Fullscreen HTML5Med Res (??MB)HTML4

Accuracy down to centimetres and even millimetres . . . Augmented GPS Everyone knows how effective – and accurate – today’s Global Navigation Satellite Systems (GNSS) are. But it wasn’t always so – and even the ~5m typical accuracy of a modern GPS is nowhere near good enough for many of today’s more demanding tasks, such as landing an aircraft, controlling driverless cars or monitoring earth movements and tides. That requires a whole new approach, called “augmentation”. T he typical GPS navigation error is actually amaz- in order of decreasing severity, are: ingly good if you consider how large the Earth is and • drift of the satellite clocks, how far above you the satellites are orbiting (even • deflection and delay of the satellite signals in the ionthough they are in low-earth orbit). But clearly, it could osphere, be a lot better. • instability in the clock of the receiving device, Many applications require much better accuracy, in some • uncertainty of the satellite orbit, cases down to the centimetre level. That includes preci• signal delay in the lower atmosphere which also desion farming, aircraft navigation, marine navigation, selfpends on the angle the satellite signals subtends to the driving cars, land surveying, construction, drone navigaatmosphere and tion, augmented reality, animal tracking and military uses. • multipath errors of the satellite signals in mountainous, GNSS devices (of which GPS is just one) are less accuheavily wooded or urban terrain. rate than they might otherwise be due to introduced errors. Some of these error sources can be partially corrected Error sources can be categorised into two types: “user equivalent range errors”, which relate to timing by the GNSS system but others cannot. So how can higher accuracy be achieved? and path differences in the radio signals received from the GNSS satellites and “dilution of precision”, which relate to a non-ideal arrangement of satellites in the sky – Augmentation to correct errors GNSS augmentation involves gatherthe receiver cannot “see” enough satellites ing information about positioning errors, to establish or maintain reliable readings. Examples of user equivalent range errors, by Dr David Maddison such as those due to ionospheric delay, 14 Silicon Chip Australia’s electronics magazine siliconchip.com.au at various locations and times. This correction information can then be transmitted to GNSS receivers where it is combined with the normal positioning information to produce a more accurate “fix”. Perhaps the simplest method for calculating the position error is to have a ground-based station with an accurately known position, constantly receiving and decoding GNSS signals. The difference between the calculated position fix and known position is the error term. Other receivers nearby are likely to have a similar error term, as many of the error sources will be the same. Therefore, by transmitting the known error term from the fixed receiver to the nearby mobile receivers, they can correct their own position fixes, to get a much more accurate position. GNSS systems currently in use include GPS (US), Galileo (EU), GLONASS (Russia) and BeiDou (China). An augmentation system may be satellite-based, in which case it is known as a Satellite-Based Augmentation System (SBAS). Or error correction information may be transmitted from ground stations, in which case it is known as a Ground-Based Augmentation System (GBAS). SBAS systems operate over wide areas such as entire countries while GBAS have more local coverage. There are a number of SBAS systems now in use, most with non-global coverage. These include: (MSAS; Japan) • Quasi-Zenith Satellite System (QZSS; Japan) • GPS Aided GEO Augmented Navigation (GAGAN; India) • System for Differential Corrections and Monitoring (SDCM; Russia) • Wide Area GPS Enhancement (WAGE; US Military) • StarFire navigation system (commercially operated by John Deere) • C-Nav Positioning Solutions (commercially operated by Oceaneering) • Starfix DGPS System and OmniSTAR system (both commercially operated by Fugro) • Wide Area Augmentation System (WAAS; USA) • European Geostationary Navigation Overlay Service (EGNOS; EU) • Multi-functional Satellite Augmentation System The non-commercial SBAS augmentation signals can be received by nearly all modern GPS and other GNSS receivers, and starting this year, by some phone models (see Fig.16). The correction signals are transmitted by geo- The navigational paradox: more accuracy is not always better! The navigational paradox states that with greater navigational precision, the likelihood of ships, aircraft or land vehicles occupying exactly the same space on designated routes increases and so does