[teqc] teqc development update ...

Lou Estey lou at unavco.org
Thu Feb 26 09:34:52 MST 2009


All,

A new qc diagnostic will be in the next teqc release allowing for
better "snapshotting" of trouble when examining the qc ASCII time
plot.  The problem is how to present expected epochs where there
is no data from any SV and no position found for the antenna (e.g.
as would result when there are no nav messages from which to compute
SV positions).  Let's take gv01 (and other gv* Galapagos sites) that
we happen to handle here at UNAVCO as an example, for day 2009-054:

ftp://data-out.unavco.org/pub/rinex/qc/2009/054/

The Galapagos network has some sort of data transfer problem such
that a large fraction of BINEX records from all gv* Galapagos sites
can be corrupted, leading to total data loss at many epochs.  Plus
gv01 (and some other gv* Galapagos sites) are not transmitting GPS
nav messages. In the previous versions of teqc, the ASCII time plot
of gv01 looks like this (from
ftp://data-out.unavco.org/pub/rinex/qc/2009/054/gv010540.09S ):

version: teqc  2009Feb11

  SV+--------|---------|--------|--------|--------|--------|--------|--------+ SV
   5|1m                                         ooooooooooooooooooooooooooooo|  5
  21|oo1m                                                     ooooooooooooooo| 21
  30|ooooIc                                         ooooooooooooooooooooooooo| 30
  18|ooooomLm                                        cooooooooooooooooooooooo| 18
  22|ooooooooooom                                               ooooooooooooo| 22
  14|ooooooooooooomIL                                         ooooooooooooooo| 14
  16|oooooooooooooooomm                                                    oo| 16
  31|oooooooooooommoIII                                            oooooooooo| 31
   6|ooooooooooooooooooooom                                        oooLLMI*Im|  6
   3|oooooooooooooooooooooom                                             112o|  3
...

although:

Poss. # of obs epochs   :   8628
Epochs w/ observations  :   5778

i.e. only 67% of the expected epochs are present.  Taking advantage
of some code improvement I made last summer to a function called
qc_missing_epochs() for the alternate qc summary line, the qc of
this site (for the same day's data) now looks like:

version: teqc  2009Feb25

  SV+--------|---------|--------|--------|--------|--------|--------|--------+ SV
   5|1m-                                        -----------------------------|  5
  21|--1m                                                     ---------------| 21
  30|----I-                                         -------------------------| 30
  18|-----mLm                                        ------------------------| 18
  22|-----------m                                               -------------| 22
  14|-------------mIL                                         ---------------| 14
  16|----------------mm                                                    --| 16
  31|------------mm-III                                            ----------| 31
   6|---------------------m                                        ---LLMI-Im|  6
   3|----------------------m-                                            112-|  3
...

i.e. shows (in the new symbol hierarchy) that there are a lot of data holes
in this data -- which is what we want to show. The '-' symbol on the SV
lines in this "qc lite" context means:

-  (lite) missing data epoch(s)

Actually, it's "suspected" missing data epochs since there's really no way
to tell without the nav messages, doing a position, finding out whether a
specific SV should be visible, etc. -- but it appears to work quite well.

This is only a change in the presentation in the qc ASCII time plot.  All
numerical qc parameters are unaffected.

--------------------

Re: http://ls.unavco.org/pipermail/teqc/2009/000790.html
subject: two new intermittent problems with TPS data

Topcon engineers took a look at the files that gave rise to
the problems teqc was having.  For at least the first case
(i.e. apparent rogue value in [RC] messages), nothing was
really wrong with the [RC] messages.  What was wrong was a
missing message, [SI], which gives the tracked satellites
for an epoch, or subsequent epochs if the list is unchanged.
The reason for the occasional missing [SI] message is that
the data was from a rover and a few random messages were
lost in transmission.  When the missing message is the [SI]
satellite list when there's a change in tracking, esp. a
reduction in the number of satellites from a previous set
of epochs, this is a problem.  The Topcon engineers suggested
some things that could be done to improve teqc's veracity
in reading corrupted TPS/JPS data, but my opinion is that
the simple fix I made in teqc seems to be "good enough" for
most cases, and, if users really want the full professional
treatment for problematic data they really should fall back
to Topcon's own translator tps2rinx.exe (for Windows only),
because, quite frankly, Topcon engineers are always going
to be far more familiar with their data than I am ever going
to be.

The second case (i.e. [SI] list show zero satellites)
apparently is allowed by the firmware but doesn't occur
often.  And I was mistaken: previous teqc versions would not
crash when this condition occurred, but teqc would cleanly
exit when this was detected, but report a rather cryptic
message of:

teqc: cannot allocate memory for Topcon CA bin (0 bytes) ... exiting

As I mentioned before, now knowing that a zero satellite
list is a valid state in TPS/JPS, this was easy to fix.

-------------------

Next release: I'm still trying to get the same GLONASS nav messages
collected from different GNSS receivers to vet teqc's reading
of Trimble's new GLONASS nav record, but this test still isn't done
and I have no idea when it might be.  If it isn't done in two
weeks, I'll go ahead and try to get out a new version of teqc
without this vetting.

-------------------

That's it for now ...

cheers,
--lou

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Louis H. Estey, Ph.D.              office:  [+001] 303-381-7456
UNAVCO, 6350 Nautilus Drive           FAX:  [+001] 303-381-7451
Boulder, CO  80301-5554            e-mail:  lou  unavco.org
    WWW:  http://www.unavco.org   http://jules.unavco.org

"If the universe is the answer, what is the question?"
                                                -- Leon Lederman
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


More information about the teqc mailing list