search

tomojitakasu / RTKLIB

2.55k stars 1.62k forks source link

agreement between PPP algorithms? #16

Open aewallin opened 10 years ago

aewallin commented 10 years ago

Hi, I want to use RTKLib for PPP-static analysis of clock offset. I did a comparison of different algorithms, see: http://www.anderswallin.net/2013/12/comparing-gps-ppp-solutions/

There is a constant offset of about 4 ns between CSRS-PPP/gLAB and the RTKLib-PPP solution. Does anyone know what settings should be used with RTKLib to get solutions that are consistent with the other PPP algorithms? The offset of 4 ns seems to be about half of the ZTD (!?) - but this may be just a coincidence ??

tomojitakasu commented 10 years ago

Such constant bias for receiver clocks is often derived from satellite-side DCBs. Please check results after applying P1-C1 DCB correction.

aewallin commented 10 years ago

Thanks, I have tried inputting DCB files to RTKPost, both the ones in the /data folder and ones I downloaded with RTKGet, but this does not remove the 4 ns offset between the RTKPost solution and the gLAB/CSRS solutions.

I would be interested in a test-case where one can show that the three programs (RTKLib/gLAB/CSRS) produce similar output. Is there a PPP-Static software test included in the RTKLib distribution?

xiangyan66 commented 8 years ago

Hi Aewallin, I wonder you fix the issue or not, or find the reasons for the difference. I think besides correcting the satellite DCB for P1-C1, actually the receiver DCB P1-C1 also exist, which will be absorbed in the receiver clock parameter. To keep the consistence, I suggest using P1 instead. C1 has the priority in RTKLIB. Regards,

aewallin commented 8 years ago

Hi, I have not done further comparisons of PPP algorithms since my original post (above). Since that test I have also learned that PPP is possible with the free GPS toolkit (http://www.gpstk.org), so there are now three open-source/free PPP implementations that I know about: rtklib, ESA gLAB, and GPStk. I am still interested in PPP for time-transfer, in particular there is new work on integer-PPP (iPPP), but I don't know if an open-source/free iPPP algorithm exists.

xiangyan66 commented 8 years ago

Hello Aewallin, I am not working on the timing. I have little knowledge on timing. As far as I know, RTKLIB have a PPP-AR mode. Probably, you can try it. regards,