geo-tester

Research tool to assess gps device behavior when passed different traffic.

Pairs great with https://github.com/era1112/pi-otg (gpio control of scripts)

Findings:

"primitive" garmin-type gps's accept the stream and decode a position
"smart" devices (ie. android handsets) quietly enter an undefined state, where the handset is unable to figure out where it is but never enters a "gps loss" state.
- The system indefinitely holds in "finding location" but never throws a "no location" alert. Some apps will continue to show that last known position. The device is also confused enough that it is unable to determine it's position through wifi location services.
- GPS debug tools show that the device recieves and decodes the beacons, enumerate SV's, but something prevents the kernel from delivering a "current location" to userland.
- In this state, the device would not know it's accurate location, and since it never formally enters a "gps loss" state it would have no means of logging the interaction.

Setup:

Download real-world RINEX data:
- https://cddis.nasa.gov/archive/gnss/data/daily/ (free login)
- download the file at: /[current year]/[highest number]/??n/brdc*
Produce research data:
```
- gps-sdr-sim -e <RINEX_FILE> -b 8 -l 0,0,200 -o ./researchtape.bin -v -s 2600000 -d 7200 -T 2021/02/23,20:34:00
```
- b8, s2600000 are rf parameters to match the hackrf, l is location in lat/long, 200 is elevation, d is length of tape in seconds, T is the start time and forces toc/toe to align, o is output file.
Modulate research data:
```
-   ./gps-sim -e rinex/*.23n -l 0,0,200 -s now -g 47 -a -d 1800 -v -i -r hackrf --disable-almanac
```
- e is input, l is loc, s is start time, g is gain, a sets amplifier on, d is duration, r is radio mode, vi is verbose interactive
- disable-almanac removes coarse almanac data from the stream (more complex to transmit/recieve, not necessary for this experiment).

Documentation:

Dependencies:

Future:

Provide feedback