International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 3
Number 3
September 2009
283
1 INTRODUCTION
The slogan “GPS/GLONASS satellite measure-
ments” has become popular recently. Judging by
news provided by the press, including Inside GNSS,
one may get the impression that the GLONASS sys-
tem is making its comeback and the number of ac-
tive satellites in the system is steadily growing. This
was the reason for an increase in our interest in the
joint utilisation of both navigation systems in prac-
tice, especially after the correction of the Russian
reference system with respect to the ITRF system
with only a centimetre shift parameters (September
2007). What remains is the problem of differences in
the time scale, but nothing seems to demonstrate that
this is particularly significant. So, what is it we did
in order to confirm the impact of observations of
GLONASS satellites on the accuracy of GNSS posi-
tioning? We conducted an experiment for which we
selected monthly data (September 2007) from the
BOGO, BOGI, and JOZ2 stations (BOGO and BO-
GI are very close to each other while JOZ2 is at a
distance of approximately 42 km). The observations
were processed using Trimble Total Control soft-
ware as the network of selected points is not vast.
The network of vectors connecting the specified
points was designated using two alternatives. The
first only used GPS observations while the second
applied both systems – GPS and GLONASS. The
quantity and configuration of GLONASS satellites
makes impossible the independent analysis of obser-
vations exclusively from the GLONASS system. In
spite of the placement of successive GLONASS sys-
tem satellites in orbit, the number of active SVS has
not changed as of this day. The Russian’s efforts are
concentrated on replacing the old type satellites with
new ones.
2 THE EXPERIMENT
What was done was a comparison of vector determi-
nations for the GPS and GPS/GLONASS data. Two
types of vectors were considered: long (forty-two
kilometre) and short (one hundred metre) ones. The
vector components, long and short, of the determina-
tions from daily cycles were characterised by a mean
error of 2 mm, and nothing seems to suggest any
change in the values of the vector components or
their accuracy characteristics in terms of both solu-
tions conducted using data exclusively from GPS
and utilising observations made using the two sys-
tems. Figure no. 1 presents changes in the “long”
vector components calculated for daily observation
cycles. The equalised coordinate values for the three
points earmarked for the experiment for solutions
using only GPS data and those using both systems
gave identical results with an accuracy of result
presentation.
Positioning Using GPS and GLONASS Systems
L. Kujawa, J. B. Rogowski & K. Kopańska
Warsaw University of Technology, Warsaw, Poland
ABSTRACT: This paper presents an experiment involving the processing of observations using the GPS and
GPS/GLONASS systems performed at the BOGO, BOGI, and JOZ2 IGS stations. Due to the small number of
GLONASS satellites, the authors failed to receive any significant improvement in positioning accuracy using
GPS and GLONASS observations jointly.
284
Figure 1. Changes in the BOGI-JOZ2 vector components in a 3D system
The lack of any discernable difference in results
for the daily solutions induced us to conduct an
analysis of DOP coefficients, making possible an as-
sessment of the “strength” of the solutions in rela-
tion to satellite numbers and configurations.
3 GDOP (GEOMETRICAL COEFFICIENT)
Figure 2. GDOP for the GLONASS system
Figure 3. GDOP for the GPS system
Figure 4. The common GDOP for GPS/ GLONASS systems
4 HDOP (2D SOLUTIONS) AND VDOP
(HEIGHT)
Figure 5. HDOP coefficient for the GLONASS system
285
Figure 6. HDOP coefficient for the GPS system
Figure 7. The common HDOP coefficient for the GLONASS
and GPS system
Figure 8. VDOP coefficient for the GLONASS system
Figure 9. VDOP coefficient for the GLONASS system
Figure 10. The common VDOP coefficient for both systems
5 PDOP (PRECISION COEFFICIENT FOR
DETERMINATION OF 3D POSITION)
Figure 11. PDOP coefficient for the GLONASS system
Figure 12. PDOP coefficient for the GPS system
Figure 13. The common PDOP coefficient for both systems
286
6 TDOP (TIME DILUTION OF PRECISION)
Figure 14. TDOP coefficient for the GLONASS system
Figure 15. TDOP coefficient for the GPS system
Figure no. 16. The common TDOP coefficient for both systems
7 DOP ANALYSIS SUMMARY
All DOP graphs are strongly correlated – firstly with
the number of observed satellites, and secondly with
their placement in the horizontal hemisphere. They
confirm that 3D or even 2D positioning using only
GLONASS satellites is, in practice, senseless. This
is confirmed by the sample Figure no. 16 which pre-
sents the number of GLONASS satellites visible
over a twenty-four hour period in the vicinity of
Warsaw.
In practice, the use of the GLONASS system in
addition to the GPS system gives very poor results.
The DOP coefficients fall insignificantly, which has
no major impact on accuracy achieved. The GPS
system provides DOP coefficients of a value below
three for the decided bulk of the time.
Figure 16. GLONASS satellite visibility in Warsaw
The only situation in which the Russian satellites
might have a major impact on accuracy would be a
situation in which the GPS satellite was low over the
horizon, while the GLONASS satellites would be
high. The measurement would be improved in such a
case. However, analysis of almanacs for GPS shows
that such situations are very rare and their duration is
very short.
8 CONCLUSIONS
The objective of this experiment was to compare
position determination in the GPS and GLONASS
systems as well as applying combined solutions. It
was demonstrated that:
− As of today, it is difficult to speak of the determi-
nation of position exclusively on the basis of
GLONASS satellites.
− The past year has seen the replacement of old
type satellites with new ones – GLONASS-M –
rather than expansion of the space sector as stated
by Inside GNSS (the number of active satellites
has not changed over the year).
− If one is to believe the promises of the Prime
Minister of Russia (dated from before the crisis),
one can have hopes that by the end of 2009 the
system will be expanded to an operational state.
However, nothing seems to support this premise.
− The compensatory effect of the two systems (al-
beit it is difficult to unequivocally state if this is
not temporary) may be considered a promising
premise for the future, when the number of
GLONASS satellites approaches an operational
level.
