PRE- AND INITIAL PROCESSING STEPS

7.3.3 Getting Started: Pre- & Initial Processing Steps

TIME-TAG SYNCHRONISATION

There are in fact two effects (section 6.3.8):

The synchronisation of two (or more) GPS receivers, so that the assumptions made in modelling of the double-differenced observations are satisfied.
The synchronisation of the observation time-tags to the time system of the satellite ephemerides.

The former condition must be satisfied at the microsecond level, while the latter only at the millisecond level. With today's code-correlating receivers this is easily accomplished. As data is collected by the receiver, an internal "navigation solution" is performed that solves for the receiver clock offset to GPS Time ( _rcj ) using pseudo-range data. Even under conditions of Selective Availability, such a calculation ensures that the receiver clock can be synchronised to GPST to the level of a few microseconds. The problem is, however, that the time-tag interval must be correctly initialised in all receivers, for example collect data on each minute, or even minute only, etc. Mixing receivers of different types may therefore be a problem unless it is absolutely clear that the time-tags (although individually accurate) correspond to the same epoch. This synchronisation is therefore a task of the field procedures. It does not mean that the observations at the receivers must be made within a microsecond of each other!

If time-tag synchronisation is not carried out within the receiver, the pseudo-range data can be processed at a pre-processing step. Unlike the standard navigation solution in which the receiver is assumed to be in motion, and a separate clock offset parameter is estimated each epoch, the pseudo-range solution has the following characteristics (section 7.2.2):

Advantage is taken of the fact that the receiver does not move during an entire observation session. Hence there are very many pseudo-range observations to estimate the three coordinate component parameters. In addition, the coordinate solution provides apriori values for the subsequent phase reduction step.
The receiver clock error, as a time-tag error (in seconds), can be modelled as a polynomial (eqn (7.2-2)):

(7.3-1)

The time-tags are then corrected for this offset, that is:

(7.3-2)

However, the data at this epoch to all the satellites are not corrected for clock error as the estimates of _rcj are not accurate enough to perform the following transformation to the sub-nanosecond accuracy level:

for pseudo-ranges to all satellites i (7.3-3a)

and

for carrier phase to all satellites i (7.3-3b)

Back To Chapter 7 Contents / Next Topic / Previous Topic