
WCDMA). While you might get a null somewhere in that bandwidth
it is exceedingly unlikely that that phase relationship will be
the same at every frequency across the entire bandwidth. If the
path length difference is at least 240 meters, for example, a
null somewhere in the 1.25 MHz bandwidth is guaranteed to be accompanied
by a peak somewhere else. Essentially, reflections don't usually
cause wideband signals to fade in total, they just cause them to
distort.

For CDMA this distortion is corrected in the time domain by the
use of a rake receiver. Essentially, separate receivers are used
to track the direct signal (if you can hear it, the usual case)
and each signficant reflection separately, and to add them up after
time-shifting them back into phase alignment to remove the distortion.
In principle, then, the reflections sort of improve your situation since
they give you additional signal to use that you wouldn't otherwise have.

> All this bouncing about would just trash the timing of anything
> trying to use timing delay for locations. These timing delays
> from the longer bouncing paths limited AMPS and digital data
> speeds to 1X for years until a new quadrature scheme was invented
> to error correct it.

But not for CDMA. CDMA (due to its broad bandwidth) can, and
does, separate the direct signal from the reflections for its
own purposes (and you can tell which is the direct signal since,
if you are hearing it at all, it will be the one which arrives
earliest), so generally with CDMA you can find exactly the
right signal to extract timing from. It is exceedingly rare
to have the direct signal attenuated entirely, and to only be
receiving significantly delayed reflections.

This is not to suggest that CDMA positioning will be nearly as
accurate as clear-sky-view GPS. CDMA towers aren't as closely
timed as the transmitters in GPS satellites, and RF that needs
to travel through or around things to get to you will have much
greater ambiguity in propagation path delays than RF from the sky
where the only significant obstruction is the ionosphere (which
can be modelled to provide partial corrections). But CDMA timing
also won't be nearly as bad as you seem to think it might.

> GPS sucks when you are under some covering that causes you to
> operate on reflected satellite signal for the same
> reason....timing delays.

I don't think so. GPS signals are broadband DSSS signals, just
like CDMA signals, and are just as resistant to a lot of this
stuff. GPS "sucks" mostly because the receivers are receiving
signals at just fantastically minute signal strengths that are
easily attenuated to nothing. And if you block 2/3's of the sky
with a roof, and have to make positioning fixes just with the
satellites in the remaining third that you can see out a window,
the geometry is such that the position fixes will quickly become
quite noisy and poor even if the timing of the satellite signals
you can still hear is still pretty good.

CDMA signals, on the other hand, are relatively quite strong and
will make it pretty far inside buildings. There's just too
much unpredictable stuff going on on the propagation path for
it to be nearly as good as clear-sky-view GPS.

Dennis Ferguson