First of all I would like to say that I am new to gis-se and as such I will try to comply as much as I can to your policy. That said I am actually doing a little project which as the title says relates to RTK-GPS. I am trying to make use of RTKLib which would be familiar to the people reading this.

Following the paper written by the creators of RTKLib I am replicating their low cost RTK receiver with of course more up-to-date parts such as U-blox LEA-6T instead of the U-blox LEA-4T. However, as I am in the UK I do not have access to a free RTK base station within 20 Km range. I therefore looked for a way of making a base station which from a logical deduction could be a second receiver (fix) acting as a base station (for double differenciating) and connected to the Rover (first moving receiver). Anyway, after a brief research I found this setup that seems to be good enough.

My questions so far are the following:

  • Would that setup work with my receiver?
  • Is it good enough? Can I still achieve cm-level precision?

If you any other source of information or examples that would help get a better grasp of RTK-GPS I would be very appreciative.

During my research, I read about the whole theory and did understand it to an acceptable level. However, I still have one thing that I cannot understand. For RTK, we perform an ambiguity resolution which consists in finding the integer cycles of the wavelenght. There are many techniques (even some working with floats) but why an integer number gives a better precision? Is it because of the PLL's? Because the wavelength for L1 for example is about 19 cm so I guess an integer number would give a precision within 19 cm, is that correct?

  • Welcome to GIS SE! I'm not sure that your question is on-topic for our site but I will leave it to the surveyors here to make the call on that. Something, I'll note is that even if on topic I think your question is too broad because it is asking multiple questions. Perhaps see what others say but if there are issues then I recommend making sure that you focus it down to your single most important question.
    – PolyGeo
    Nov 23, 2014 at 21:36

3 Answers 3


To go back to the basics of GPS positioning, you need to know the distance between the receiver antenna and each of the satellites the receiver is tracking. You need a minimum of 4 satellites to determine your position.

The distance antenna-satellite is equal to a number of full wavelengths plus a partial wavelength.

The receiver can only measure the last wave, or actually the partial wavelength.

The ambiguity is the number of full wavelengths separating the antenna from the satellite. So resolving the ambiguity is determining the number of full wavelengths.


As I understand your question, at 19 cm wave length, you need to measure the phase to get down to cm (or even mm) accurate positioning.

However, the measure phase is actually always of py 2*pi*i where "i" is an integer.

In other words, to measure the actual phase of the wave, you first need to figure which wave needs to measured.


Yes, the M6T, or the newer M8T modules should work.

Yes, the newer modules should work as well or better to get cm-level accuracy out of an RTK setup as in the paper.

The fancier, RTK compatible, modules like the LEA-4T, LEA-6T or the newer NEO-M8T resolve the partial wavelengths/phase information to get sub-wavelength precision out of the signals. There is a coarse 1023 bit 1.023MHz signal transmitted on the 1557MHz L1 frequency that helps resolve ambiguities along a 1023-wavelength signal train. The receiver modules do all the math to solve and produce 'pseudorange' information for each satellite, resolving the integer ambiguity and reporting the phase information.

The RTKLIB library solves the simultaneous equations of position and time of the several satellites, and with the two receiver stations, estimates corrections to the pseudoranges to get cm-scale position solutions between the two receivers.

So you need the software to resolve the ambiguities to not be off by a wavelength, but the phase information along the wavelength and along the still provides sub-wavelength precision.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.