Replies to This Discussion

Permalink Reply by Morli on February 5, 2010 at 7:43pm

Ken,
putting it differently , what do I need to make my own diy WAAS ground station? ( some thing I wanted to exploit so called extra accuracy in my GPS)

• ▶ Reply

Permalink Reply by Mike on February 6, 2010 at 7:53am

A starting point would be to put the two GPS units in a clear area a metre apart. Log the two GPS units for a while then compare the results and see what the correlation is.

Everything I have read indicates that using two GPS units as you have proposed does not work but no harm in trying at least for your own self satisfaction!

• ▶ Reply

Permalink Reply by brakar on February 7, 2010 at 4:34pm

Jason, I think your idea is theoreticaly very interesting. I have little idea of how it would work in practice, but a few considerations;
- as I understand it, the 406 GPS has a 1Hz update freq. If the plane is moving 10 m/s there will be somewhere between 0 to 10m diff in the relative gps possitions from the two gps's depending on how syncronized they are.
- There might be a transmission problem. The ardupilots are currently using one way telemetry I think, so there will be some work related to feeding data from gps1 to gps2. (Or is the 406 waas enabled, so you can send data directly on the waas freq?, if so isn't the gps reading such datas already? and isn't the waas freq. restricted - for protection against interference?).

PS; if such a systems worked, it might have been of great help in order to make automatic landing possible also for diy systems.

• ▶ Reply

Permalink Reply by Simon Wood on February 9, 2010 at 10:49am

WAAS signals are 'in-band' from stational satellites at L1 frequencies (1.5GHz), they model the north american content's atmospherics on a grid based structure (20KM squares IIRC).

DPGS/LAAS uses another transmission and models atmospherics at a point. DGPS beacons are at around 250KHz and normally require a (serially connected) RX only radio in addition to GPS. LAAS can be whatever frequency you are licensed for and could be WiFi, Xigbee, etc.

There is no reason why 'big brother' would seek you out unless you are operating on a non-licensed frequency.
Simon.

• ▶ Reply

Permalink Reply by Simon Wood on February 8, 2010 at 12:36pm

Hi Jason,
I'm afraid 2 recievers outputing computed GPS position will not help you, however if you can place one (or both) the units in a mode where they output pseudorange and code measurements then you can do some 'heavy lifting' processing to compute accurate corrections.

The corrections could be of two forms:\
1). LAAS - basically DPGS corrections but highly specific to your position. A fixed 'reference station' knows it position highly accurately, and therefore can sense variation in the timing of the GPS signals from each satellite and build up a model of the atmosphere and communicated this to a 'rover'. Most 'rover' GPS receivers should be able to process this type of correction. I don't know of any 'free' software which can do this, but would love to find out that there is some...

2). RTK - pretty much the same as above, however the base sends details of each satellite (rather than a modeled atmosphere) and the rover must do some pretty heavy processing it's self. The RTK stream itself is fairly low bandwidth (around 200 bytes/s). There is a 'free' project (http://gpspp.sakura.ne.jp/rtklib/rtklib.htm) that can perform the calculations, and is said to work on a Beagle board.

Depending on what you are trying to achieve you may be able to record the pseudorange data in flight and post process, for example if you are trying to work out accurately where a photo was taken. You may even be able to 'downlink' the pseudorange data from the drone and compute position on the ground, but I fear that any latency will make the position irrelevnt.

Cheers,
Mungewell.

• ▶ Reply

ModeratorPermalink Reply by Jack Crossfire on February 8, 2010 at 8:35pm

Do a Goog on Poor Man's DGPS.

• ▶ Reply