Journal of Global Postonng Systems (2004) Vol. 3, No. 1-2: 232-241 A Performance Analyss of Future Global Navgaton Satellte Systems Cedrc Seynat 1, Allson Kealy 2, Kefe Zhang 3 1 GPSat Systems Australa, Sute 1/22 Aberdeen Road, McLeod, Vctora, emal: cedrc.seynat@gpsatsys.com.au, Tel: +61 (0)3 9455 0041 Fax: +61 (0)3 9455 0042 2 Department of Geomatcs, The Unversty Of Melbourne, Vctora, Australa emal: akealy@unmelb.edu.au, Tel: +61 (0)3 8344 6804 Fax: + 61 (0)3 9347 2916 3 School of Mathematcal and Geospatal Scences, RMIT Unversty, Vctora, Australa emal: kefe.zhang@rmt.edu.au, Tel: +61 (0)3 9925 3272 Receved: 15 Nov 2004 / Accepted: 3 Feb 2005 Abstract. For an ncreasng number of applcatons, the performance characterstcs of current generaton Global Navgaton Satellte Systems (GNSS) cannot meet full avalablty, accuracy, relablty, ntegrty and vulnerablty requrements. It s antcpated however that around 2010 the next generaton of GNSS wll offer around one hundred satelltes for postonng and navgaton. Ths ncludes constellatons from the US modernsed Global Postonng System, the Russan Glonass, the European Galleo, the Japanese Quas- Zenth Satellte System and the Chnese Bedou. It s predcted that the performance characterstcs of GNSS wll be sgnfcantly mproved. To maxmse the potental utlty offered by ths ntegrated nfrastructure, ths paper presents an approach adopted n Australa to quantfy the performance mprovements that wll be avalable n the future. It presents the desgn of a GNSS smulaton toolkt developed n Australa and the performance expectatons of future GNSS for a number of mportant applcatons wthn the Asa Pacfc regon. In quantfyng the mprovement n performance realsed by combned systems, ths paper proposes a practcal approach to facltate the development of nnovatve applcatons based on future GNSS. Key words: GPS, Galleo, GNSS, Smulaton 1 Introducton Currently, there are only two satellte navgaton systems n operaton, the Global Postonng System (GPS), and the Russan equvalent Glonass. The GPS sgnal s free but ts avalablty s not guaranteed and currently most users are prepared to accept ths rsk. However, as satellte navgaton becomes a vtal technology across a number of crtcal ndustral sectors, the prospect of, for example, a naton s transport nfrastructure becomng dependent on ths technology s a strategc rsk that most ndustral countres are not wllng to accept. Ths argument ntated the Galleo programme n Europe. Galleo s a Global Navgaton Satellte System (GNSS) and 30 satelltes orbtng the Earth at an alttude of 23,616km (three spares) wll transmt a navgaton sgnal that can be receved almost anywhere, and from whch a recever can determne ts poston and tme. Unlke GPS, Galleo wll also offer a guarantee of servce to users who are wllng to pay for t (e.g. commercal servce CS, and Publc Regulated Servce PRS) n addton to a free sgnal smlar to that of GPS (Open Servce - OS and Safety of Lfe servce - SoL). Galleo wll be avalable to the publc n 2010 (European Commsson, 2003). Despte many techncal dfferences between these three GNSS systems, the commonalty of the carrer frequences they use creates the potental for the future development of an nteroperable GNSS recever, as llustrated n Table 1, whch compares the servces avalable and assocated sgnals both now and at around 2015. Asde from these three core GNSS nfrastructures, a number of addtonal space-based navgaton systems are also under development through varous natonal programmes. Japan s currently developng the Quas- Zenth Satellte System (QZSS), wth three satelltes placed n a specal orbt that maxmses coverage over Japan. The QZSS wll complement other exstng GNSS systems over Japan, but at the same tme, these satelltes wll also be avalable over Australa and the South East Asan regon (Petrovsk 2003). In Chna, the Bedou
