The paper reviews some objectives and preparatory activities of the International GNSS Service (IGS), in relation to the developingEuropean Global Navigation Satellite System Galileo. Experience already acquired in the IGS in monitoring multiple global navigationsatellite systems (GPS and GLONASS) is reviewed. Some signicant features of the Galileo system are outlined, including constellation
ities encompass multiple Global Navigation Satellite Sys-
The next step in this development is the introduction ofGalileo into the IGS activities. A number of preparations
dling a very large global network of sensor stations and
Analysis Centres are working together on the GalileoGeodetic Service Provider (GGSP) Prototype, aimed atdening the reference frame for the Galileo productsand the approach to maintaining that frame over the20-year lifetime of the system. Within the IGS, Galileois a focus of the GNSS Working Groups planning, as
* Corresponding author. Tel.: +49 06151 902272; fax: +49 06151903129.
E-mail address: email@example.com (J.M. Dow).
Advances in Space Research 39tems (GNSS), the IGS Governing Board changed theorganizations name to the International GNSS Service.The acronym IGS, formerly the International GPS Service,was retained. IGSs pilot project using GLONASS(IGLOS) was ocially terminated by the Board in Decem-ber 2005, because GLONASS had been successfully andsuciently incorporated into the routine daily IGS datacollection, dissemination and processing cycles, and ithad become a standard part of the IGS product oering.The IGS had indeed become an International GNSSService.
the complex data analysis required to produce on a rou-tine basis high accuracy, short latency products, the Inter-national GNSS Service has already made a signicantcontribution to the development of the Galileo system.IGS products have been and are being used extensivelyas a reference in the development and verication of algo-rithms and systems, both ground and space. Data fromthe IGS station network was the basis of the Galileo Sys-tem Test Bed, Version 1 (GSTB-V1), and Galileo receiversare currently being deployed at a number of IGS sites forsupport of the GSTB-V2 Mission activities. Several IGSdesign, and some areas are identied in which Galileo will inuence IGS operations in the future and in which the IGS and its activeelements can continue to contribute to new GNSS developments. These include the IGS GNSS Working Group, the Galileo System TestBed GSTB-V2 and the Galileo Geodetic Service Provider (GGSP) Prototype. 2007 Published by Elsevier Ltd on behalf of COSPAR.
Keywords: Galileo; IGS; GNSS; International GNSS Service; GPS; GLONASS; IGLOS; IGEX
In March 2005, in recognition of the fact that its activ-
have been already initiated in this direction. These will bethe subject of this paper.
Through its unique experience and expertise in han-Galileo and the IGS: Taking advant
J.M. Dow a,*, R.E. Neilana ESA/European Space Operat
b IGS Central Bureau, Jet Propulsion Laboratoryc Technical Univer
Received 20 December 2006; received in revi
Abstract0273-1177/$30 2007 Published by Elsevier Ltd on behalf of COSPAR.doi:10.1016/j.asr.2007.04.064ge of multiple GNSS constellations
, R. Weber c, G. Gendt d
Centre, Darmstadt, Germany
lifornia Institute of Technology, Pasadena, USA
, Vienna, Austria
form 17 April 2007; accepted 18 April 2007
(IGEX-98), which had been successfully completed in1998 but had continued in a reduced mode thereafter
acewell as being a signicant element of the IGS StrategicPlan for the years 20082012.
2. International GNSS service
The IGS provides, with a high reliability, accurateGNSS satellite orbit and clock solutions, both in nearreal-time and o-line, the latter with delays of 17 h (rapidproducts) and 11 days (nal products), respectively. Typi-cal accuracies of the rapid and nal products are 3 cmand 0.1 ns for orbits and clocks. The near real-time prod-ucts are based on solutions made every 6 h by combiningsolutions provided by a number of independent AnalysisCentres, each based on pseudo-range and phase data col-lected from a global network of stations tracking the GNSSsatellite constellations. The delay with respect to the latestdata in these solutions is 3 h. Real-time products (orbit andclock corrections) are provided through a 24 h predictionfrom the combination solution with sub-decimetre andnanosecond-level accuracy.
These products are well suited to a wide range of appli-cations, including:
Access to and continued improvement of the Interna-tional Terrestrial Reference Frame (ITRF), throughhighly accurate and continuous measurements ofground station positions and velocities and Earth Orien-tation Parameters (polar motion and length of day).
Monitoring of deformations of the solid Earth. Monitoring of sea level variations. Orbit determination of satellites in low Earth orbit. Continuous mapping of the variations in the electroncontent of the Earths ionosphere.
Weather forecasting and climate research, through mon-itoring of troposphere zenith biases.