the risk of collision. Solutions to this problem include requiring different vehicles on the same route to incorporate slight deviations from the nominal route, such as being offset from the route by a certain distance, and improved traffic management and collision avoidance systems. Fig.1: WAAS system showing original ground reference stations (yellow) and newer ground reference stations (red) added in Mexico, Canada and Alaska in 2008 to extend service area. Correction signals are generated by the ground stations, sent to the ground uplink stations and then sent to geostationary satellites where they are retransmitted to WAAS-enabled GNSS receivers. siliconchip.com.au Australia’s electronics magazine September 2018  15 stationary satellites (the positioning satellites are in lowEarth orbit). SBAS signals for GPS are transmitted on either the L1 frequency bands centred at 1575.42MHz or in some cases, the L5 band centred at 1176.45MHz which is a protected “safety of life” aviation band. These are the same frequency bands on which navigational data from the GPS satellites is received. In contrast to SBAS signals, receiving GBAS signals typically requires specialised equipment. Note that some SBAS systems are certified for aviation use by the International Civil Aviation Organisation (ICAO) and meet certain standards. Non-aviation use SBAS systems may use propriety technology which cannot be certified under ICAO standards. New Australian system In the 2018 Federal Budget, $225 million was allocated over four years to Geoscience Australia for the development A history of GNSS augmentation In the early days of GPS (which was developed by the US military but made available to civilian users worldwide), there was a concern that enemies of the USA would use GPS to guide missiles to targets within the USA or their allies. This lead to a deliberate signal degradation being imposed on the GPS service known as “Selective Availability” (SA), which made civilian GPS much less accurate than the military version. Originally, it was thought that the uncertainty in position would be about 100m but with better receiver designs, it became closer to 20-30m. This still wasn’t accurate enough for some users and Differential GPS or DGPS was developed as the US military insisted SA must remain, despite a lot of pressure from other US Government agencies as well as civilian users. It was eventually realised that the offset in the deliberately degraded SA signal was relatively constant and varied slowly, so if there was a land-based transmitter at a precisely known location, it could calculate the offset and transmit it to a suitable GPS receiver which would then apply the offset to the calculated position. That, along with measurements due to ionospheric delays, also transmitted by the base station to the receiver, enabled an accuracy of 5m even with SA enabled, as long as the receiver was suitably close to a DGPS transmitter site. As it wasn’t easy to provide DGPS transmitters at all the sites Fig.2: error estimates for civilian users of GPS hours before and after Selective Availability was permanently turned off. 16 Silicon Chip of an Australian GNSS augmentation system. This is intended to cover continental Australia as well as the Cocos Islands, Christmas Island and the Australian Antarctic Territories. Separately, Geoscience Australia is also running a twoyear project in conjunction with Land Information New Zealand (LINZ), to be completed in January 2019, to test two positioning technologies: next-generation SBAS and Precise Point Positioning (PPP). Note that there are already some commercial SBAS services available in Australia (eg, John Deere’s StarFire, described below). Like other SBAS systems, the new Australian system will take into account Australia’s continental drift (to the north-east at 7cm per year) which has put Australia’s official map grid out of kilter with its true position on the Earth’s surface (see panel for more details). Geoscience Australia is assessing the suitability of SBAS technology for agriculture, aviation, construction, maritime, required (as the range was tens of kilometres), the US FAA started investigating transmitting correction signals by satellite, which lead to the development of the Wide Area Augmentation System (WAAS) and eventually other similar satellite-based augmentation systems (SBAS) which are discussed in this article. By the mid1990s, it was apparent that DGPS had rendered SA of little value, which led to a decision to permanently turn it off on May 2nd, 2000. There are several types of DGNSS (Differential GNSS). Classical DGNSS, using an accurately surveyed reference station, can achieve position accuracies of 1m at distances up to tens of