goGPS MATLAB version 0.4.3 is available for download.
This is mainly a release to fix a bug that emerged since version 0.4.2: if the rover RINEX file contained approximate coordinates in the header, these were used as approximate coordinates for each processing epoch, also for kinematic datasets. If the rover moved farther away than 30-40 kilometers from its starting position, the least squares solution RMS error exceeded a hard-coded threshold of 4 m, causing the subsequent epochs to be discarded. Please note that previous versions of goGPS already used the RINEX header coordinates in this way, but the 4 m threshold (and the consequent discarding of “bad” epochs) was introduced only since version 0.4.2. Now the rover approximate coordinates are always estimated from code pseudoranges, disregarding the RINEX header coordinates.
This release also serves as a reference for this paper, published on the “GPS Toolbox” column of the GPS Solutions journal: http://link.springer.com/article/10.1007/s10291-015-0469-x
Changelog since v0.4.2:
Added the leap second that will be introduced between the 30th of June and the 1st of July 2015
Various bug-fixes, among which:
– the observation synchronization was not working properly when syncing obs with different time rates
– fixing an error that occurred when reading antennas that missed the azimuth-dependent section
– restored the possibility for the rover to use RINEX header coordinates (if available); now the estimation of approximate coordinates (either from the RINEX ones or from the center of the Earth) is done iteratively until the code point-positioning sigma02 goes below a fixed threshold (now set to (4m)^2), or until the number of iterations reaches a maximum of 5 (whichever comes first)
– the SP3 reading function had not been updated to use the continuous GPS time
Note 1: goGPS binary observation data (*_obs_* and *_eph_*) format was not modified, thus old datasets can be used with the new version without any modification.
Note 2: GUI settings files are also unchanged.