[net.space] How does Geostar work?

VLSI%DEC-MARLBORO@sri-unix.UUCP (07/31/84)

From:  John Redford <VLSI at DEC-MARLBORO>

Can someone on the net tell us how the Geostar navigation system works?
The accuracies claimed for it are much higher than those for the military
Navstar system (less than 1 meter positioning accuracy versus 30 meters).
However, the Geostar ground units are simpler (they do not need an accurate
time base) and they use fewer satellites (four versus eighteen).  How can
they do so much better with so much less?  Technical answers only please,
not flames about the inferiority of anything done by the government.

John Redford
DEC-Hudson
   --------

dietz%USC-CSE%ECLA@sri-unix.UUCP (08/02/84)

Geostar works by triangulation.  A signal addressed to a portable
tranceiver is sent from the central ground station via one of the
satellites.  The ground unit then emits a very short omnidirectional
microwave pulse (with ID information included).  All four satellites
receive this pulse, and transmit it back to the ground station.  Time
delays are used to compute position.  The position is then transmitted
back to the ground unit.

The Geostar ground unit is simply a microwave tranceiver capable of
emitting high power (500 watts, I believe) short duration pulses, along
with some fast control logic to detect when the unit is being polled
and to receive and display position information.  Average power
consumption is low, because the pulses are so short (a microsecond?).

The Navstar system uses passive ground units, since they must operate
in combat conditions where radio silence is critical.  These passive
units must do the triangulation themselves, so they are expensive.

Paul Dietz
dietz%usc-cse@usc-ecl

henry@utzoo.UUCP (Henry Spencer) (08/05/84)

> Can someone on the net tell us how the Geostar navigation system works?
> The accuracies claimed for it are much higher than those for the military
> Navstar system (less than 1 meter positioning accuracy versus 30 meters).
> However, the Geostar ground units are simpler (they do not need an accurate
> time base) and they use fewer satellites (four versus eighteen).  How can
> they do so much better with so much less?  Technical answers only please,
> not flames about the inferiority of anything done by the government.

The two systems are almost duals of each other, in the mathematical sense
of "precise opposites".  Navstar works by having the user terminal ("ground
unit") time the signals from several different satellites, whose positions
are known accurately.  Geostar works by having the satellites and the
central ground computer facility time the signals from the user terminal.

The Geostar system puts most of the smarts, and most of the cost, in the
satellites and the central ground computers.  Navstar has dumb satellites
and no central ground facility, and consequently needs smart user terminals
with accurate clocks.  Geostar's user terminals are basically just digital
radios for talking to the satellites and sending them pulses to be timed.

The greater accuracy of Geostar is mostly a question of being able to
invest more in accuracy when the investment is centralized, I believe.
They may get some benefit from timing one signal at a time instead of
having to broadcast continuous position-reference signals.  And it
probably helps that they use Clarke-orbit ("geostationary") satellites,
whose positions do not change rapidly.

Geostar would probably need a few more satellites for truly global
coverage; the four-satellite requirement is basically for coverage of
the Americas, I think.  And I suspect that Geostar's accuracy is poor
near the poles, where the satellites are nearly on (or actually below!)
the horizon.  Navstar, being a military system, has to think about
these things, hence large numbers of satellites in non-Clarke orbits.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry