DPRG
DPRG List  



[DPRG] Long Haul Way Points

Subject: [DPRG] Long Haul Way Points
From: John Harrison johnh630 at comcast.net
Date: Sat Sep 16 22:30:49 CDT 2006

Hello everyone,

 

On Friday we marked off the 4 locations at Fair Park for the contest using 3
separate GPS receivers.

 

The first receiver was one of Kip’s (I did not get the make or model).  It
is a sports class receiver that is WAAS enabled, but for the measurements on
Friday the WAAS was not enabled.  The measurements were strictly
pseudo-range.

 

The second was a Raven, Invicta 115 (Raven bought the old Star Link company
out of Austin, TX).  This would be considered a mid level receiver.  It is
single frequency, WAAS enabled, Carrier phase tracked.

 

The last was a John Deere, StarFire SF2 (John Deere bought the NavCom
company and leases bandwidth on satellites for their proprietary DGPS
correction service).  It is a dual frequency, WAAS and proprietary DPGS
enabled, Carrier phase tracked receiver.  This is considered a high end
(self contained) receiver.  The next level of accuracy is the RTK receivers,
but they are not self contained.

 

After allowing each of the receivers to acquire their signals for about ½
hour, we started the test at point 1 then moved onto point 2, point 4, and
then point 3.  This sequence appeared to be the best to give the receivers a
chance at maintaining RF lock.  We were only marginally successful.  This is
a very tough area to be trying to rely on GPS for navigation.  There are
many trees to generate RF nulls and reflections, and also a bunch of
concrete walls and side walks to magnify the multi-path problem.  When we
lost a lock during a move, we allowed the receivers to require and settle
out again.  I am defining “loosing a lock” by a switch in the satellite that
the receiver was using for DGPS corrections, not a loss of GPS satellites.
None of the receivers lost GPS lock.  Kip’s was not using a DGPS satellite
so a move did not affect his.  The Invicta 115 never lost the WAAS lock.
The Star Fire switched between 3 separate satellites.  I am assuming it was
using the satellite with the greatest signal/noise ratio.

 

All of the receivers were at ground level (the worse configuration possible)
and were sitting within 2 ft of each other.  Some of the offset errors in
the data can be contributed to the 2 ft distance from each other will the
test was performed.  All receivers were tested concurrently.

 

In general, I was surprised on how well the receivers did.  All of them kept
GPS lock (even at point 3, this is a killer).  The errors are higher than
the published errors that I have seen, but given the terrain, it does not
surprise me.  I would consider the errors we found to be “short duration”
errors.  If we went back on another day and repeated the test the errors
could be much higher, particularly for Kip’s receiver without any DGPS.  I
don’t think that the DGPS corrected receivers would not be that much
different.

 

Remember that this was a static test.  My experience with GPS receivers
suggest that the correlation algorithms in these receivers identify when the
receiver is not moving.  In this state the receivers start filtering more to
lower the error.  Once the receiver is in motion the filter time constant
has to be reduced and the error goes up.  This is just a theory of mine but
it seems to be valid.  I have measured a 0.5 – 1.0 second phase lag between
actual position and reported position.  At 3-4 MPH this is not a big deal,
but at 30-40 MPH this can be significant.

 

The summary of the data follows.  I have a spreadsheet with all of the data
but it is fairly large (8+ MB) so I did not attach it.  If the club has an
ftp sight or some other mechanism to distribute the data I am happy to
provide it.  It is VERY interesting to plot the errors vs. time.  It is
obvious that the signals are time correlated.  The two different summaries
are for different coordinates.  The first summary uses longitude and
latitude.  The angles are in decimal degrees.  The second summary uses
distance in feet.  All distances are relative to the mean location of point
4 (as measured by the Star Fire receiver).  I used this for my reference
point because the Star Fire had the lowest standard deviation of error at
this point (presumably the most accurate fix).

 

An interesting note
 Kip’s receiver varies the precision that it reports the
locations with.  It must monitor the standard deviation of the output and
adjusts accordingly.  At all 4 locations the output has a standard deviation
of zero!!

 

I hope that the formatting of the data survives the Email.  If it gets too
messed up. let me know and I will try some other way of getting the results
to you.

 

John Harrison

 

 

 

Long Haul Way points (summary)


Point 1                                                              

            Receiver           Collection time  Mean Latitude   Standard
Deviation        Mean Longitude            Standard Deviation

                                    (seconds)         (decimal degrees)
(decimal degrees)          (decimal degrees)          (decimal degrees)

 

            Star Fire            771.2    32.77717382      3.6763E-06
96.76192813      1.0390E-06

            Raven   729.6    32.77714548      1.7407E-05        96.76192736
8.5547E-06

            Kip's     759.0    32.77724333      0.0000E+00       96.76198667
0.0000E+00

 

 

 

Point 2                                                              

            Receiver           Collection time  Mean Latitude   Standard
Deviation        Mean Longitude            Standard Deviation

                                    (seconds)         (decimal degrees)
(decimal degrees)          (decimal degrees)          (decimal degrees)

 

            Star Fire            680.4    32.77752958      2.2805E-06
96.76207409      4.9159E-06

            Raven   643.2    32.77752633      6.2034E-06        96.76208473
6.5539E-06

            Kip's     602.0    32.77755333      0.0000E+00       96.76210167
0.0000E+00

 

 

 

Point 3                                                              

            Receiver           Collection time  Mean Latitude   Standard
Deviation        Mean Longitude            Standard Deviation

                                    (seconds)         (decimal degrees)
(decimal degrees)          (decimal degrees)          (decimal degrees)

 

            Star Fire            859.8    32.77767411      1.0043E-05
96.76264638      8.5782E-06

            Raven   775.6    32.77770969      4.6791E-05        96.76264229
1.2261E-05

            Kip's     852.0    32.77772000      0.0000E+00       96.76261167
0.0000E+00

 

 

 

Point 4                                                              

            Receiver           Collection time  Mean Latitude   Standard
Deviation        Mean Longitude            Standard Deviation

                                    (seconds)         (decimal degrees)
(decimal degrees)          (decimal degrees)          (decimal degrees)

 

            Star Fire            707.4    32.77788595      1.7148E-07
96.76246211      7.0455E-07

            Raven   674.0    32.77789382      1.0394E-06        96.76247302
2.3170E-06

            Kip's     753.0    32.77785000      0.0000E+00       96.76239667
0.0000E+00

 

 

 

Long Haul Way points in rectangular coordinates (summary)


Point 1                                                              

            Receiver           Collection time  E-W      Standard Deviation
N-S       Standard Deviation

                                    (seconds)         (feet)    (feet)
(feet)    (feet)

 

            Star Fire            771.2    164.1309           0.3194
-258.6005          1.3350

            Raven   729.6    164.3695           2.6295  -268.8895
6.3209

            Kip's     759.0    146.1391           0.0000  -233.3571
0.0000

 

 

 

Point 2                                                              

            Receiver           Collection time  E-W      Standard Deviation
N-S       Standard Deviation

                                    (seconds)         (feet)    (feet)
(feet)    (feet)

 

            Star Fire            680.4    119.2676           1.5110
-129.4114          0.8281

            Raven   643.2    115.9984           2.0145  -130.5929
2.2527

            Kip's     602.0    110.7913           0.0000  -120.7853
0.0000

 

 

 

Point 3                                                              

            Receiver           Collection time  E-W      Standard Deviation
N-S       Standard Deviation

                                    (seconds)         (feet)    (feet)
(feet)    (feet)

 

            Star Fire            859.8    -56.6373            2.6367
-76.9276            3.6471

            Raven   775.6    -55.3822            3.7687  -64.0062
16.9913

            Kip's     852.0    -45.9688            0.0000  -60.2628
0.0000

 

 

 

Point 4                                                              

            Receiver           Collection time  E-W      Standard Deviation
N-S       Standard Deviation

                                    (seconds)         (feet)    (feet)
(feet)    (feet)

 

            Star Fire            707.4    0.0000  0.2165  0.0000  0.0623

            Raven   674.0    -3.3526 0.7122  2.8582  0.3775

            Kip's     753.0    20.1163 0.0000  -13.0553            0.0000

 

 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.dprg.org/pipermail/dprglist/attachments/20060916/a6dd74a3/attachment.html

More information about the DPRG mailing list