A network of some 350 globally distributed ground sta-tions equipped with geodetic quality GNSS antennas andreceivers, of which many are driven by stable atomic clocks(rubidium, caesium, hydrogen masers), provides in nearreal-time the data needed to produce the IGS productsand support these applications. Data are typically retrievedthrough the hierarchy of IGS Data Centres, mainly hourlybut also daily, with 1 s or 30 s observation rates. A signif-icant sub-set of stations is now providing data in real-time,with 1 s sampling rate.
Further background can be obtained from the IGS web-site (http://igscb.jpl.nasa.gov includes the IGS Terms ofReference and current Strategic Plan, as well as networkperformance and much other information); the IGS Anal-ysis Centre Coordinator website http://www.gfz-pots-dam.de/pb1/igsacc/index_igsacc.html; the proceedings ofthe latest IGS Workshop, held in Darmstadt 812 May2006 (Springer et al., 2007; see also http://nng.esoc.e-sa.de/ws2006/programme.html); Dow et al. (2005); and
1546 J.M. Dow et al. / Advances in SpBeutler et al. (1999). Current information on the ITRF isprovided by Altamimi et al. (2005).through the continued involvement of several of the IGEXdata providers and Analysis Centres; see Willis et al.(1999), Slater et al. (2000, 2004), Weber and Springer(2001), Weber et al. (2005). A dedicated laser tracking cam-paign through cooperation with the International LaserRanging Service (ILRS, see Pearlman et al., 2002) and col-laboration with timing laboratories added to the value andrelevance of the IGEX campaign.
The objectives of IGLOS were to:
Set up a global network of GLONASS (or rather, com-bined GLONASS and GPS) receivers, with improvedgeographic distribution compared with the IGEXnetwork.
Produce precise orbit and clock products with a delay ofless than 3 weeks and orbit accuracy of 1020 cm, formonitoring system performance and for geodeticapplications.
Calibrate GLONASS receivers and antennas. Monitor the accuracy and stability of the GLONASSgeodetic reference frame and time system versus thoseof GPS.
The IGLOS project was of intrinsic interest, despite thesmall number of active GLONASS satellites available atthe time, but was also considered a rst step towards incor-poration of additional GNSSs into the IGS processingcycle, looking in particular to the eventual deployment ofGalileo. In the meantime, the GLONASS constellation isbeing built up again with more robust spacecraft; and thereare excellent prospects of a complete constellation of 24satellites again within a few years.
The objectives of the IGLOS Pilot Project were achievedand the project was concluded. The processing of GLON-ASS data in the IGS continues on a routine basis. A graph-ical summary of the orbit accuracy obtained over a periodof several years is given in Fig. 1. It should be noted thatresults of the Russian Mission Control Centre (MCC) arebased on satellite laser ranging (SLR) data only and applyto those few satellites of the GLONASS constellationwhich are regularly tracked by SLR.
4. Galileo orbit constellation
The Galileo spacecraft will y in a Walker constellation27/3/1, i.e. with 27 satellites distributed in 3 orbit planesseparated by 120, with a phase angle of 360/27 = 13.33. GLONASS in IGS
The International GLONASS Service Pilot Project(IGLOS) was initiated by the IGS in February 2000 as acontinuation of the International GLONASS Experiment
Research 39 (2007) 15451551between the positions of the rst satellite in consecutiveplanes. In addition, each plane will have a (normally active)
aceJ.M. Dow et al. / Advances in Spspare satellite, bringing the total number of spacecraft to30.
In order to avoid the gravitational resonance associatedwith a 12 h orbital period, the Galileo satellites will have asemi-major axis of about 29,600 km (altitude some3000 km higher than GPS), and an orbital period of 14 h5 min, giving a repeat of the ground track in about 10 days,corresponding to 17 orbits. The relatively short repeat per-iod is convenient for mission planning purposes. The con-stellation lifetime is 20 years, while individual satelliteshave a design lifetime of 12 years.
The requirements on constellation stability are asfollows:
Distance between consecutive satellites in one plane:variation within 3.
Distance between satellites in adjacent planes: variationwithin 3.
Fig. 1. (a and b) IGLOS (GLONASS)Research 39 (2007) 15451551 1547 Inclination and ascending node: variations within 3.
In order to minimise the need for orbital manoeuvres,small osets are applied to the nominal initial orbital ele-ments of each satellite. In this way the above criteria canbe satised for a period of years much longer than wouldotherwise be possible.