kilometres from the station. RTK (Real Time Kinematic) corrections use carrier phase measurements from the GPS satellites to achieve centimetre accuracy as long as a reference station is close to the receiver, preferably within 15km. WARTK (Wide Area RTK) allows for stations to be up to 500-900km distant. SBAS as described in this article, of which WAAS (USA) was the first of its kind, are gradually replacing those DGNSS systems which only work over short distances (ie, Ground Based Augmentation Systems or GBAS). SBAS works over continental areas and eventually should be available globally. These systems can be thought of as wide-area DGNSS systems. Fig.3: ground station showing 24 hours of data and scatter of positional data on May 1st (left) and May 3rd (right) before and after Selective Availability (SA) was turned off. With SA 95% of the points fell within 45m and with it switched off 95% frll within 6.3m. Australia’s electronics magazine siliconchip.com.au mining, rail, road, utilities and consumer use. According to Geoscience Australia, the specific elements of the test system are: • An L-Band satellite transmitter operated by Inmarsa • The operation of a satellite uplink at Uralla NSW by Lockheed Martin • A positioning correction service operated by GMV and Geoscience Australia • A GNSS ground tracking infrastructure operated by Geoscience Australia and LINZ • A testing program partnership between Geoscience Australia and FrontierSI. • LINZ overseeing the SBAS test program in New Zealand Testing has so far confirmed the expected accuracies for both second generation SBAS and PPP. Specific technologies being tested are: • Single frequency L1 “legacy” SBAS, equivalent to current WAAS and EGNOS systems, to improve position accuracy to one metre or less • Next-generation SBAS L1/L5 dual-frequency multi-constellation (DFMC) involving GPS and Galileo with the correction signal transmitted on L5 (see Fig.17). • PPP service using GPS and Galileo with correction data transmitted on L1 and L5 and an expected accuracy of 10cm or better Most GPS users in Australia should be able to see improvements from the “legacy” SBAS system right now but it is not currently certified for “safety of life” applications. Many GPS devices will use this data without any intervention but on my handheld GPS, I had to enable the option to use WAAS/EGNOS (see Fig.19). Despite the confusing names (WAAS/EGNOS are not available in Australia), the option enables SBAS, not necessarily those two systems in particular. More specialised equipment will be required to use next-generation SBAS and PPP. Overview of existing SBAS systems WAAS – USA The Wide Area Augmentation System (see Fig.1) was officially developed to improve the accuracy of GPS fixes used by aircraft. Testing of the system started in 1999 and it was commissioned in 2003. Ground reference stations measure inaccuracies in the GPS signals and send the corrections to master ground stations. These send the corrections on to the WAAS satellites every five seconds (or less) and they then transmit the signals to WAAS-enabled receivers. The WAAS specification requires a position error of no more than 7.6 metres both horizontally and vertically 95% of the time but typical accuracy figures achieved throughout the contiguous US and most parts of Alaska are 1.0 metre horizontally and 1.5 metres vertically. Since this is primarily an air navigation system, system integrity is of critical importance and if significant errors are detected in the GPS or WAAS system, a warning signal must be sent to users within 6.2 seconds to indicate that the navigational data is invalid. The system must also have Fig.4: European EGNOS system ground stations. RIMS are Ranging & Integrity Monitoring Stations that receive signals from US GPS satellites, MCC are Mission Control Centres for data processing and calculation of correction; and NLES are Navigation Land Earth Stations where data is sent to geostationary satellites for retransmission to end users. siliconchip.com.au a high level of availability, equivalent to downtime of no more than five minutes per year. EGNOS – EU The European Geostationary Navigation Overlay Service was developed by the European Space Agency and the European Organisation for the Safety of Air Navigation (EUROCONTROL) which started operations in 2005. It involves 40 Ranging and Integrity Monitoring Stations (RIMS) ground stations, four Mission Control Centres (MCC), six Navigation Land Earth Stations (NLES) and uses three geostationary satellites (see Fig.4). It provides correction data for the GPS (US), GLONASS (Russian Federation) and Galileo (European Union) GNSS systems. The system is designed to provide no less than seven metres horizontal accuracy