Seynat et al.: A Performance Analyss of Future Global Navgaton Satellte Systems 233 navgaton system s also beng developed. Current Bedou satellte navgaton and postonng system conssts two geosynchronous satelltes based on the DFH-3 bus. There shall be four satelltes, two operatonal and two backups upon completon of the system (Chnese Defence Today, 2004). In addton, many augmentatons to GNSS are ether under development, or have been commssoned, or are already n use: Space Based Augmentaton Systems (SBAS) have been bult (or are beng deployed) by the US, Europe, Inda, Japan, and Chna. Ground-Based Augmentaton Systems (GBAS) offer tremendous performance benefts to the avaton sector and have led to the development of the Amercan Wde Area Augmentaton System (WAAS), Local Area Augmentaton Systems (LAAS) and the Australan Ground-Based Regonal Augmentaton System (GRAS). Ths bref overvew of current and future navgaton nfrastructures llustrates the huge potental that exsts for future navgaton and postonng applcatons. The vast majorty of the world wll be users of these exstng systems. The fundamental questons then are: Whch system or systems should a country use? ; How to choose a combnaton of the systems? ; What are the benefts and respectve merts of those systems?. There s no smple answer to these questons, as the best soluton wll undoubtedly depend on the targeted applcaton, whch has ts own requrements n terms of accuracy, relablty, robustness, cost and other applcaton-specfc crtera. What can be provded, however, s a means whereby parameters that descrbe these performance requrements can be computed. Hgh-accuracy software smulatons are a cost-effectve and precse approach of determnng the performance characterstcs attanable from the future GNSS, and have been recognsed as an approprate pre-development tool for satellte navgaton systems and applcatons n Japan (Petrovsk 2003) and Europe (Seynat 2003). The techncal benefts of ths approach le n the fact that the smulatons are reproducble and totally controlled, and parameters can be changed ndvdually f necessary for an n-depth understandng of the underlyng effects. Ths paper ntroduces a smulaton toolkt developed to conduct a qualtatve assessment of the performance characterstcs of the future GNSS nfrastructure. The desgn and development of the smulator archtecture as well as the models descrbng all nfluencng effects on the performance of GNSS are presented. Fnally, representatve results over Australa are demonstrated and future developments are outlne. Table 1: Current and future GPS, Glonass and Galleo servces and sgnals for 2004 and planned for 2015 GPS Glonass GALILEO Servces 2004 2015 2004 2015 2004 2015 Basc Postonng (unencrypted) Integrty/safety (unencrypted) Commercal/value-added (encrypted) Securty/mltary (unencrypted) SPS CA PPS P(Y) L2 P(Y) SPS CA L2C L5 PPS P(Y) L2 P(Y) M L2 M SP HP L2 SP L2 3 rd Sgnal Integrty message HP L2 Unknown OS E5a E5a SoL E5b E5a CS E6 PRS E6 SPS-standard postonng servce, PPS precse poston servce, SP standard precson, HP-hgh precson, OS open servce, SoL safety of lfe servce, CS commercal servce, PRS publc regulated servce 2 Smulaton technology and methodology Current and future users of navgaton technologes need to understand and quantfy the performance they can expect from the canddate systems, used ndvdually or n combnatons. The GNSS Smulaton Tool (GST) developed n ths research ams to provde a toolkt that reproduces the performance behavour of exstng and planned GNSS, n order to support the development of
