MPC OM-20000072 Manual De Usuario

There are two types of differential positioning algorithms: pseudorange and carrier phase. In both of 
these approaches, the “quality” of the positioning solution generally increases with the number of 
satellites which can be simultaneously viewed by both the reference and remote station receivers. As 
well, the quality of the positioning solution increases if the distribution of satellites in the sky is 
favorable; this distribution is quantified by a figure of merit, the Position Dilution of Precision 
(PDOP), which is defined in such a way that the lower the PDOP, the better the solution.

Due to the many different applications for differential positioning systems, two types of position 
solutions are possible. NovAtel’s carrier-phase algorithms can generate both matched and low-latency 
position solutions, while NovAtel’s pseudorange algorithms generate only low-latency solutions. 
These are described below:

1.

The matched position solution is computed at the remote station when the observation in-
formation for a given epoch has arrived from the reference station via the data link. Matched
observation set pairs are observations by both the reference and remote stations which are
matched by time epoch, and contain the same satellites. The matched position solution is
the most accurate one available to the operator of the remote station, but it has an inherent
latency – the sum of time delays between the moment that the reference station makes an
observation and the moment that the differential information is processed at the remote sta-
tion. This latency depends on the computing speed of the reference station receiver, the
rates at which data is transmitted through the various links, and the computing speed of the
remote station; the overall delay is on the order of one second. Furthermore, this position
cannot be computed any more often than the observations are sent from the reference sta-
tion. Typically, the update rate is one solution every two seconds.

2.

The low latency position solution is based on a prediction from the reference station. Instead
of waiting for the observations to arrive from the reference station, a model (based on pre-
vious reference station observations) is used to estimate what the observations will be at a
given time epoch. These estimated reference station observations are combined with actual
measurements taken at the remote station to provide the position solution. Because only the
reference station observations are predicted, the remote station’s dynamics will be accurate-
ly reflected. The latency in this case (the time delay between the moment that a measure-
ment is made by the remote station and the moment that a position is made available) is
determined only by the remote processor’s computational capacity; the overall delay is of
the order of a hundred milliseconds. Low-latency position solutions can be computed more
often than matched position solutions; the update rate can reach 10 solutions per second.
The low-latency positions will be provided for data gaps between matched positions of up
to 30 seconds (for a carrier-phase solution) or 60 seconds (for a pseudorange solution, un-
less adjusted using the DGPSTIMEOUT command). A general guideline for the additional
error incurred due to the extrapolation process is shown in Table 4.

0-2 seconds

1 cm/sec

2-7 seconds

2 cm/sec

7-30 seconds

5 cm/sec