but in practice, it is around one to two metres (see Fig.5 ). Work is currently underway to extend EGNOS coverage to southern Africa. EGNOS is primarily of value to aviation users as, due to the low angle of the geostationary EGNOS satellites over Fig.5: present coverage area of EGNOS showing horizontal and vertical position accuracy (HPE and VPE) at less than 3m and 4m respectively and the probability of achieving this accuracy. This data is sent out to EGNOS users and is frequently updated. In practice 1 and 2-metre accuracy is achieved. Australia’s electronics magazine September 2018  17 Is special equipment required to receive SBAS signals? Nearly every GNSS receiver made today is SBAS-enabled (for non-subscription services) and they are automatically configured to receive and use the signals with no extra hardware or software required. You do, however, need to be within an SBAS service area. There is also typically an accuracy difference between consumer grade GNSS receivers and professional and aerospace grade receivers. Some mobile phones are starting to support SBAS and the first to do so use the Broadcom BCM47755 receiver chip. the horizon, it is difficult to get reception on the ground in urban areas, especially in central and northern Europe. To overcome this problem, SISNeT (Signal in Space through the Internet) was developed, which transmits EGNOS corrections over the internet to users, primarily via wireless phone networks. SISNeT can be implemented via software on a smart mobile phone with an internet connection and a built-in GPS, or built into specialised navigation devices. In 2011, EGNOS was certified for “Safety of Life” applications such as aircraft navigation and landing under instrument flight conditions using a GPS approach to a runway. From 2020 onward, experiments will start on EGNOS Version 3 with dual frequency downlinks on both the L1 (existing) and the L5 bands as well as the use of multiple constellations (other GNSS systems). MSAS and QZSS – Japan The Multi-functional Satellite Augmentation System has operated since September 2007. A typical navigation fix obtained is within 1.5-2.0m accuracy. It is primarily used for aviation purposes (see Fig.6). Japan’s other SBAS system is the Quasi-Zenith Satellite System, which is designed to work with the GPS system. As distinct from MSAS, it is primarily intended to be used in the heavily built-up urban areas in Japan’s cities and mountainous regions where it is difficult to lock onto a Fig.6: MSAS system architecture. Note that there is a monitoring and ranging station (MRS) in Australia. While correction data is not valid for Australia, with the addition of extra ground reference stations this system has been determined to be able to be expanded for use in Australia. Image credit: Irene Hidalgo. 18 Silicon Chip Fig.7: ground track of non-geostationary QZSS constellation satellites. geostationary SBAS satellite low on the horizon. The satellite orbits are set up so that one satellite will always be over Japan at high elevation so it can be seen from within urban canyons. To achieve this, they were launched in inclined elliptical geosynchronous orbits and follow asymmetrical figure eight patterns as seen on the ground (see Fig.7). The first satellite was launched in 2010 and then an additional three satellites were launched in 2017 with the four satellite system expected to become fully operational this year (2018). The QZSS system is compatible with existing GPS receivers with no modification. The system is designed to be able to achieve sub-metre accuracy (see Fig.8). The positioning services offered by QZSS include the Satellite Positioning Service which will provide the same accuracy as GPS, the Sub-Meter Level Augmentation Service with an accuracy of 2-3m, the Centimeter Level Augmentation Service with an accuracy of about 10cm and Position Technology Verification Services for new positioning technologies as they are developed. Fig.8: coverage availability (i.e, the proportion of time a navigational fix can be obtained) in Ginza using GPS alone, using GPS and Galileo together, GPS enhanced with QZSS; and GPS enhanced with combined Galileo signals plus QZSS. The more blue in the images the better. Image source: JAXA, Japan Aerospace Exploration Agency. Australia’s electronics magazine siliconchip.com.au GAGAN – India The use of L1, L5 and L6 signals Currently, SBAS systems that use GPS satellites operate on the L1 band (centred at 1575.42MHz) but in the future, they will also use the L5 band, centred at 1176.45MHz (some already do). If an SBAS system observes both frequencies simultaneously, it is possible to directly measure the ionospheric delay of a GNSS signal to a much greater degree than just