234 Journal of Global Postonng Systems next generaton navgaton applcatons. The man objectves of the GST are: To provde an accurate, ndependent tool capable of analysng customsed applcaton scenaros, based on realstc models of all effects relevant to the performance of an applcaton of the satellte navgaton technology; To produce precse techncal data on the performance of navgaton applcatons n the form of predefned or custom-defned navgaton scenaros; To generate smulated navgaton data as a real-world navgaton recever would capture, n a format famlar to applcaton developers, n support of algorthm development and testng; and To allow external programs and fle formats to connect to the GST, n order to maxmse the re-use of exstng expertse and data sharng. To acheve these objectves, at the core of the GST s a set of models relevant to the descrpton of navgaton systems. These models are desgned to be fully confgurable by the user. The GST also has a set of data analyss tools and external fle readers that ether ntalse some of the GST model parameters or fully replace a model, dependng on the specfc fle used. In addton, the GST can also be confgured to use models developed externally by thrd partes. These external models can take the form of a Wndows DLL or an executable (.exe ) fle. orbtal parameters of the GPS constellaton, as provded n the YUMA format from the Internet (USCGNC, 2004). Sgnal-n-space: The satellte sgnal s defned n the GST by a set of sgnal characterstcs: The frequency f of the carrer; The modulaton scheme: Bnary Phase Shft Keyng (BPSK) or Bnary Offset Carrer (BOC); and In case of the BOC modulaton, the two ntegers n and m, multples of the base frequency 1.023MHz, defnng the sub-carrer frequency fs and the code rate fc, Navgaton message: The navgaton message sent by the satellte s not modelled n the GST. Ths means that the satellte ephemers and clock are not transmtted to the recever model and therefore not used n the poston computaton of the recever. However the ephemers error s represented n the GST as an error of δ. eph Fles readers Inputs Models and algorthms Data analyss and algorthms Fgure 2: Functonal dagram of the GST software archtecture 3.2 Envronment models Fgure 1: Functonal dagram of the GST software archtecture 3.1 Space Segment External models Output and vsualsaton Orbts: The GST computaton of the satelltes coordnates s based on the standard Kepleran orbt parameters. Satellte postons n the GST can also be mported from actual data provded from real GPS satelltes, such as those provded on a daly bass by the Internatonal GPS Servce (IGS, 2004). The orbt fles are generated usng the standard sp3 format, and the GST contans a sp3 fle reader. The values from the sp3 fles can be nput n the GST n place of the computed Kepleran coordnates. Alternatvely, the GST can use The term envronment refers here to the propagaton medum through whch the satellte sgnals travel before reachng the recever antenna. The man effects consdered are the transmsson delay due to onospherc effects, tropospherc effects and multpath. Ionosphere: The transmsson delay nferred by the onosphere s based on the knowledge of the total electron content (TEC) along the transmsson path of the sgnal (Parknson 1996). In order to obtan an accurate estmate of the TEC, two approaches have been mplemented n the GST: The frst mplementaton uses a global onospherc model, NeQuck (Letnger 1996), as an external model plugged-n the GST archtecture. The second mplementaton uses measured data of the vertcal Total Electron Content (vtec) regularly avalable from publc nternet sources and publshed n the Ionospherc Map Exchange (IONEX) format.
