|
Cellular forums Home > Archive > GPS > September 2006 > What's wrong here
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
|
|
|
| Just came back from a walk today and downloaded this track onto my map.
http://img291.imageshack.us/img291/...unter005wc3.jpg
Notice the area along "Christies Creek" were I basically followed the
creek all the way. Under the dense foliage I did get a "weak signal"
error, but notice were the line goes from red to purple and dives up the
side of the hill then back down into the creek. If you lose the signal
wouldn't it naturally just join the last two known points?
Also notice from Waypoint 30 to Waypoint 31 were again it dives down to
the road and then back up to the ridgeline I was following. This area
was more open and I don't think I was losing signal then (although It
could have as it was raining and all the trees were wet)
GPS was a Geko201 downloaded to OziExplorer.
I suppose the basic question is if you lose a signal shouldn't it just
join up the last 2 points, why does it seem to randomly go of at an
angle and then come back again?
Thanks......
| |
| Wayne R. 2006-09-26, 7:33 am |
| This is called "multi-path error".
GPS receivers depend solely on calculated time delay to figure out
where they are. The time a signal takes to get from a satellite to the
receiver tells the thing the distance between the two. The satellites
also know exactly where they are, and pass that info to your receiver.
Put this info together from many satellites and your receiver will
know exactly where it is, give or take a bit. Your receiver knows 1)
where all the satellites are, and 2) how long signals take to get from
each - that's the basis of GPS function.
So, if such signals bounce around the rock walls of natural or
artificial canyons (and especially if direct signal paths are also
blocked), what's received are from longer paths - length causes delay.
Indirect signal paths - due to bouncing - add delay & create a mess.
This delay causes the error you see from your last hike.
On Tue, 26 Sep 2006 06:02:24 GMT, gaza <gaza@hotmail.com> wrote (with
clarity & insight):
>Just came back from a walk today and downloaded this track onto my map.
>http://img291.imageshack.us/img291/...unter005wc3.jpg
>
>Notice the area along "Christies Creek" were I basically followed the
>creek all the way. Under the dense foliage I did get a "weak signal"
>error, but notice were the line goes from red to purple and dives up the
>side of the hill then back down into the creek. If you lose the signal
>wouldn't it naturally just join the last two known points?
>
>Also notice from Waypoint 30 to Waypoint 31 were again it dives down to
>the road and then back up to the ridgeline I was following. This area
>was more open and I don't think I was losing signal then (although It
>could have as it was raining and all the trees were wet)
>
>GPS was a Geko201 downloaded to OziExplorer.
>
>I suppose the basic question is if you lose a signal shouldn't it just
>join up the last 2 points, why does it seem to randomly go of at an
>angle and then come back again?
>Thanks......
| |
| Pieter Litchfield 2006-09-26, 7:33 am |
| I agree completely with the problem description below. In addition, wet
leaves are perhaps one of the best signal blockers around - it doesn't take
a dense forest to significantly attenuate the signal. This may cause the
loss of signal from a satellite that was providing a geometrically superior
solution in favor of one that was providing a less accurate solution. The
net result is an increase in the size of the errors encountered under these
conditions. In response (I hope) to your last question, the GPS is taking a
position fix every second (or so), and always "connects the dots." If it
looses signal completely for several seconds, it will connect to the last
known point. Many GPSs have some "avereraging" function bulit in which may
tend to continue the path travelled if a siganl is momentarily lost. It is
quite possible that with canopy and/or multipath errors, the track would
show a significant deviation from what you know is the path walked. The GPS
has no way of "knowing" that the erroneous position fixes are wrong, so it
simply continues to connect the dots, producing the bad trail.
There are no easy fixes. I believe some GPSs have better bad signal
rejection circuits than others. Some antenna designs may be more effective
under canopy than others. However, multipath errors are very common in
canyon or highrise building areas.
"Wayne R." <wruffner@KomKast.net> wrote in message
news:g9thh21gkliucub
d5b9rc3mklbfvc1ppj9@
4ax.com...[color=darkred]
> This is called "multi-path error".
>
> GPS receivers depend solely on calculated time delay to figure out
> where they are. The time a signal takes to get from a satellite to the
> receiver tells the thing the distance between the two. The satellites
> also know exactly where they are, and pass that info to your receiver.
> Put this info together from many satellites and your receiver will
> know exactly where it is, give or take a bit. Your receiver knows 1)
> where all the satellites are, and 2) how long signals take to get from
> each - that's the basis of GPS function.
>
> So, if such signals bounce around the rock walls of natural or
> artificial canyons (and especially if direct signal paths are also
> blocked), what's received are from longer paths - length causes delay.
> Indirect signal paths - due to bouncing - add delay & create a mess.
>
> This delay causes the error you see from your last hike.
>
>
> On Tue, 26 Sep 2006 06:02:24 GMT, gaza <gaza@hotmail.com> wrote (with
> clarity & insight):
>
| |
| Martin Bodenstedt 2006-09-26, 10:33 am |
| Wayne R. schrieb:
> GPS receivers depend solely on calculated time delay to figure out
> where they are. The time a signal takes to get from a satellite to the
> receiver tells the thing the distance between the two. The satellites
> also know exactly where they are, and pass that info to your receiver.