using the L1 alone. Also, the L5 signal is more immune to ionospheric storms and the use of two frequencies gives some redundancy in case one of the transmission bands suffers from interference. Since 2009, all new GPS satellites have been equipped to transmit navigational data using L5 signals. The Japanese QZSS system transmits an L6-band signal at 1278.75MHz with a data rate of 2000bps and if utilised, has the capacity to deliver real-time accuracy of 5cm horizontally and 10cm vertically, using PPP techniques. Fig.9: GAGAN system architecture. India was the fourth country to establish an SBAS system after the US, EU and Japan, with its GPS Aided Geo Augmented Navigation system, starting July 2013. It is managed by the Airports Authority of India and is primarily designed for air navigation but has other uses. It meets the requirements of international aviation bodies for “safety of life” operations and has a horizontal accuracy of 1.5m and 2.5m vertical (see Fig.10). GAGAN uses three geostationary satellites transmitting on the L1 and L5 bands, 15 Indian Reference Stations (INRES), the Indian Master Control Centre (INMCC) comprising three sites to process the correction data from INRES and three Indian Land Uplink Stations (INLUS) to transmit data to the GAGAN satellites (see Fig.9). An additional function of GAGAN is for ionospheric research. The ionosphere is relatively unstable over the Indian region and data will be used to design better algorithms for ionospheric corrections. Fig.11: location of SDCM ground stations around the world. tioning (PPP) for GLONASS. This technique is of interest because traditional techniques used with SBAS (real-time kinematics) lead to greater inaccuracy the further a user is from a base station, so a high density of base stations is required. PPP does not require any base stations to work and an algorithm is used that accurately incorporates numerous effects known to affect GNSS signals such as tropospheric refraction, earth crust movements and ocean tides, antenna phase centre shifts, phase spin and relativistic effects. PPP can provide centimetre level accuracy without needing base stations (see section on NASA GipsyX and panel on PPP). Fig.10: planned performance of GAGAN within specified coverage areas within the Indian Flight Information Region (FIR) for aviation. APV is Approach with Vertical guidance and RNP is Required Navigation Performance. SDCM – Russia The System for Differential Correction and Monitoring is designed to provide correction and integrity data for both GPS (USA) and Russia’s GLONASS system. It has 19 ground stations in Russia and four abroad, with a processing centre in Moscow (see Figs. 11 & 12). SDCM can provide a positioning accuracy of 1-1.5m horizontally and 2-3m vertically for normal users but can provide centimetre level accuracy within 200km of ground stations. Correction data can also be delivered over the Internet via SISNeT. Work is also underway to develop Precise Point Posisiliconchip.com.au Fig.12: availability of SDCM in coverage area, mostly over the Russian Federation. Australia’s electronics magazine September 2018  19 Precise Point Positioning (PPP) PPP is an alternative method for providing correction data to GNSS receivers. Whereas DGNSS requires ground reference stations with precisely known locations to obtain corrections, no reference stations are needed for PPP. In DGNSS, satellite orbit and clock errors are determined or estimated and transmitted to the receiver (called the “rover”), whereby the receiver applies the corrections to raw observations at the rover. In PPP, position coordinates are calculated with respect to the navigation satellite’s reference frame in space, not a specific ground reference station. Therefore, PPP should work globally, unlike SBAS which has a specific service area depending on how many ground reference stations have been installed. PPP requires precise mathematical models, such as NASA’s GipsyX, which take into account a large number of very subtle sources of error (see main text). After a control centre calculates the corrections, they are transmitted to the rover. The extremely accurate calculations made with PPP enables a higher level of accuracy than DGNSS. Another advantage of PPP is the possibility of reduced cost because a network of ground reference stations does not need to be maintained and corrections can possibly be transmitted to the rover with less bandwidth required than for DGNSS. A disadvantage of PPP at the moment is relatively long times to obtain a position fix or “convergence”. but just uses phase information of the two signals to make the calculations. The internal position fix calculated within the GPS receiver may be further enhanced