Seynat et al.: A Performance Analyss of Future Global Navgaton Satellte Systems 235 The two mplementatons descrbed above for modellng of the onospherc effect on the sgnal transmsson are complementary rather than exclusve. Whle TEC maps provded by the IONEX fles are more representatve of the actual stuaton that one wants to smulate wth the GST on a specfc date, tme, and locaton, the NeQuck model, s more generc and can be used to smulate a data and tme set n the future. Troposphere: The sgnal transmsson delay due to tropospherc effects s modelled usng the refracton ndex of the troposphere (D Govann 1990). The Hopfeld model s used to estmate the refracton ndex (Hofmann 1998). The model s dependent on the estmates of local temperature, atmospherc pressure and water vapour partal pressure. In the GST these values can be set manually or alternatvely, the GST can read meteorologcal data collected at ground reference statons and output n the Recever Independent Exchange (RINEX) format (Gurtner 2002). Multpath: Multpath s hghly dependent on the local envronment surroundng the navgaton recever. Desgned to be a generc tool, the GST models the multpath effect as ts end effect on the measured range. Multpath causes a range measurement error, whch can be solated from other error sources n actual recever measurements. The model used n the GST s an emprcal model based on observatons of tme seres of multpath range error (Parknson 1996). The range error caused by multpath,, s modelled as follows: δ δ mult ( η ) () t = b + a n ()cos t k () t (1) mult mult mult c where b mult s the multpath bas error, a mult s the ampltude of the multpath error, η s the satellte elevaton angle at the recever locaton, k s a factor requred to adjust the elevaton dependence, and n c (t) s correlated Gaussan nose. The model n Equaton (1) s approprate to represent the range error caused by the multpath for several reasons. Frst, the elevaton dependence s modelled n such a way that satellte at a low elevaton causes hgher multpath errors, as t s observed from actual measurements. The factor k allows adjustng the peak of ths elevaton dependence. Secondly, the use of correlated Gaussan nose also represents effects seen n actual observatons. Typcal correlaton tmes observed n multpath effects are of the order of a few mnutes. To smplfy the tunng of ths model, the GST proposes several sets of default values correspondng to low, medum, and strong multpath envronments. Expert users are also allowed to change parameters separately f they wsh. 3.3 User Segment Recever coordnates: Recever coordnates are nput n the GST as a lattude, longtude, and heght n WGS-84. Maskng angle: The GST allows the defnton of a customsed maskng profle to represent the specfc stuaton at a partcular locaton. Recever carrer-to-nose rato: The total sgnal carrerto-nose rato s a measure of the sgnal qualty and nfluences the trackng error. It s modelled as the sum of the sgnal-to-nose rato where no jammng s present, plus the jammer-to nose rato. Recever trackng error: The modellng of the accuracy of the range measurement n the GST assumes a Non- Coherent Early-Late Processng. Two models are mplemented, one for the GPS BPSK modulaton scheme (Kaplan 1996) and one for the Galleo BOC scheme (Betz 2000). Recever poston: The recever model computes an estmaton of ts poston based on measured range between tself and each satellte n vew. In the current mplementaton of the GST, the measured range s modelled as the true satellte-to-recever range plus the sum of the errors descrbed n the prevous sectons. Then the poston s estmated usng a weghted least square algorthm. User Equvalent Range Error: The models descrbed n the prevous sectons account precsely for the dfferent error sources affectng the navgaton soluton. An alternatve, less accurate, modellng of these errors s commonly used when smulatons need to be run for long smulated perods and over large geographcal areas. In ths alternatve method, the errors are grouped n a User Equvalent Range Error (UERE). In the GST the UERE for each satellte s a functon of ts elevaton. Publshed UERE values for GPS and Galleo were used (Shaw 2000, Benedcto 2000). 3.4 GST predefned fgures of merts To descrbe the output of the GST a number of standard performance parameters are computed. Avalablty of navgaton soluton: The recever can make an estmate of ts poston and clock bas when at least 4 satelltes are n vew. The nstantaneous avalablty of the navgaton soluton anav () t s a flag set to 1 f at least four satelltes are vsble at a tme t, and set to 0 otherwse: The avalablty of the navgaton soluton n a tme wndow s the percentage a () 1 of the tme, n the tme wndow, when nav t =.
