Hacker News new | ask | show | jobs
by tamcap 1422 days ago
> how do you measure your own gps “accuracy”?

Because GPS is way cool. First, there is the official site: https://www.gps.gov/systems/gps/performance/accuracy/ I am far from being an expert, but being a flying nerd, that's what I can tell you:

The more satellites in sight you have access to, the cooler stuff you can do with GPS. Minimum to establish 2D (on globe projection 2D) location is 3 sats, minimum for altitude is 4 sats. There are mathematical simulations you can do to see if, and when, a given aircraft will have access to 4 sats (or more), called RAIM: https://www.mygdc.com/new/static/usermanuals/web_help/RAIM_P...

Now, things get even better if you use 4+ sats, or sats from different constellations (I am not sure how mature is cross-constellation data validation, however, so let's pretend 1 constellation for now). With 5 sats, the receiver can calculate the position even more precisely, determine the level of URE, etc. You can even detect that one source is unreliable, and discard it (effectively falling back to 4 sats). With even more sats you can (theoretically) do cool data analysis, where you can determine those factors even more precisely.

Now, add the inertial factors. Many passenger have either in-software (history of previous positions) or in-hardware (think accelerometers, magnetometers) inertial data sources. A sophisticated flight management software (FMS) can cross-corelate the inertial data with GNSS data and determine when the latter becomes unreliable.

All of those measurements and factors can subsequently be reported via ADS-B or ACARS, to help with fleet management, and general predictive maintenance. While aviation technology might sometimes look way out of date at a first glance, the modern airplanes have some really cool technology hiding in it.

1 comments

Friend of mine is doing ocean bottom surveys. He uses pretty decent GPS receiver, but also gets real-time (or applies it later) "correction" data from ground based stations that calculate error caused by atmosphere, which gets him down to 1cm accuracy.
That's SBAS/WAAS! another cool thing

This is trying to measure ionosphere disturbances which affect signal propagation rates via ground reference stations. The real-time model is then calculated and distributed using ground stations and geostationary satellites.

GPS is way, way cool, and it's a pity the extent of its engineering is not better known.