[teqc] RINEX nav files question.

Lou Estey lou at unavco.org
Tue Nov 30 09:50:59 MST 2010


Raúl and everyone,

>> I'd like to know why when rinexing Leica mdb files with TEQC, the nav
>> file generated is different that nav file generated with a Leica
>> software; more specifically, the differences are in ToW data in some
>> satellites.
>> I have attached two files, one of them generated with Leica software
>> (ALG13060.10n) and the other one generated with TEQC (ALG23060.10n).
>> Please, look at the 15th line in ALG13060.10n and 12th line in
>> ALG13060.10n, for example.
>
> Is the original data in Leica LB2 or MDB format? My notes on Leica formats
> show that the ToW (time of week of the specific nav message transmission)
> is not stored with the nav message. To get an estimate of when this might
> be, I use other time tags in these formats to estimate when the nav message
> was received, assuming that the messages are output as they are collected
> by the receiver. (And the GPS ToW only needs to be to 30 second
> granularity.)
> So it's an approximation based on what's known.
>
> As to what Spider is doing, you'll have to ask Leica.

It looks to me like Leica's Spider software just copies the ToE (time of ephemeris)
directly as the "ToW" value for every GPS nav message.  (This clearly is not correct
from a RINEX specification point-of-view.)

>> The consequence of these differences is that the navigation positions
>> obtained with +eepx flag and a unique observation file are very
>> different, being much more dispersed with nav file from TEQC than nav
>> file from Leica software.
>
> "more dispersed"?

Raúl reported off-line to me:
 > With (Leica-produced) ALGC13060.10n, 95% of the positions is nearer than 7.49 m. to average position
 > With (teqc-produced)  ALGC23060.10n, 95% of the positions is nearer than 85.83 m. to average position
 > I have files that these results differ from 7 m. to 150 m.

So, Leica using the ToE as the ToW somehow leads to better epoch-by-epoch
positioning in the current teqc code than using a more realistic ToW.

Offhand, I'd have to say there's a high probability of some sort of GPS nav message
ordering bug in teqc when doing the antenna point-positioning. It would probably
only show up doing this sort of epoch-by-epoch test.  It seems that using a "wrong"
RINEX GPS nav file where the ToW is forced equal to the ToE (and equal to the ToC ==
"time of clock", which is assumed necessary by teqc, even though this is not, AFAIK,
in the IS-GPS-200) the ordering bugs goes away and the point-positioning ends up
being done better.  (Understand that all this is just an educated guess at this point.)

This will take some digging ...

--lou


More information about the teqc mailing list