with external correction signals, depending on the level of accuracy chosen and therefore subscription fee paid. According to a John Deere (Australia) online brochure for the StarFire 6000 receiver, the following levels of service are available: • SF1: ±150mm accuracy, no repeatability; position drifts over time. No subscription is required. Initial position determination takes 10 minutes. • SF2: ±50mm accuracy, no repeatability; not available for StarFire 6000 receiver, subscription required to receive correction signals. Initial position determination takes 90 minutes. • SF3: ±30mm accuracy with in-season repeatability, subscription required to receive correction signals. Initial position determination takes less than 30 minutes. • Radio RTK: ±25mm accuracy with long-term repeatability, subscription required to receive correction signals unless using own base station (see Fig.13). Initial position determination takes less than one minute. • Mobile RTK: ±25mm accuracy with long-term repeatability when mobile phone signal available, subscription required to receive correction signals. Initial position determination takes less than one minute. Starfire – John Deere (commercial) GipsyX (PPP) – NASA The StarFire navigation system is commercially operated by John Deere and used in precision agriculture for vehicle guidance (See Figs. 13 & 14). Also see the article about Agbots in the June 2018 issue of SILICON CHIP for more information on its usage: www.siliconchip.com.au/Article/11097 StarFire broadcasts correction signals from satellites on L-band frequencies to give high levels of position accuracy. John Deere operates a number of ground reference stations around the world, including Australia, to generate the correction signals. Unlike other SBAS systems, the correction accuracy is said to be independent of the distance from a ground station. StarFire receivers use L1 and L2 frequencies from GPS satellites. The encrypted military P(Y) signal on L1 is used in conjunction with the P(Y) L2 signal to accurately calculate ionospheric delays. It cannot decrypt the P(Y) signal GipsyX is a set of real-time GNSS data processing techniques and software developed by NASA to obtain global corrections for GNSS satellites. It improves the accuracy of GNSS systems such as GPS and GLONASS (Galileo and BeiDou support is being developed) by modelling complex and subtle effects that lower the accuracy of GNSS devices. GipsyX enables Precise Point Positioning (PPP; see separate explanatory panel). Effects taken into account include: Fig.13: John Deere StarFire RTK base station that acts similarly to other SBAS base stations. It provides a repeatable 2.5cm accuracy. 20 Silicon Chip • Short-term and long-term changes in the Earth’s orientation, including polar motion and variations in Earth’s axial rotation angle (UT1). • Solid Earth body tide deformations. • Ocean tide loading deformations. • Transmitter and receiver antenna calibrations. • Satellite attitude variations. • Phase wind-up, which relates to the fact that satellites Fig.14: John Deere guidance display inside tractor or similar vehicle. Australia’s electronics magazine siliconchip.com.au • • • • • must rotate to keep their solar panels pointed toward the Sun. This rotation causes the phase of the radio signal to change with respect to the receiving antenna and this is misinterpreted as a variation in range, with an error of around 10cm. Quaternion compensation for vehicle attitude such as spacecraft and aircraft. A quaternion represents the relative rotation of two coordinate systems such as between a spacecraft and a fixed frame of reference such as earth or another spacecraft; only rotational orientation is considered. General relativity (as described by Albert Einstein). Crustal plate motion (eg, Australia moving north-east at around 7cm per year). Second order ionospheric corrections. First order ionosphere delay corrections based on the L1 and L5 transmissions can give centimetre level accuracy but second order effects need to be taken into account for millimetre accuracy. These stem from the change in polarisation of radio waves as they travel through the Earth’s magnetic field (Faraday rotation), leading to an error of 1-10mm. The effects of a dry or wet troposphere (the lowest 6-10km of the atmosphere) on signal delay. This involves one of several mapping functions; either GPT (Global Pressure and Temperature model), GMF (Global Mapping Function), VMF (Vienna Mapping Function) or NMF (Niell Mapping Function). Additionally, GipsyX takes into account for orbiting space vehicle complex force models that include: • • • • • • • High-order Earth static gravity fields. Atmospheric drag. Solid earth, ocean, and pole tide gravity