I am currently working with QField on my tablet (Samsung Galaxy Tab S5e) and my smartphone (Samsung Galaxy A40). Since the positioning accuracy on my devices is rather low (absolute best case ~3m on an open field with clear sky), I am looking for lowcost ways to improve it in real time. Due to the high costs, I don't want to use an external GNSS receiver.

I've read that with RTK networks accuracies can be improved to sub-meter levels if you have raw measurements from your GNSS receiver, use a Ntrip-client and have access to correction data. This discussion: GNSS raw measurements on Android: Does real time/post processing Differential techniques possible?, and this article and the corresponding video: https://www.gpsworld.com/how-to-achieve-1-meter-accuracy-in-android/; https://www.youtube.com/watch?v=vywGgSrGODU gave me an overview of the topic, but I am still missing some things:

In the mentioned article/video from 2018 it says that there might be apps in a few years which will offer high-accuracy positioning with your smartphone. Does anyone know of an app that has been developed in the meantime and that achieves sub-meter accuracies with the GNSS-data received from a smartphone?


3 Answers 3


This thread on the GPSTest Google Group discusses this topic: https://groups.google.com/g/gpstest_android/c/uT2fmvu7ktA

That thread lists several apps, although I don't know if any of them have definitively proven sub-meter accuracy:

I've been tracking open-source GNSS apps here:


...but I don't know of any that have proven sub-meter accuracy using integrated GPS.


Submeter accuracy with tablet/phone during regular field use is unlikely anytime soon (2024 perspective).

Generally speaking, GPS accuracy is affected by several factors

  1. GPS receiver precision, based on ability to extract real-time information from the GPS signal "received". This is where the apps mentioned try to do more.

  2. GPS receiver ability to lock in on an adequate number of satellites. This is influenced by antenna and potentially shielding/interference inside the phone/tablet from other components, and of course line of sight from where you are to the sky. And whether the device uses all available satellite constallations (GPS, GLONASS, BEIDOU, Galileo, QZSS).

  3. Signal bounce off obstacles (multipath errors). This is a big deal in built-up areas (bounce from buildings) but also in valleys and in dense forest.

  4. Atmospheric interference, especially from the ionosphere. This can be corrected since it varies slowly with time. However, free solutions "in the field" are limited to SBAS, only in some areas of the world, and with accuracy on the metre rather than submetre level (some say 60cm).

Professional surveyors use paid subscriptions and/or post-processing to cut down the effect of #4 from ~1 metre to centimetres. This also requires that the point be "occupied" for some time for readings to stabilize. Therefore, they make use of a fixed base station at a point with good visibility (#2) and no interference (#3) and then locate specific points of interest using a moving rover in the field, which assesses position relative to the base station.

Absent that, consumer-grade single-device GPS accuracy is primarily limited by #2 and #3, and so incremental improvements in #4 or #1 are of limited value.

I've personally had some success using a cheap (relatively) external GPS like the Garmin GLO 2 with my iphone. The improvement comes not from #1 or #4, but since having a dedicated external device allows me to improve visibility and interference (#2 and #3). I put it on my hat or on a long stick, and on occasion have even put on a rock in a clearing and then on my iphone used the Measure and Theodolite app to find my phone's position relative to the GPS!

If SBAS (and therefore metre-level, maybe 60cm) accuracy is worth it to you, Arrow makes devices that use SBAS. But they are expensive, and you're still stuck with #2 and #3.

If you really need centimetre level, you need surveyor-quality gear and the budget that goes along with it.

All of this explains why efforts on primarily #1 (and maybe sort-of #4) for consumer tablets/phones have tended to wither on the vine...it just isn't the limiting factor, except in rare cases.


The easiest way to get horizontal accuracy down to 2cm use an affordable gnss reciever with RTK correction like the ublox F9P (see https://www.ardusimple.com/ or https://www.sparkfun.com/).

  • 2cm is really unlikely without an expensive (survey or higher-grade Mapping / GIS antenna). 10cm to 30cm, maybe. Better than that, not easily?
    – Trams
    Commented Aug 17, 2023 at 22:47

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.