Unable to get centimeter level accuracy in the kinematic mode

[Amjad-58]

I am using ZED F9P Ublox and during Kiematic mode I do not get centimeter level accuracy. I collect data in city and my rover and base station has almost 30 km distance Please guide me what I am missing?

[SergeyDolin]

Hello, @Amjad-58, the accuracy of determining coordinates largely depends on the antenna you use. If you want to get a centimeter level of accuracy in kinematic mode, then I’m afraid you need a fairly high-quality antenna, the parameters of the phase center variation, which is in the atx file. When using the relative method, you should not exceed the distance from the base station of 10-15 km, since at large distances the estimates of the ionospheric delay, phase ambiguities and other parameters will not be equivalent, which will not allow achieving high accuracy when solving double difference equations. And speaking of accuracy, what do you mean? By internal convergence or external convergence?

[Amjad-58]

Hello @SergeyDolin,

I am using UBLOX's ANN-MB-00 multi-band antenna paired with the ZED F9P board. Unfortunately, the available public base stations are quite distant, approximately 25 to 30 km from my location.

For achieving centimeter-level accuracy, I collect raw GNSS data from UBLOX and process it using RTKLIB. I convert the raw data to RINEX format using RTCNV and then upload the files – rover (UBLOX) observation file, and base station .obs and .nav files – to RTKPOST. I set the positioning mode as kinematic since my rover is in motion while the base station remains stationary.

However, I am facing a challenge as I'm unable to achieve centimeter-level accuracy (Fix solution). I only get a float solution. Interestingly, when I keep the rover stationary and collect data for 10 to 15 minutes, I manage to achieve a 10 to 15 percent fix solution.

I am seeking your advice on how to achieve a 100 percent fix solution. What parameters or considerations should I focus on? I am currently using external convergence through RTKLIB, and I'm not familiar with internal convergence.

Your guidance on optimizing the settings or any additional steps to enhance accuracy would be greatly appreciated.

[SergeyDolin]

@Amjad-58 I'm afraid that with your configuration it will be impossible to achieve a 100% fixed solution.

I have collected several studies for you in your configuration, except that the base station in these works was very close. If you do not find the answer to your question in these works, then you can write to me by email (Sergeydolin@icloud.com) and I will try to help you understand your task in more detail.

1) https://www.researchgate.net/publication/354203105_Evaluation_of_Low-Cost_GNSS_Receiver_under_Demanding_Conditions_in_RTK_Network_Mode

2) https://www.researchgate.net/publication/359736981_Behavior_of_Low-Cost_Receivers_in_Base-Rover_Configuration_with_Geodetic-Grade_Antennas

3) https://www.researchgate.net/publication/343456912_Testing_Multi-Frequency_Low-Cost_GNSS_Receivers_for_Geodetic_Monitoring_Purposes

4) https://www.researchgate.net/publication/346509581_Performance_Evaluation_of_Low-Cost_and_Real-Time_Multi-GNSS_Advanced_Demonstration_Tool_for_Orbit_and_Clock_Analysis-Precise_Point_Positioning_MADOCA-PPP_Receiver_Systems

[SergeyDolin]

@Amjad-58 Have you tried using an exact ephemeris file when processing? Also try to include zenith tropospheric delay in the estimate. The solution should, in theory, be better. It is also very important to give the exact coordinates of the base station, and the antenna offset parameters, in order to increase the number of ambiguity resolutions.

[Amjad-58]

No, I haven't attempted those methods. My approach involves obtaining RINEX data directly from the base station with a significant baseline. I configure the settings in kinematic mode, turning off ionospheric and tropospheric delays, and setting the ambiguity for GPS as "Fix and Hold." Afterward, I initiate the execution. Are you suggesting that I should upload the precise ephemeris files from the base station? If so, could you provide guidance on how to do this? Additionally, I've been choosing the RINEX header for the base station coordinates. but I can give the exact coordinates as well. and I think I can also put the zenith option for tropospheric settings 1 of RTKPOST.

[SergeyDolin]

@Amjad-58

Accurate ephemeris files and other files for high-precision methods can be found here: https://igs.bkg.bund.de/root_ftp/IGS/

Precise coordinates for the observation epoch are required. Set up the configuration as in the pictures

[Amjad-58]

Hi there, I am unable to download clock/orbit products from the IGS. I don't know their directory setting. I mean, I know my week number, but in the IGS, when I select a particular week number directory, I cannot find my particular day. Can you guide me on this? [image: image.png]

El mié, 28 feb 2024 a las 11:01, Sergei Dolin @.***>) escribió:

@Amjad-58 https://github.com/Amjad-58

Accurate ephemeris files and other files for high-precision methods can be found here: https://igs.bkg.bund.de/root_ftp/IGS/

Precise coordinates for the observation epoch are required. Set up the configuration as in the pictures 4.png (view on web) https://github.com/tomojitakasu/RTKLIB/assets/23137315/5d8c26a8-1479-44ba-bd35-5c82e24d1af3 3.png (view on web) https://github.com/tomojitakasu/RTKLIB/assets/23137315/bb667e8a-33f8-4fee-a97d-76fbc79695ff 2.png (view on web) https://github.com/tomojitakasu/RTKLIB/assets/23137315/11d7a561-825c-45a4-86ed-dafbc3ca10af 1.png (view on web) https://github.com/tomojitakasu/RTKLIB/assets/23137315/17293555-b176-4700-8afe-7066c83dac36

— Reply to this email directly, view it on GitHub https://github.com/tomojitakasu/RTKLIB/issues/740#issuecomment-1968631076, or unsubscribe https://github.com/notifications/unsubscribe-auth/A3OUQA3ZQLBWIXAIFRGK5ULYV357ZAVCNFSM6AAAAABDWS6INCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNRYGYZTCMBXGY . You are receiving this because you were mentioned.Message ID: @.***>

-- Amjad Hussain Doctoral Researcher Deusto Smart Mobility Research Group University of Deusto @. @.>

[image: Deusto] http://www.deusto.es/

@.***>