It seems we finally had a chance to test how well the T-RAIM in our GPS receivers worked, and for some reason nobody told me about it until three months later...
This plot shows data from GPS.FREEBSD.DK's clockstats file. X-axis is seconds past midnight UTC, the data covers 18:00Z to 22:00Z. Red points are PPS timestamps as measured by the computer. Green and Blue lines show the T-raim one sigma estimate from the receiver. |
|
Same data, but zoomed in. |
|
Data from GPS.DIX.DK, it uses interrupt timestamps so it fares a lot worse. |
|
Same data, but zoomed in. |