Hacker News Reader

Comment by rzimmerman | Hacker News Reader

GPS receivers also have to correct for things like light traveling slower through plasma in the ionosphere. Receivers on the ground actually solve for their positions in 4 dimensions (they need to solve for time!). GPS is so cool.

View on HN

That's part of what the 4th connected satellite does. Also this is one of the better descriptions of trilateration I've seen. http://gis.stackexchange.com/questions/12866/why-does-gps-po...

Andrew Holme explains a some of these things succinctly on his home made GPS receiver project page[0]. Search for "Solving for user position", where he even mentions ionospheric propagation. He also has an appendix showing his math[1]

[0] http://www.aholme.co.uk/GPS/Main.htm [1] http://www.aholme.co.uk/GPS/user_position_solution.pdf

This can also be useful in unpredictable ways. One is the North Korean nuclear test a couple of years back[1].

"The blast generated infrasonic waves that propagated all the way to the upper atmosphere causing small variations in the density of electrons there.

By analyzing the signals from GPS satellites collected at ground-based monitoring stations in South Korea and Japan, scientists [...] confirmed the ionospheric disturbance generated by the North Korean test."

[1] http://gpsworld.com/researchers-see-ionospheric-signature-of...

Any idea if large-scale weather effects (like the jet stream) can cause noticeable accuracy problems?

It does, in fact GPS positioning error, derived from fixed GPS stations, is often used as an input for weather forecast models like WRF. The error is closely binded to the atmospheric water content. Using the different azimuths from the GPS constellation you can even build up a 3d representation of the water content on different layers of the atmosphere.

This is possibly the most interesting thing I've read on HN. I knew that a similar principle is used for CT scans, but I never thought of blowing it up to Earth-scale. Thank you.

Thanks. This work is interesting http://tel.archives-ouvertes.fr/tel-00444081 Unfortunately the main thesis is in french, however there are included articles in english and lots of references. I am currently working on a project using a high density GPS stations network we deployed in the south east of France. We have had good results for nowcasting and to map in 3d the atmospheric weather content. No publication yet though.

2d maps of rainfall have also been derived from interpolated GSM antennas, using the signal loss.

I've done a ton of research into precision positioning with GPS (centimeter resolution). Have you investigated using NOAA CORS data in your project? http://geodesy.noaa.gov/CORS/

Disclaimer: GPS is one of those parts of science that gives me the nerd chills.

Cors data is limited to the USA, we used RGP data, a network of ground stations maintained by the IGN which is analogous to Cors. The IGS gathers data throughout the world to enable global applications but it looks like the data is more complete when accessed through national networks.

I worked a few years ago on precision guidance. I was supposed to assess the performance of automatically guided tractors. Sub centimetric guidance on 4 tons huge tractors was pretty tricky to test. On what exactly did you work?

You also have satellites with GPS receivers, so that they can measure the effects when the GPS signal skims the atmosphere. This allows humidity and temperature to be measured with very high accuracy, with a very good vertical resolution.

http://en.wikipedia.org/wiki/GNSS_radio_occultation

You're making a good distinction here. The research on using GPS occultation for sensing water vapor that I'm most familiar with uses two satellites, and does not use ground stations.

I am biased here, because I know the people who are developing this technology, so I'd be welcome to further information on this point! E.g., I took a look at the thesis abstract pointed to by @kubiii, but it wasn't specific enough to tell what their exact sensing method was.

There are some generally good-quality summaries of GPS radio occultation here: http://geooptics.com/?page_id=148

The main advantages of GPS occultation for H2O sensing are that it's relatively bias-free (compared to microwave radiometers), that it can see through clouds, and that it has better usable vertical resolution.

Yes, that's exactly it. Two satellites - one is a standard GPS satellite, and the other is a satellite with a GPS receiver.

I wonder how many implementations of the calculations there are. Do all the commercial GPS and phone manufacturers write this on their own, or is there a tested version they can license?

A significant portion of the calculation is usually done in hardware (less 'written' more 'designed'), and yes these things are licensable. Most, though, have no interest in fabricating their own chips (and can't get much out of customizing one) and buy chips off the shelf. There certainly are several independent implementations.

This article has been posted here a couple of times:

http://www.aholme.co.uk/GPS/Main.htm

discussions:

https://hn.algolia.com/?q=gps#!/story/forever/0/http%3A%2F%2...

It mentions an earlier project also done by an individual:

http://lea.hamradio.si/~s53mv/navsats/theory.html

So even a few individuals are able to wrap their heads around the problem (I guess commercial projects would also tend to have one or two people doing significant amounts of the work).

A single, mildly impressive engineering graduate student should be able to construct a functional hardware GPS receiver from scratch in a year or so.

Reply on news.ycombinator.com