Theory: Creating "LPS" transmitter for GPS receivers

Question

As I understand it, GPS works by having transmitters with known locations send a timestamp, and calculating the time it took to receive it, so that you can triangulate your own position, (essentially an intersect of three or more circles with known centers).

I am interested in whether it is possible to create a "local" gps sender and using it with an unmodified gps receiver.

Sci-Fi Example to clarify

Let's assume that "suddenly" a country finds itself transported 100 years back in time, and all GPSs are rendered useless. Would they be able to (for example) establish GPS transmitter towers without having to reconfigure the GPSes, or would each GPS unit have to be programmed with information (eg. the coordinates) about the new transmitters?

Answer 1

whether it is possible to create a "local" gps sender and using it with an unmodified gps receiver.

Yes, if use means fooling a (C/A only) receiver into believing it receives a real signal. This is what GNSS simulators do for test purpose.

No, if use means determine an unknown position with an unmodified (NAVSTAR-) receiver from signals sent by transmitters stationary in ECEF (earth centered earth fixed) reference frame.

The reason is, that the CNAV and LNAV messages, telling the orbital elements of the space vehicles, are restricted to orbits in Earth Centered Inertial (ECI) reference frame. Even when sending the mean motion A as zero (which is possible), there is no way to cram the position of an earth based transmitter into the message.

Your hypothetical country would have to develop a custom Interface Specification and modify receivers to interpret this message.

Earth based navigation systems using NAVSTAR-like modulation may operate with known transmitter positions or use such a custom message, but this is not covered by the GPS/NAVSTAR space/user-segment definition.

Answer 2

Search for GNSS simulators. They are widely used for testing GPS receivers, software, phones, etc. If you make GPSs, you can't just walk outside and see if it works. For a controlled test of time-to-fix or position accuracy or sensitivity you need a simulator, and they're widely available.

That said, your objective of actual navigation for many unmodified receivers probably wouldn't work. A GPS receiver knows how to predict the orbits of the transmitters based on a few orbital parameters: transmitters on poles would not be possible to model as orbits.

In summary, you can fool one GPS easily, take it on a tour. But you can't make real signals to be correctly interpreted by many unmodified receivers.

Answer 3

A local GPS transmitter, of which there are many, at harbour entrances, runway threshholds etc, tend to get called 'pseudolites'(Pseudo-Satellites)