[FOSS-GPS] Post-processing RINEX to simulate RTK

[Homan, Thomas]

Hi Antonio,

Huh? You've confused me a bit so could I ask for some clarification on a few things please. I'm basing my experience on use of our Trimble and TopCon RTK equipment over the past decade.

Who is the manufacturer of your GPS equipment that you used to collect the Rinex data?

Please define 'very demanding' in terms of horizontal and vertical accuracy requirements. I wouldn't go near GPS equipment for laying out bolt hole patterns for mounting stuff to a slab. All GPS RTK equipment has a positioning tolerance. For our Trimble gear the stated spec is +-(5mm +1ppm (per km of distance from base to rover))

RTK by it's very nature does not utilize a intermediate PC. The rover applies a correcting factor that was developed by the base and broadcast in some fashion. Adding the PC would have caused many fits. When we download our field data collectors into the software a vector is show that includes various statistical values such as accuracy estimation and 3d covariance matrix between the base and rover. What does the RTK baseline data look like when you download your data collector into the mfr's processing software? Our Trimble data collectors generate a *.dc file that can be analyzed. Interestingly RTK vectors are not stored as a final point but as a delta x,y,z vector from the base station along with the various other statistical data. At least for our Trimble and TopCon units, the actual measurement occurs only on L1 no matter what the marketing adverts claim. As most RTK work is done at distances under 10km from base to rover adding the second  or additional  frequencies can actually add noise into the solution.

Did you store the Rinex observations during the period when you were collecting the original RTK data or during a later session? If you collected the Rinex data at the same time then you are wasting your time, statistically speaking. You need 2 independent datasets to do analytical processes. Using the same data more than once would only display errors in your processing techniques, not the measurements developed during the RTK session. 20Hz data collection was probably fine but the RTK processing of our 2 different mfr's work at 1 or 2 HZ. As I'm sure you saw, 20Hz data fills up memory cards VERY quickly.

To answer your question, the processing software that come with your GPS units should be able to do the work depending on the features that were licensed with it. If your software won't do the job you might look at the GraphNav series of software to do your validation with. I know our Aerial Imagery contractors utilize it (or another software package google GPS aerotriangulation software) in a form similar to yours only they are keeping track of an aircraft in flight coupled to INS and other stuff to determine camera position during a imagery collection campaign. 20Hz data collection is a very normal collection speed and the software works just fine with it. I was able to get a 30day demo from the GraphNav reseller without too much hassle.

Regards,

Tom



From: foss-gps-bounces at lists.osgeo.org [mailto:foss-gps-bounces at lists.osgeo.org] On Behalf Of António Pestana
Sent: Monday, October 24, 2011 10:57 AM
To: foss-gps at lists.osgeo.org
Subject: [FOSS-GPS] Post-processing RINEX to simulate RTK

Sorry if this is not the right forum to post my question. I'm a newcomer and I am not by no means a GNSS expert. But I want to learn.

I want to evaluate the quality of RTK differential phase positioning for some very demanding engineering applications. Unfortunately I was not able to connect rover and reference station to the same computer, so I could not use commercial GNSS software to do the RTK processing of the observations. As a workaround I decided to store the observations (collected at both sites) in RINEX format files. Now I have a bunch of RINEX files, obs (phase at 20 Hz) and nav, from base and rover.

I want to simulate real-time (as if they were obtained in the field) double-differences double-frequencies phase positioning of the rover by processing these RINEX files. I also need one position for each epoch of observations and, for now, I do not want to use any kind of "smoothing filter". Can I use a post-processing application to do the job ? If this is not possible, can you explain me why?

Regards

Antonio


PS:  Some days ago I've found RTKLIB. It's a very interesting application. It would be nice if I could use it to solve my problem.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/foss-gps/attachments/20111025/5b05e441/attachment.html