Post Processing with RTKLIB

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Post Processing with RTKLIB

RTKPostProcess
Dear Everyone,

I would like to ask something about RTK and its possibilities.
I have a device that can receive GNSS data(.binr), and I have connection to the local RTK network via NTRIP(rtcm3).
I receive GNSS data with a device (with limited internet connection, so when we get the first position we are able to send it to the server, but every other will be uploaded in the next 1-2 day), and I need the base station data which can be captured through NTRIP.
What i want to do is, capture the GNSS data with the device, send the first position to our server. After this hopefully we will be able to capture the nearest RTK station data to the devices first position.
In theory, in this point, we have the GNSS data which can be uploaded in every one or two day, and we have the base station data. With these I hope we will be able to get a good FIX solution during post processing.
One of the biggest issues with this for me is the synchronization.
I would like to hear someone with a little bit more experience in RTK about this issue. What do you think about this, is this possible, or should I search for another solution to my problem?

Thank you for your answer!

Best regards, Greg
Reply | Threaded
Open this post in threaded view
|

Re: Post Processing with RTKLIB

DavidKelley@ITSware.net
Post processing raw observation data gathered in the field has been very common decades and you should have no problems using RTKLIB (the RTKPOST tool) for that.  Many "old timers" still do it that way for various reasons and when we are in the filed we are checking as we go but we old "trust" the post processed results.  And post processioning means you can also use final orbits and other data products that are readily available to get better results.  All of this is moot if your field data is itself corrupted.  And you can easily turn RTCM3.x into RINEX files with other RTKLIB tools and find 3rd party processing services to solve it for you.

Time synchronization will not in fact be an issue.  RTCM 3.x provides the precise time over the current GPS week, so you have to tell the tool (RTKLIB or any other such tool) what week a file came from but the file name conventions used often help with that.  There are a few other messages and tricks, but this solves 99% of your issues.  The navigation filter will then determine the precise time.  

The data from the "nearest RTK" station of course depends on where you are.  Many tools (RTKLIB, SNIP, and often the NTRIP Caster device itself) will provide a means to capture the raw RTCM data for your post processing needs.  NGS CORS even runs a long term FTP site with such data, but the data is decimated.  Depending on your application that may not be an issue.  

So yes this is quite possible and routinely done. But you have not told the group anything more about the devices in the field (the rovers) and what span of location and how often the positional estimates are wanted, what precision is required, etc. Many of the practical details and useful advice you might get will turn on that.    Regards, DCK
Reply | Threaded
Open this post in threaded view
|

Re: Post Processing with RTKLIB

RTKPostProcess
Thank you for your answer,
The device contains an NV08C-CSM GNSS module, with 2JM013-010/113-UFL antenna. The desired precision is 2cm, and we need them as often as possible, we want a really precise route from the FIX data.
My problem with the synchronization is that we can capture both the rover and the base station data, but because we capture it in two different place, it maybe cause some problem when we want to process it. For example because of the slow internet connection or anything else the .tag file we create for this probably will have a few second delay.

Best regards,
Greg
Reply | Threaded
Open this post in threaded view
|

Re: Post Processing with RTKLIB

DavidKelley@ITSware.net
In this case you are planning to use low lost L1 only device with a ten dollar patch antenna.  Search for Michele Bavaro to find practical use tips on that chip and RTKLIB.  By now you probably know L1/L2 is to be preferred, and that you need to be <15km (better yet <10km) from your base stations to get fixes with this setup.  We have done RTK mobile <2cm work on vehicles with the NV08 and RTKLIB, but we tend to use better antenna (see https://www.use-snip.com/kb/knowledge-base/question-what-antenna-should-i-use/).  You do not state your data rates, only "as often as possible" - to some people that means 1Hz, to others much more (that chip does 20hz with the normal increase in measurement noise floor). One presumes you have a generally clear sky? No forest cover? Is this a drone, a ground vehicle or a manpack or other? This directly effects the expectations of signal loss you will have.  You will end up with lots of data, in the NV08 propriety format, that you will then translate to RTCM3 or to RINEX for use in post. And RTKLIB does all that just fine.

If you are processing "at the end of the day" then the slow internet should not matter here (see prior reply about time tags, I do not think you will need to use the tag option), but perhaps I do not understand your config. It is likely you will want to record the RTCM streams from all the base stations your device may be near and select which is best in post.  You have not mentioned what the span of rover operation would be.  BTW, a point I failed to note in my first post of that you need not send a precise / good position to the NTRIP Caster and you only need to send it once.  For most needs, a 1km will suit, so in fact the unaided fix from this device in NMEA-183 format will be fine (or you can add it by hand). The Caster will then pick the best station to send to you (and you will get the exact location for that data).