236 Journal of Global Postonng Systems Contnuty of navgaton soluton: The contnuty of the navgaton soluton s a quanttatve estmate whether the navgaton soluton can be computed wthout nterrupton. The nstantaneous contnuty of the cnav navgaton soluton, () t, s a flag set to 1 f the navgaton soluton s avalable at the tme t and tme t- DT, and set to 0 otherwse. The contnuty of the navgaton soluton n a tme wndow s the percentage c () 1 of the tme n the tme wndow when nav t =. Avalablty of Accuracy: The avalablty of the accuracy expresses whether the poston estmate made by the recever s accurate enough to be used for a partcular applcaton. Dfferent applcatons have dfferent accuracy requrements, and f the poston computed by the recever s not accurate enough for a gven applcaton to be successfully carred out, then satellte navgaton s not approprate. The nstantaneous avalablty of accuracy, aacc () t, s a flag set to 1 f the postonng error of the recever s less than a user-defned threshold, and set to 0 otherwse. The avalablty of the accuracy n a tme wndow s the percentage of the tme when aacc() t = 1. Contnuty of accuracy: The contnuty of the accuracy s equvalent to the contnuty of navgaton soluton defned prevously, usng the avalablty of accuracy to estmate contnuty. Dluton of precson: The Dluton Of Precson (DOP) s a well-known ndcator of the geometry of the satelltes n vew from a recever. Poston error: The nstantaneous poston error s the scalar dfference between the recever true poston and the poston estmated by the sngle pont postonng algorthm of the recever model. 3.5 Model Valdaton Each of the models above has been carefully tested ndvdually. Also, the GST results have been valdated by ndependent sources. It s beyond the scope of ths paper to present a detaled valdaton report of the models. Valdaton has been carred out n the followng areas: Recever and satellte geometry: coordnates, lne-ofsght vectors, and DOP values were compared aganst those obtaned from a separate software tool (Satellte ToolKt) Recever RAIM avalablty fgures were compared aganst a servce volume smulator developed n Europe (Lozou 2002). World and Europe performance maps publshed n (Lachapelle 2002) and (Leonard 2003) were successfully reproduced wth the GST The smulated recever pseudorange and carrer measurements were nput n the GPS/Glonass/SBAS data Toolkt Teqc (UNAVCO 2004) and successfully tested for qualty. Other comparsons have been conducted and further nvestgaton s currently beng conducted by the authors. 4 Smulaton Results Ths secton presents a sample of the typcal smulaton results obtaned from the GST spannng across the entre sute of the models descrbed earler. The am of ths secton s to llustrate the capabltes of the GST n smulatons focused on Australa. The results presented here cover a wde range of topcs and ths wll be the frst nvestgaton undertaken n Australa. 4.1 Avalablty of Navgaton Soluton over Australa The global avalablty of the navgaton soluton over Australa s dependent upon the number of satelltes n vew and the mask angle. Smulatons were run at a 40 mask angle for the Galleo constellaton only, the GPS constellaton only, and for the combned GPS+Galleo constellaton. Smulatons of 24 hours were run. At a 40 mask angle, both the GPS and Galleo constellaton when used alone offers reduced avalablty of the navgaton soluton. The GPS constellaton s unusable most of the tme (about 20% avalablty on average), and the Galleo constellaton offers about 60% avalablty almost everywhere except n the northernmost regon of the country. The clear lattude dependence shown n Fgure 3(b) s due to the sphercal symmetry of the Galleo constellaton wth respect to the Earth s centre. The advantage of usng a combned GPS and Galleo constellaton s clearly shown on Fgure 3, as the avalablty of the navgaton soluton remans 100% globally.