fields. Solar and terrestrial radiation pressure. General relativity. Third body effects from the Sun, Moon and other planets. Custom and general models of spacecraft shape. C-Nav – Oceaneering (commercial) C-Nav Positioning Solutions is commercially operated by Oceaneering. It uses the technique of Precise Point Positioning (see panel) and is generally known as GcGPS or Globally corrected GPS. It generates correction data by a proprietary implementation of NASA’s GipsyX software and it broadcasts orbit and clock corrections for all GNSS satellites simultaneously from its own satellites. It is available all over the world from 72°N to 72°S latitude. Typical accuracy is better than 5cm horizontally and 15cm vertically. It has over 40,000 users worldwide, on a subscription basis. Proprietary receivers are required to use this system. C-Nav works as follows. Worldwide Global GPS Network (GGN) ground reference stations collect dual frequency L1 and L2 data (other frequencies such as L5 may be used). This data enables ionospheric and other measurement to be made. The raw data is transmitted to two “hot” Network Processing Hubs plus a backup hub via the internet. Independent Refraction Corrected Orbit and Atomic Clock Offset corrections for all GPS satellites are then computed by the Network Processing Hubs. Corrections are then sent via an uplink to geostationary satellites whereupon they are retransmitted to users (see Fig.18). siliconchip.com.au How changes in the Earth’s shape affect accuracy With navigation systems becoming so accurate, it is important to consider what frame navigational data is referenced to since the Earth is constantly changing shape due to continental drift, uplift, subsidence and other factors. This affects both the notional altitude and position at any point near the Earth’s surface. The GPS system was originally referenced to the US Department of Defense World Geodetic System of 1984 or WGS 84 (now called WGS 84 [Original]). It was actually defined in 1987 with a world survey done using Doppler satellite surveying techniques. WGS 84 (Original) was upgraded in accuracy using GPS measurements in 1994, to WGS 84 (G730). It was again upgraded to WGS 84 (G873) in 1996 to be more closely aligned with the International Earth Rotation Service (IERS) Terrestrial Reference Frame (ITRF) 94. It was then called WGS 84 (G873) and used from 1997. In 2002, WGS 84 (G1150) was implemented and followed by WGS 84 (G1674) from 2012. Unfortunately, the Earth continues to change shape and the difference in position using WGS 84 (Original) at the present can be 1-2 metres, perhaps more. The International Earth Rotation Service (IERS) computes the positions for specific sites on the Earth on a regular basis and the data is fed into the International Terrestrial Reference Frame (ITRF) for the current epoch (time period). The ITRF is an internationally accepted standard and the most accurate geocentric reference system, and so it is the reference frame used for SBAS corrections. WGS 84 (G1674) agrees with ITRF to within about 10cm. In Australia, the current reference frame is the Geocentric Datum of Australia 1994 (GDA94). However, since this was established in 1994, the Australian tectonic plate has shifted by 1.6m meaning that Australian coordinates are no longer aligned with GNSS coordinates such as GPS (based on WGS 84), making high accuracy navigation impossible. Australia has therefore implemented the Geocentric Datum of Australia 2020 (GDA2020), based on the projected position of the Australian continent on the Earth’s surface in 2020. If this datum is used now, the offset from GNSS coordinates such as GPS will be 20cm but they will converge in 2020. GDA2020 is closely aligned with ITRF2014. Starfix – Fugro (commercial) The Starfix system by Fugro is a commercial system primarily aimed at navigation for offshore construction vessels, survey operations, pipe laying and cable laying activities, seismic surveys, dive support and installation and monitoring of floating storage of offshore oil and gas at the point of production. Their correction data is delivered via satellite or the Internet in a proprietary compressed format. It works with GPS, GLONASS, BeiDou and Galileo. Centimetre, decimetre and sub-metre accuracies are available. For regions at high altitude beyond about 75°N or 75°S, beyond the reach of their geostationary satellites, correction data is delivered by Iridium satellites which are in polar orbit and have global coverage. A variety of services are available: Australia’s electronics magazine September 2018  21 Possible future for SBAS Fig.15 (below left) shows SBAS coverage in 2013 while Fig.16 (right) shows the predicted coverage (at the time) for 2020-2025, showing