Abstract :
[en] For several years, the number of Global Navigation Satellite Systems (GNSS) has been increasing,
opening new perspectives in the field of precise positioning. Particularly, the European system,
Galileo, is experiencing a prompt expansion with the launch, in 2015 and 2016, of 8 satellites
belonging to the new Full Operational Capability (FOC) generation. Broadcasting new signals,
with new modulations, the first studies addressing this system reveal promising level of precisions
on both code and carrier phase observables.
Yet, Galileo is far from being the only GNSS undergoing a noteworthy revolution. Alternatively,
the Chinese program BeiDou, still in a developing phase, as well as the American GPS, currently
undergoing a modernization of its signals, also knew major progress these last two years. Indeed,
7 new satellites have reached the initial BeiDou constellation, bringing to 20 the number of active
satellites. Among them, 5 spacecraft inaugurated the Phase III generation, broadcasting the new
B1, B2 and B3 signals. Regarding GPS, the block IIF, made of L5 signal broadcasting satellites,
kept expanding but at a less steep rate than BeiDou or Galileo.
In this study, we first estimated the individual precisions of each signals broadcast by the aforementioned
GNSS. For this purpose, we created two short baselines of approximatively 5 meters
between similar type receivers. We combined their measurements to form double differences, leaving
in the position equations only multipath and receiver noise. The great expectations regarding
Galileo’s quality turned into affirmations as long as we studied this system. As a matter of a fact,
the code pseudoranges values of the 4 signals of Galileo we have considered (E1, E5a, E5b, E5a+b)
presented outstanding precisions (from 5 to 17 centimetres on code pseudoranges with an elevation
mask of 10 degrees) when compared to GPS (from 12 to 20 centimetres on codes pseudoranges)
and BeiDou (from 26 to 40 centimetres for codes and for phases) in identical conditions.
Even though the precision of Galileo observables is noticeable, the influence of the poor geometry
of the satellite constellation degrades in a significant way the resulting precision of the position
estimated, no matter the recent increase in the number of satellites. Indeed, in this incomplete
constellation, the addition of new satellites results in longer visibility period but not in larger
number of satellites observed at a single epoch. Combining Galileo with GPS or BeiDou is a
way to solve this issue, as the three systems have been designed to be compatible. Therefore,
multi-GNSS relative positioning based on overlapping frequencies should entail better accuracy
and reliability in position estimations. However, the differences between satellite systems induce
inter-system biases (ISBs) inside the multi-GNSS equations of observation.
The overlapping frequencies of these GNSS are the L1 and L5 frequencies of GPS with the E1 and
E5a signals of Galileo, respectively. As far as BeiDou is concerned, the B2 signal of emitted by the
Phase II BeiDou satellites corresponds to the E5b frequency of Galileo. Regarding the new Phase
III satellites, the B2 frequencies will correspond to the Galileo E5a+b signal and B1 of BeiDou will
be compatible with E1 of Galileo and GPS. The combined use of these overlapping frequencies in
zero baseline double differences (ZB DD) based on a unique pivot satellite is employed to resolve
ISBs. This model removes all the satellite- and receiver-dependent error sources by differentiating
and the zero baseline configuration allows atmospheric and multipath effects elimination.
We conducted this study on the L1/E1, L5/E5a, B1(phase II)/E5b overlapping frequencies. Our
receivers were not able to receive the phase III signals of BeiDou satellites. An analysis of the
long-term stability of ISBs (GPS- Galileo and Galileo - BeiDou) was conducted on various pairs
of receivers over large time spans. The possible influence of temperature variations inside the
receivers over ISB values was also investigated. Our study is based on the 6 multi-GNSS receivers
(2 Septentrio PolaRx4, 1 Septentrio PolaRxS, 1 Septentrio PolaRx5 and 2 Trimble NetR9) installed
on the roof of our building in Liege. The estimated ISBs are then used as corrections in the multi-
GNSS observation model and the resulting accuracy of multi-GNSS positioning is compared to
GPS, Galileo and BeiDou standalone solutions.