Seynat et al.: A Performance Analyss of Future Global Navgaton Satellte Systems 237 (a) GPS (a) Galleo (b) Galleo (b) GPS (c) Galleo+GPS Fgure 3 - Avalablty of Navgaton Soluton over Australa for a 40 mask angle (c) Galleo+GPS Fgure 4-95% Percentle accuracy for a dual-frequency (a) L1/E5a Galleo recever, (b) L1/L5 GPS recever and (c) combned Galleo/GPS recever at the same 2 frequences
238 Journal of Global Postonng Systems 4.2 Postonng Accuracy The postonng accuracy presented n ths secton was obtaned by usng the GST n Servce Volume mode. The UERE budget used for ths purpose was a dual frequency L1/E5a Galleo recever and dual frequency L1/L5 GPS recever. Of course such recevers do not exst yet but ther expected performance was presented n (Shaw, 2000; Benedcto, 2000) and are used here. A constant mask angle of 20 was used, for a smulaton tme of 24 hours. Fgure 4 shows that the 95% accuracy to be expected from Galleo or GPS dual-frequency recevers s of the same order of magntude (about 6m across the country). The combned performance of Galleo and GPS s expected to be of about 2-3m, whch s a sgnfcant mprovement compared to each ndvdual system The effect of usng a combned GPS/Galleo recever s also to remove geographcal dsparty across the country. The tme traces of postonng accuracy (not dsplayed here) are also more consstent, wth less fluctuaton. Ths effect of combned use s as mportant as the absolute gan n accuracy tself. Users usng Galleo/GPS recevers can expect a consstent performance of ther system, whch may reduce the need for costly augmentatons n areas and at tmes of the day where ndvdual systems fal to delver the requred performance level. 4.3 Impact of QZSS on navgaton performance n Australa The QZSS has the prmary objectve of augmentng GPS over Japan so that satellte avalablty n dense urban areas lke Tokyo remans hgh. The orbt of the QZSS satelltes s such that there wll always be at least one, and often 2 satelltes n vew and at hgh elevaton from recevers located n Japan. However, the QZSS satelltes also pass above Australa, and for that reason they offer the potental of augmentng GPS n Australa as well. Ths potental s assessed here wth the GST. The desgn of the QZSS constellaton orbts s presented by Petrovsk (2003), wth several optons beng consdered ncludng an 8-shape orbt used n subsequent GST smulaton shown here. The QZSS satelltes wll pass above or near Australa. It s therefore possble to envsage that those satelltes can be used n Australa, n the same way as they wll be used n Japan,.e. as an augmentaton to GPS. The number of vsble QZSS satelltes and ther elevaton s presented here for four Australan state captals, namely Melbourne, Sydney, Darwn and Perth. These locatons have been chosen because they are wdespread across the country and because the benefts of usng QZSS n urban areas wll be llustrated later, partcularly n low vsblty stuatons that can occur n Melbourne or Sydney. Fgure 6 ndcates that QZSS satelltes n an 8-shape orbt wll be well vsble from Australa. In Melbourne and Sydney, at least one satellte wll be vsble above 60 at all tmes. In Darwn, all 3 satelltes wll be vsble at all tmes from locatons wth mask angles lower than 25. The fgures from Perth are also excellent. Fgure 5 shows that QZSS s partcularly relevant to Australan GNSS users. In areas of reduced vsblty, the avalablty of a sngle addtonal satellte n all tmes can make a dfference between gettng a poston fx or not. At least one QZSS satellte wll be almost always n vew at any tme, even n urban canyons. Wth ths prelmnary result based on geometry, t appears that QZSS usage n Australa should be nvestgated further, at both techncal and poltcal levels. To llustrate the navgaton mprovement resulted from the combned use of GPS and QZSS n Australa, the smulaton results n an urban area s presented n Fgure 6, for the locaton of Melbourne, and a mask profle representng an urban canyon. Fgure 6 shows the avalablty and contnuty of the accuracy n an urban envronment at the locaton of Melbourne for the GPS only and combned GPS/QZSS cases. The smulaton