[aprssig] Position Ambituity in APRS!

[Hans Johnson]

Hi All,

Just my $0.02

I think what's going on here is there isn't a clear distinction between
precision and accuracy.  It's quite possible to have quite a precise
measurement of your location, but that doesn't necessarily mean your
measurement is all that accurate.

For better or worse, it would have been nice if APRS allowed for full
precision numbers plus some way to express one's confidence in the accuracy
of those numbers.  Maybe considering it to be Gaussian, and defining the 1
sigma radius or similar.  Of course, I wouldn't expect anyone to have hand
calculated this in the days of yore, so I can completely understand why
something like this wasn't included in the standard. :)

On the other hand, it would make "position ambiguity" fairly easy, add
random noise to your position, and turn down your accuracy estimation.

V/R,

Hans Johnson

On Jan 9, 2008 3:17 PM, Robert Bruninga <bruninga at usna.edu> wrote:

> > If I have my Kenwood at home
> > (35 57.458N  083 33.430W) set to
> > POS AMB 4, it would transmit 35  .  N\083  .  W.
> >
> > Using your logic, my actual location could
> > (probably would) be outside of your
> > 60 mile circle, right?
>
> The kenwood implementation for transmitting position ambiguity
> is not the recommended way to transmit position ambiguity.  It
> truncates.  So it will transmit a 60 mile ambiguous position at
> 35N and 83W.  But you are actually 57 nautical miles north of
> that position, and so you will just barely be inside the 60 nm
> ambiguity circle.
>
> If Kenwood had rounded instead of truncating, then you would be
> very near to the center of the circle which would have been at
> 36N and 83W.
>
> But even talking about being inside or outside the circle means
> you still do not understand the concept.  The circle is not a
> discrete boundary.  It is a special graphical process intended
> to convey to you that the position is not well known and so it
> could be anywhere in that generatl area within 60 miles or so...
>
>  Talking about positions "within that circle" is meaningless...
>
> Bob
>
> > William McKeehan
> > KI4HDU
> > http://mckeehan.homeip.net
> >
> > On Wed, January 9, 2008 12:36 pm, Robert Bruninga wrote:
> > >>> No, it is preferred to be a circle of the
> > >>> right ambiguity size centered on approximately
> > >>> 35N and 83W.  The word "approximately"
> > >>> is because it should be randomized there.
> > >>> Probably within say half or less of the
> > >>> radius of ambiguity...
> > >>
> > >> If you eliminate the "randomized" portion,
> > >> where would you plot a station that
> > >> gave you 35N and 83W as it's position?
> > >> 3400.00N and 08300.00W ?
> > >
> > > If the guy is saying his best estimate of his position to
> the
> > > nearest whole degree is at 35N I don't see any reason why
> you
> > > would suggest plotting him at 34N.  If he says he is at 35N
> 83W,
> > > then plot him at 35N and 83W with a 60 mile ambiguity
> circle.
> > >
> > > Your software then has to use whatever internal math it
> needs to
> > > be able to activate the right pixels on the screen.  This
> > > includes offsets, scale, mercator projection distrotion
> > > elimination, etc.  But in my code, the entry for for "35"
> > > degrees is just 35.
> > >
> > > Good luck.
> > > Bob
> > >
> > >> >> So if my APRS packet says 35  .  N\083  .  W,
> > >> >> I could actually be at 3467.85N\08272.45W ?
> > >> >
> > >> > Unfortunately, neither of those is a valid posit.
> Minutes
> > > can
> > >> > only be 00 to 59.  But assuming you meant
> > > 3457.85N\08253.45W,
> > >> > then yes, the coordinates of  35N and 83W are closest to
> > > that
> > >> > position.
> > >>
> > >> Yes, I was thinking in base 100, not 60 when I made up
> those
> > > numbers.
> > >>
> > >> [big snip]
> > >>
> > >> --
> > >> William McKeehan
> > >> KI4HDU
> > >> http://mckeehan.homeip.net
> > >>
> > >>
> > >
> > >
> > >
> >
> >
>
>
> _______________________________________________
> aprssig mailing list
> aprssig at lists.tapr.org
> https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig
>



-- 
---------------------------------------------
Hans Johnson (hjohnson at sfu.ca)
B.ASc, Computer Engineering
Simon Fraser University

... Si hoc legere scis numium eruditionis habes. -- Anonymous
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/aprssig_lists.tapr.org/attachments/20080109/37385f56/attachment.html>