I am using GPS receiver named G STAR-IV Globalsat BU-353-S4 but it gives erroneous position.

How can I get better position in Real Time using any filtering techniques?

I saw some filtering algorithms but most of the algorithms talk about base station and I don't know about base station. Please see the image. The green is the true path and red points are GPS position data.enter image description here

  • You will need to give much more precise information to get a useful answer. How do you know it is erroneous? It is consistently erroneous? What filtering algorithms have you looked at? Etc. – John Powell Sep 19 '16 at 7:20
  • I am not familiar with that specific receiver, but did a quick search and if it is the Globalsat BU-353-S4 that is the first hit, then it looks like it will be a few meters of accuracy (at best) when utilizing SBAS corrections. It doesn't appear that this receiver will utilize any other sort of differential correction. – Johnson5144 Sep 20 '16 at 12:56
  • @Johnson5144 Thank you very for your answer. Yes that is Globalsat BU-353-S4. I am using in Matlab environment. but how can I utilize SBAS? – Learning Sep 21 '16 at 5:17
  • 1
    It looks like there is a utility called GPS Information that installs. There is a toggle switch to utilize WAAS/EGNOS. You will want to be certain that is toggled on in order to receive SBAS corrections. – Johnson5144 Sep 21 '16 at 13:02
  • 1
    Once the receiver is configured to utilize SBAS corrections and sending the NMEA data, that is all that is needed. Unless you unplug the receiver and start over, it would continue to utilize SBAS corrections until reconfigured. This device is not meant for high accuracy positioning, so even with SBAS corrections you will get a few meters of accuracy at best, depending on the environment you are in. – Johnson5144 Sep 22 '16 at 13:19

We don't say "erroneous position" for GPS data, but "innacurate". This is because, being a measurement, it'll always have an amount of error, no matter what one does. What one can do is work around these errors, that is, define an acceptable Margin of Error and try to bring your derived positions to within it.

There is, however, very equipement-specific. There is no simple statistical filtering one can do here (if that is what you're looking for). There are three ways to bring you precision up:

1. Use a better signal: There are two types of signals in GPS - open, civilian-use signals, and encrypted, military-use signals. The open signals give much lower precision than the encrypted ones, and there isn't really much you can do about it if you're using a civilian receptor.

2. Work with phase information: The codes themselves can only take you so far, for the extra mile you have to derive your final position using carrier phase differentiation. This is what the talk about "base station" you read was referring to. This, however, is only an option on survey-grade receivers, which yours isn't. Again, nothing you can do about it.

3. Correct for outside errors sources: There are many sources of error in the GPS system that accumulate. Some of them can be calculated by external stations, and either transmited to your receiver for real-time corrections, or post-processed. These transmissions can be done through other satellites (SBAS) or direct point-to-point radio (GBAS). Of these, your receiver accepts two SBAS providers, WAAS and EGNOS. These, however, will only work at North America and Europe, respectively.

Given your receiver, turning on SBAS is your best approach here. A typical code-only receiver should give about 15m of precision, with SBAS you can expect something like 3m on a good day. If that is good enough for you, then you're set. If not, then you'll need to upgrade your equipment, to a much more expensive one at that. Sadly, it's the limitations of the system.

  1. During my experiment I did mistake without utilizing SBAS correction. Utilizing the SBAS correction we can minimize the error.

  2. Another way is to use google map. Using Google Map We can select some point (e.g., source point, turning point, destination point), then store some positions of the path connecting the selected points. Apply clustering technique to minimize error

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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