uses a sngle frequency recever and the plot dsplayed here s relevant to mass market applcatons, such as personal moblty. The gan n avalablty and contnuty of accuracy s clear from these plots, although for an accuracy threshold of 12m, 100% avalablty and contnuty s not acheved at all tmes even wth combned QZSS+GPS recevers. Addtonal systems, such as the addton of Galleo or pseudoltes, may be requred dependng on the applcaton consdered. Investgaton nto pseudolte networks n urban envronments s a potental feld of study that can be made wth the GST. Overall, the results presented n the prevous three sectons show the sgnfcant mprovement n navgaton performance to be expected from a recever capable of trackng several ndependent systems. More than n postonng accuracy, the mprovement s n the consstency of the performance and therefore n the relablty of the applcaton from the end-user perspectve. Ths s an mportant consderaton from a marketng perspectve, as users are more lkely to adopt the new technology f ts relablty s certfed 4.4 Raw data generaton for post-processng by thrd party software Based on ts accurate models and ts end-to-end capablty, the GST can be used to generate smulated measurement as they would be receved by a real recever. The GST uses the wdely accepted RINEX 2.1 format to output smulated data. The GST generates the RINEX observaton and navgaton fles. Part of the
Seynat et al.: A Performance Analyss of Future Global Navgaton Satellte Systems 239 valdaton exercse outlned n Secton 3.5 was to produce RINEX fles and nput them n the TEQC software for qualty check. More generally, the avalablty of smulated Galleo and GPS data opens the possblty for research organsatons to start testng algorthms and applcatons early, n readness for the future avalablty of the hardware. Ths approach saves development and testng effort and can be acheved usng the GST. Such capabltes wll be llustrated n further publcatons. (a) Melbourne (b)sydney (c)darwn Fgure 5 8-shape QZSS satellte elevaton (left) and number of vsble satelltes (rght) at Melbourne, Sydney and Darwn n Australa
240 Journal of Global Postonng Systems Fgure 6 Avalablty of Accuracy and Contnuty of Accuracy from a recever n an urban area at the locaton of Melbourne, obtaned from an accuracy threshold of 12m 5 Conclusons Ths paper presents the GNSS Smulaton Tool, a software tool of navgaton systems that ams to provde a means for research organsatons and ndustry to develop navgaton applcatons usng the current and emergng technologes. The reason to buld such a tool today orgnates from the fact that navgaton technology, and especally satellte navgaton, s now at a major crossroad: new satellte systems are beng bult n Europe, Japan, Inda, Chna and the Unted States. Current systems are beng mantaned, augmented and rapdly mproved, n Russa and the Unted States. Addng to the advent of these new technologes s the boomng need n today s socety for postonng nformaton. It s the dual growth of the technology and the market that makes the future of navgaton multdscplnary and challengng. The current GST tself cannot handle fully the complexty and varety of the challenges to come n the near future. However, t provdes a low-cost, flexble tool to provde specfc answers to a number of performance related questons. The potental areas of further development for the GST are numerous. One of the most promsng applcatons s the generaton of raw pseudorange and carrer phase measurements, formatted n RINEX 2.1, to be drectly usable for algorthm development and applcaton certfcaton. The capablty to generate RINEX fles s already presented n the GST, but further models and valdaton can be added, such as the possblty to smulate satellte outages, or the random ntroducton of corrupt measurements, that a real recever would experence. Also, the Galleo RINEX format, when t