Almost.
The relevant information is the difference between signals from
different satellites as seen from the receiver.
The satellites all use highly accurate clocks synchronized to a common base.
Therefore the calculated position of a receiver is the more accurate the
wider the satellites are spaced apart as seen from the receiver (aka
"footprint").
If one or more signals is not received directly but as a reflection only
that of course falsifies the computet position. But modern GPS chip sets
discover this discrepancy and ignore the signal for the current
computation...
--
Martin Bodenstedt
(www.die-bodenstedts.de / www.maboko.de)
| |
| Sam Wormley 2006-09-26, 10:33 am |
| gaza wrote:
> Just came back from a walk today and downloaded this track onto my map.
> http://img291.imageshack.us/img291/...unter005wc3.jpg
>
> Notice the area along "Christies Creek" were I basically followed the
> creek all the way. Under the dense foliage I did get a "weak signal"
> error, but notice were the line goes from red to purple and dives up the
> side of the hill then back down into the creek. If you lose the signal
> wouldn't it naturally just join the last two known points?
>
> Also notice from Waypoint 30 to Waypoint 31 were again it dives down to
> the road and then back up to the ridgeline I was following. This area
> was more open and I don't think I was losing signal then (although It
> could have as it was raining and all the trees were wet)
>
> GPS was a Geko201 downloaded to OziExplorer.
>
> I suppose the basic question is if you lose a signal shouldn't it just
> join up the last 2 points, why does it seem to randomly go of at an
> angle and then come back again?
> Thanks......
Perhaps it goes into 2D mode for lack of signals and is using the
wrong altitude. You have here an excellent example of the weakness
of GPS in areas of signal obstruction and high multipath error.
| |
| Mike Scott 2006-09-26, 10:33 am |
| Sam Wormley wrote:
>...
> Perhaps it goes into 2D mode for lack of signals and is using the
> wrong altitude. You have here an excellent example of the weakness
> of GPS in areas of signal obstruction and high multipath error.
>
Just the sort of environment where in-car GPS boxes are most needed, and
perform worst. But I digress.
I was reminded about a track recorded by my GPS72 out walking. On
analysis, a section of the tracklog showed we were doing 10mph - where
in fact we were having problems with a footpath overgrown with brambles
and nettles, and doing maybe 2mph if we were lucky. Doesn't inspire
confidence.
--
Please use the corrected version of the address below for replies.
Replies to the header address will be junked, as will mail from
various domains listed at www.scottsonline.org.uk
Mike Scott Harlow Essex England.(unet -a-t- scottsonline.org.uk)
| |
| Sam Wormley 2006-09-26, 10:33 am |
| Mike Scott wrote:
> Sam Wormley wrote:
>
> Just the sort of environment where in-car GPS boxes are most needed, and
> perform worst. But I digress.
An excellent digression--the nav system in my Toyota Prius integrates
INS and wants to stay on a road exhibiting good performance with loss
of signal including start up during TTFF, tunnels and other obstruction.
When traveling a road that isn't in the map data base, the displayed
location pointer tries to cling to a road in the data base... till I'm
a few hundred feet in the new direction.
>
> I was reminded about a track recorded by my GPS72 out walking. On
> analysis, a section of the tracklog showed we were doing 10mph - where
> in fact we were having problems with a footpath overgrown with brambles
> and nettles, and doing maybe 2mph if we were lucky. Doesn't inspire
> confidence.
>
| |
| Jack Erbes 2006-09-26, 10:33 am |
| gaza wrote:
> Just came back from a walk today and downloaded this track onto my map.
> http://img291.imageshack.us/img291/...unter005wc3.jpg
>
> Notice the area along "Christies Creek" were I basically followed the
> creek all the way. Under the dense foliage I did get a "weak signal"
> error, but notice were the line goes from red to purple and dives up the
> side of the hill then back down into the creek. If you lose the signal
> wouldn't it naturally just join the last two known points?
>
> Also notice from Waypoint 30 to Waypoint 31 were again it dives down to
> the road and then back up to the ridgeline I was following. This area
> was more open and I don't think I was losing signal then (although It
> could have as it was raining and all the trees were wet)
>
> GPS was a Geko201 downloaded to OziExplorer.
>
> I suppose the basic question is if you lose a signal shouldn't it just
> join up the last 2 points, why does it seem to randomly go of at an
> angle and then come back again?
I think the "wild excursions" that are straight line segments are
representative of your EPE growing much larger and/or you are losing
your fix while, at the same time, track points are not being collected.
What are the options for track point collection on the Geko 201? If you
set the trackpoint collection to distance (instead of time) and at finer
intervals, you'll get quite a different picture. You'll see the loss of
fix straight line segments the same but with more track points. When
the EPE gets larger you'll also see more trackpoints more widely
dispersed. And those may be something closer to your actual movements.
If you can set your track point collection rate so that you use most of
(but do not exceed) your track point limits during a hike, it will give
you a lot more data points.
Jack
--
Jack Erbes in Ellsworth, Maine, USA (jackerbes at adelphia dot net)
(also receiving email at jacker at midmaine dot com)
|
|
|
|
|