62/78/Montana Track Point Repeatabililty

Discussion in 'Mapping & Navigation' started by Countdown, Nov 7, 2013.

  1. atlas cached

    atlas cached OX Ambassador

    Joined:
    Jun 16, 2012
    Oddometer:
    688
    Do you not understand how GPS works? All GPS/GLONASS satellites are in motion, while WAAS/EGNOS satellites are geo-stationary.

    This is why your GPSr is constantly downloading new ephemeris data, which informs the GPSr when and where each satellite will be, and is required before any position can be calculated.

    The WAAS/EGNOS satellites transmit correction data that allow the GPSr to calculate current position more precisely.

    When calculating your position on Earth, the only unknown is your position. The GPSr knows precisely where each satellite is.
    #21
  2. Countdown

    Countdown Long timer

    Joined:
    Oct 20, 2003
    Oddometer:
    6,089
    Location:
    Carson City/Ridgecrest
    Yes since all 8-12 satellites sending data are in motion, they probably are the cause of an error term that varies. I think the calculation uses data from all satellites with a signal at the receiver this could very well appear as a random error that changes in any direction for each calculation resulting the the "jitter" observed. I would think other errors like temperature or atmosphere would not change at such a fast rate.
    #22
  3. atlas cached

    atlas cached OX Ambassador

    Joined:
    Jun 16, 2012
    Oddometer:
    688
    [​IMG]
    #23
  4. mcnut

    mcnut Long timer

    Joined:
    Apr 26, 2006
    Oddometer:
    1,394
    Location:
    Bakersfield CA & Sammamish WA
    Well I would hope so, I certainly have. The sat constellation was designed to be dynamic (ever changing) so that given poor geometry at a specific location, in time it will improve, and conversely good become less good over time.

    That said with a good view of the sky the unit should perform to predicted and published standards. Also a stationary unit (again with a good view of the sky) should not be reporting to the user that is has moved 10s of meters at the same time it is reporting an error +- several meters.

    Bruce
    #24
  5. Countdown

    Countdown Long timer

    Joined:
    Oct 20, 2003
    Oddometer:
    6,089
    Location:
    Carson City/Ridgecrest
    Does any one know what the units dynamic reported accuracy is based on?

    In my example (in most cases) it was only reporting moving less than 10 ft but usually once while stopped it would jump about 20-40 feet. However I was in tall pines with high horizions. I think most false reported movement was within the accuracy but the big one were not.
    #25
  6. mcnut

    mcnut Long timer

    Joined:
    Apr 26, 2006
    Oddometer:
    1,394
    Location:
    Bakersfield CA & Sammamish WA

    Sorry, you asked;

    http://en.wikipedia.org/wiki/Error_analysis_for_the_Global_Positioning_System

    From the wiki:

    "However, the advancement of technology means that today, civilian GPS fixes under a clear view of the sky are on average accurate to about 5 meters (16 ft) horizontally."

    I like to keep it simple: If I know the GPS is stationary, I would deem it less then desirable if the unit is telling me I have moved. Worse yet would be if the GPS is showing a small error (DOP? of less then say 5 meters) at the same time it is showing a change in position of 15 meters (and one likely cause of maps spinning) which I believe some on the Montana thread have reported.

    Jerry, I know you are thinking software and SteveAZ is thinking hardware so it would be nice to be able zero in on this.

    As for your questions regarding track handling I would suggest you look at post #9981 in the "New Garmin Montana" thread. Might be a cross platform issue, another reason I dislike mega-threads.

    Bruce
    #26
  7. Countdown

    Countdown Long timer

    Joined:
    Oct 20, 2003
    Oddometer:
    6,089
    Location:
    Carson City/Ridgecrest
    Jerry
    #27
  8. Countdown

    Countdown Long timer

    Joined:
    Oct 20, 2003
    Oddometer:
    6,089
    Location:
    Carson City/Ridgecrest
    Completed and in depth review of the wikipedia error analysis

    http://en.wikipedia.org/wiki/Error_a...tioning_System

    As suspected this analysis was for static error and did not address much about short term dynamic change in error.

    The error terms that could reasonably have a dynamic aspect are:

    Atmospheric effects both ionosphere and troposphere. Changes due to solar rediation could be rather fast but the paper stated the obviouis; "Humidity also causes a variable delay ---- this effect both is more localized and changes more quickly than ionospheric effects. In high wind conditions, the amount of water in the path could change quite fast.

    Multipath effects. The paper stated; "multipath effects are much less severe in moving vehicles. When the GPS antenna is moving, the false solutions using reflected signals quickly fail to converge." This could be the key to quick changes in the pos error. Although the receiver is stationary, the angle to each satellite does change in 2-5 seconds and could cause a reflected signal to come and go.

    Interference. Some motorcycles do have a lot of EMI.

    In summary, I believe that the displayed accuracy on the Satellite page of the receiver is only the PDOP calculatad by the receiver from known positions. The actual error can be much higher and can be dynamic.

    Given this possible "Jitter" in each pos calculation of a stationary antenna, the question is, does the receiver try to determine if a move actually occured? If so, is there a difference between last generation and current generation units?
    #28
  9. 250senuf

    250senuf Long timer

    Joined:
    Feb 17, 2008
    Oddometer:
    2,491
    Location:
    Nelson area, Kootenai, BC, Canada
  10. wbbnm

    wbbnm Long timer

    Joined:
    Dec 18, 2007
    Oddometer:
    2,457
    Location:
    Abq NM
    On the one hand I don't see how the GPS can be expected to really know if it is stationary or if somebody on foot is looking in a small area for a geocache or something.

    On the other hand the Montana has profiles and maybe it could be smart enough to know that somebody using a motorcycle or automobile profile is not driving around in 20 ft circles.
    #30
  11. Countdown

    Countdown Long timer

    Joined:
    Oct 20, 2003
    Oddometer:
    6,089
    Location:
    Carson City/Ridgecrest
    I think you answered your own question. Obviouisly it can not know, it is up to the requirements writer to specify the algorithum.

    With the exception of geocacheing, hikers and up do not need a new trackpoint for a < 5 meters.

    With Auto, they seem to throw out track points if moving in a straight line. This could be inhanced to add time and detect slow down, stop, and speed up.

    If one size (60/76 seems to work) does not fit all, 2-4 profiles for various applications would work for 99.99% of users.
    #31