near-global access. This includes WAAS, EGNOS and MSAS with an enhanced system including SDCM and GAGAN as well as dual frequency • Starfix.L1 is a single-frequency system using L1 and can provide a position fix within one metre. • Starfix.XP2 uses GPS and GLONASS and obtains orbit and clock corrections from a third party with further corrections by Fugro software. It uses Precise Point Positioning (PPP; see panel). Accuracies of better than 10cm horizontally and 20cm vertically can be obtained. • Starfix.G4 uses GPS, GLONASS, Galileo and BeiDou with clock and orbit corrections provided from Fugro’s own network of ground reference stations, with additional corrections provided by proprietary software. Accuracies better than 10cm horizontally and vertically can GNSS (L1 and L5 bands) and an expanded network of stations in the Southern Hemisphere. The figures come from the European Space Agency and do not include any possible contribution from the Australian SBAS system under development, as it pre-dates the announcement. be obtained. • Starfix.G2 is a subset of Starfix.G4 but uses only GPS and GLONASS. • Starfix G2+ uses GPS and GLONASS with clock and orbit corrections enhanced with carrier phase corrections from the Starfix.G4 network, plus in-house augmentation algorithms. Better than 3cm horizontal and 6cm vertical accuracy can be achieved. OmniSTAR – Trimble (commercial) The OmniSTAR system, owned by Trimble, is another commercial augmented GNSS service. OmniSTAR correction signals are proprietary in nature and service is avail- Fig.17: existing free-to-air SBAS service areas showing positions of geostationary satellites that transmit correction data. Australia, Antarctica, Africa and South America are the four main land masses not currently covered by SBAS. Initiatives are under way to provide SBAS in Africa as an extension of EGNOS, South America with SACCSA (Solución de Aumentación para Caribe, Centro y Sudamérica / Augmentation Solution for the Caribbean, Central and South America) and Malaysia and South Korea with KASS (Korean Augmentation Satellite System to be in place by 2021). 22 Silicon Chip Australia’s electronics magazine siliconchip.com.au Fig.18: Coverage for single frequency (L1) and dual frequency (L5) SBAS test. Image source: Geoscience Australia. able in most areas of the world, including Australia. Their services include: • OmniSTAR HP, their premium service uses an L1/L2 dual frequency receiver. It has an accuracy of 10cm. • OmniSTAR G2 uses GLONASS satellites and correction data and is suitable for use in areas with limited satellite visibility such as mountainous regions, heavily vegetated and built-up areas. Accuracy better than 10cm is possible. • OmniSTAR XP is a dual-frequency system with orbit and clock correction, with a long-term repeatability of 10cm and is suitable for precision agriculture. • OmniSTAR VBS is the basic single frequency service using L1 and receives correction data from regional ground reference stations. An accuracy of better than 1m can be achieved. Other augmentation systems Wide Area GPS Enhancement (WAGE) is an obsolescent system of the US Military with an uncertain service status. It is used to improve the horizontal accuracy of the encrypted GPS signal used by the military, on specialised receiv- Fig.19: a typical hand held consumer GPS display showing the positions and signal strength of the satellite signals being received. Note the “D” in the signal strength bars indicating a correction signal (or differential signal) is being received for that particular satellite. The correction signal is transmitted by a different satellite than the GPS satellites. This correction signal is being received because of the SBAS test bed now operating in Australia. ers. The military GPS signal is encrypted to prevent an enemy spoofing the signal to cause an inaccurate position fix. Modern standard GPS receivers outperform WAGE. WAGE has been superseded by Talon NAMATH, about which there is little published information and any existing WAGE users are being encouraged to use it. Conclusion As shown in the panel on future predictions of SBAS availability, in the near future, enhanced or augmented GPS will be available over all occupied areas of the Earth’s surface and most of the oceans. This will mean that pretty much everyone will be able to determine his or her own position to within about 1m on the Earth’s surface, making vehicle and personal navigation substantially more reliable. It will also enable many new technologies which are not practical with the present ~5m typical inaccuracy, as deSC scribed in the introduction. Fig.20: overview of C-Nav system. siliconchip.com.au Australia’s electronics magazine September 2018  23