Seynat et al.: A Performance Analyss of Future Global Navgaton Satellte Systems 241 becomes avalable, can be easly ncorporated nto the GST. For applcaton development, the ntroducton of pseudoltes n the smulaton would be a useful addton to the tool. From the modellng perspectve, onosphere scntllaton, troposphere rregulartes (e.g. ranfall), clock errors, and refned recever algorthms are other nterestng developments currently beng consdered. The smulaton results presented here are just a subset of the entre sute of outputs that the GST can provde. They have shown the quanttatve benefts of complementng GPS wth another system such as Galleo. Future avalablty of complementary systems s partcularly relevant to the South-East Asa regons and countres such as Australa are n a prvleged poston to develop nnovatve applcatons based on an nteroperable use of future GNSS. 6 Acknowledgements GST valdaton uses the servce volume smulaton tool developed by VEGA Plc and kndly provded by John Lozou. John s thanked for provdng the tool and hs nsghtful comments on valdaton and smulaton. Partal fnancal support from Vctoran Partnershp for Advanced Computng (VPAC) and Corporatve Research Centre for Mcro-technology endorsed to A/Prof Zhang s hghly acknowledged. References Barnes J, Rzos C, Wang J, Small D, Vogt G, and Gambale N (2004) Hgh precson ndoor and outdoor postonng usng LocataNet, Journal of Global Postonng Systems, 2(2):73-82 Bartone CG (1996) Advanced pseudolte for dual-use precson approach applcatons, Proceedngs of 9th Int. Tech. Meetng of the Satellte Dvson of the U.S. Inst. of Navgaton, Kansas Cty, Mssour, 17-20 Sept., 95-105 Benedcto J., Dnwddy S., Gatt G., Lucas R., Lugert M., Galleo: Satellte System Desgn and Technology Developments, European Space Agency, November 2000, http://ravel.esrn.esa.t/docs/galleo_world_paper_dec_200 0.pdf. Chnese Defence Today (2004) BD-1 Navgaton Satellte, http://www.snodefence.com/space/ spacecraft/bd1.asp. D Govann G., Radcella S., An analytcal model of the electron densty profle n the onosphere, Advances n Space Research, Volume 10, 1990. European Comsson (2003), The Galle Project, Galleo Desgn Consoldaton, http://europa.eu.nt/comm/dgs/ energy_transport/galleo/doc/galle_brochure.pdf. Featherstone WE, Krby JF, Zhang KF, Kearsley AHW, Gllland JR (1997) The quest for a new Australan gravmetrc geod, n: Segawa J, Fujmoto H, Okubo S (eds), Gravty, Geod and Marne Geodesy, Sprnger, Berln, 581-588 Gurtner W., RINEX: The Recever Independent Exchange Format Verson 2.10, ftp//gscb.jpl.nasa.gov/gscb/ data/format/ (2002) Hofmann-Wellenhof B., Lchtenegger H., Collns J., GPS Theory and Practce, Sprnger-Verlag, 1998. IGS webste: http://gscb.jpl.nasa.gov (last vsted November 2004) Kaplan E., Understandng GPS Prncples, Artech House, 1996. Lachapelle G., Cannon M., O Keefe K., Alves P., How wll Galleo Improve Postonng Performance? GPS World, September 2002, 38-48. Letnger R., Radcella S.(1996), NeQuck Ionospherc Model, Software Documentaton, http://www.tu.nt/ ITUR/software/study-groups/rsg3/databanks/onosph/ Rec531/NeQuck_ITUR_software.pdf. Leonard A., Krag H., Lachapelle G., O'Keefe K., Huth C., Seynat C., Impact of GPS and Galleo Orbtal Plane Drfts on Interoperablty Performance Parameters, GNSS 2003 Conference, Graz, Austra, 22-25 Aprl 2003. Lozou J. Servce Volume Tool for RAIM computaton, Prvate Communcaton. Parknson B., Splker J. (Edtors), Global Postonng System: Theory and Applcatons, Volume 1, Progress n Astronautcs and Aeronautcs Volume 163, 1996. Petrovsk I. (2003), QZSS - Japan's New Integrated Communcaton and Postonng Servce for Moble Users, GPS world, June 2003 ssue. Seynat C., Pdgeon A., and Lozou J. (2003), Combned use of Galleo and GPS: towards nnovatve navgaton research, Proceedngs of SatNav2003, the 6th Internatonal Symposum on Satellte Navgaton Technology Includng Moble Postonng & Locaton Servces, Melbourne, Australa, 22-25 July 2003 Shaw M., Sandhoo K., Turner D., GPS Modernzaton, Proceedngs of The Royal Insttute of Navgaton GNSS-2000, Ednburgh, Scotland, May 2000. UNAVCO Faclty, Boulder, Colorado, Teqc homepage, http://www.unavco.org/faclty/software/teqc/teqc.html (last vsted November 2004) Unted States Coast Guard Navgaton Center, GPS Almanac Informaton, http://www.navcen.uscg.gov/gps/almanacs. htm (last vsted November 2004)