In view of the very long interval to be expectedbetween correction manoeuvres (typically 12 years), it isof interest to use IGS (and ILRS) experience of GLON-ASS and radiation pressure modelling uncertainties toinvestigate the prediction uncertainty of a spacecraft insuch an orbit. This has been carried out by tting reallaser data from an Etalon satellite (in a GLONASS-typeorbit), showing that residuals of only a few km after 12years can be expected. It is clear that, since Etalon is aspherical satellite with a very low area-to-mass ratio, radi-ation pressure modelling for Etalon is signicantly simpler
nal orbit accuracy, till July 2006.
aceand more accurate than it can ever be for Galileo. How-ever, the accuracy obtained in predicting over 12 years isabout a factor 100 better than the Galileo in-orbit posi-tion deadband allows, so that we can be relatively con-dent that the orbit prediction accuracy will indeed besucient to allow a correct denition of the a priori orbitparameters for operation without frequent maintenancemanoeuvres.
Further details about the Galileo orbit selection processcan be found in Zandbergen et al. (2004).
5. IGS GNSS working group
A GNSS Working Group (WG) was established by theIGS in 2003, with the objectives to:
Prepare a consolidated feedback to next generation nav-igation satellite systems (Galileo, GPS, GLONASS)based on relevant experience of providing highest accu-racy products for the existing systems.
Reect on opportunities of upcoming GNSSs to iden-tify improvements in IGS products.
Establish information exchange and stimulate coopera-tion between IGS and entities involved in the manage-ment and technical set-up of Galileo, as well asmodernisation of GPS and GLONASS.
Typical technical tasks were to include study of:
Standard formats (RINEX extension to Version 3, SP3,IONEX, SINEX, real-time, . . .) and their application toGalileo.
How to deal with intra- and inter-system biases. Combined GPS/GLONASS processing. Galileo System Test Bed (GSTB-V2) and In-orbit Vali-dation (IOV): e.g. exploiting the analogy of GPS/Gali-leo-IOV constellation with GPS/GLONASS reducedconstellation scenario for verication of navigationprocessing.
The issue of intra- and inter-system biases becomesmuch more complex for Galileo and future GPS satellites:
Galileo will oer E1, E5a, E5b, E6. GPS will have biases between C1A, P1, P2, L2C/CM,L5-Code.
GPS/Galileo as well as GPS/GLONASS system timeosets will have to be monitored.
Additional frequency-dependent antenna phase centreosets and variations will have to be established andmonitored.
A central question for the WG will be: which biases canbe monitored by the IGS?
In the GSTB-V2 and IOV phases of Galileo, it is
1548 J.M. Dow et al. / Advances in Spexpected that IGS, through its Analysis Centres, cancontribute to improving orbital models, in particularradiation pressure models, and to monitoring the varia-tion of oset between centre of mass and centre ofphase, as well as dierential group delay. Exploringthe challenges of hybrid GNSS data processing duringthe Galileo IOV phase and identifying the possibleimprovements to IGS products will be key activities inthis context.
Galileo will provide phase and pseudo-range measure-ments with lower measurement noise and multi-patherrors (allowing more accurate precise point positioning,for example). Multi-frequency measurement combina-tions will give better integrity for kinematic applications,while the increased number of satellites (up to 60 withGPS and Galileo together) will increase navigationrobustness in case of non-optimal horizon masking andallow derivation of better scientic products (e.g. iono-sphere maps, horizontal gradients for troposphere zenithdelays).
Information on the Galileo system design is provided byOehler et al. (2006), and other contributions in the sameProceedings.
6. Galileo System Test Bed
The Galileo System Test Bed, Version 1 (GSTB-V1) uti-lised GPS signals to simulate critical aspects in a prototypeGalileo ground segment. Central to this experimental oper-ational activity, which was carried out in the NavigationLaboratory at ESA/ESTEC (The Netherlands), was theuse of data from a subset of the global IGS network. Atotal of 41 stations, operated by GeoForschungsZentrumPotsdam (GFZ), the European Space Operations Centre(ESA/ESOC), the Centre National dEtudes Spatiales(CNES) and other IGS contributors, were congured toprovide 1 s data in near real-time (1 h latency). A summaryof the GSTB-V1 system and test results can be found inFalcone et al. (2004).
The Galileo System Test Bed, Version 2 (GSTB-V2)goes one step further, by incorporation of up to 2 Galileotest satellites (Giove-A and in future Giove-B), along witha network of 13 sites equipped with antennas and receiv-ers able to track the Giove signals in addition to those ofGPS; see Piriz et al. (2006). Two European IGS centreswith long experience of operating global GNSS networks(GFZ and ESOC) were entrusted with the establishmentof most of those new tracking sites, as part of a consor-tium led by Galileo Industri...