How does differential GPS positioning work

5.5 Differential GPS

The number of GPS application areas and the Accuracy requirements have increased in the course of the last few years to the overall system. They are sufficient for many scientific work areas guaranteed 10 - 50m Position errors are no longer possible, as there are many measuring methods large-scale maps include and there the Positional accuracy about the Character width (line width) at about 3m lies (e.g. German basemap DGK-5). In the geodetic area, the position accuracy (error) should even be in Millimeter range lie!

One therefore has for some areas of application additional facilities and Systems designed, which significantly improve the accuracy of civilian positioning:

  • Differential GPS (DGPS) as local or Wide range system
  • Pseudolites (terrestrial pseudo GPS satellites)
  • Extensive system expansion of the GPS: Wide Area Augmentation Systems, e.g. WAAS (am.), GAGAN (jap.), MSAS (ind.) andEGNOS (europ).
  • Integrity checks on the receiving end (Receiver Autonomous Integrity Monitoring)

Of the options mentioned above, this represents DGPS for most geoscientific applications one economical solution because it is new by means of minor modification of the original GPS system Location accuracies of <2m and many modern GPS receivers are technically prepared for DGPS operation. For this reason, the Basics of DGPS technology explained.

The basic principle of the DGPS

For the DGPS-assisted Geospatial data collection in terrain are often used as a topographical basisdigital official basic measuring sheets or cards related (e.g. cadastral plans, TK-50, -25 or DGK-5 ...), which, according to their scale, show inaccuracies regarding the location of spatial objects. The medium positional accuracy is already included with the official land survey based on the DGK-5 about 3m! A conventional handheld GPS (e.g. Garmin 60csx) achieved Accuracies of approx. 3m with EGNOS correction switched on - is thus roughly within the character accuracy of a digital DGK-5.

Note: The theoretical positional accuracy of objects on a map can, starting from the Character accuracy and the scale, determine at any time:

Theor. Positional accuracy = drawing accuracy x scale / 1,000 m

(For a map on a scale of 1: 25,000, the drawing accuracy is around 0.2mm; this results in a theoretical positional accuracy of +/- 5m!)

The DGPS procedure now offers the possibility with relatively little effort medium quality to increase the location determination well beyond the range of the drawing accuracy of a conventional map (0.1-2 meters)! Be for this purpose Error compensation quantities used to calculate the pseudo-distances (position), which as a additional correction signal is made available to the GPS receivers.

The procedure is for that civil usersthat with the Carrier L1 and the C / A code work must have been developed. The technical facility of the Reference transmitter is operationally independent. There are therefore a large number of fixed or movable reference transmitters in Germany alone, which have a corresponding, type-dependent Correction signal radiate.

The procedure is based on the Comparison of the GPS position of a geodetically fixed reference station with the distance or coordinates calculated by the GPS. The resulting differences are recorded as Correction signal (data) sent out to Correction of the terrain data used in the GPS (Fig. 5.5.1).

Fig. 5.5.1: Principle of the DGPS structure and the transmission paths for GPS and correction signals (Mansfeld, 1998)

The calculated Correction values apply only to the relevant reference station, whereby the accuracy of the correction data depends on the accuracy of the geodetically determined position (WGS-84) of the phase center the antenna on the transmitter (Base transmitter and receiver) is.

According to Brinkkötter-Runde (1995), the decisive factor for error compensation quantities is that at neighboring observation points a large part of the error influences attributable to the GPS measured variables are the same (proximity can, however, in the real case mean several kilometers!). Is this correlation within a finite distance high, she canrelative position determination of the points to each other high accuracy respectively.

Is now the position of a reference point geodetic in WGS-84 known and will be there contemporaneous an (almost) identical GPS signal as processed on the mobile receiver can be obtained from the resulting difference (!) to the other points with identical reception characteristics the relative accuracy of the points to one another to oneabsolute positional accuracy be mapped in the field.

Note: In general, the greater the distance to the transmitter (see Fig. 5.5.7), the greater the inaccuracy of the correction in the GPS receiver, since the measurement conditions can no longer be regarded as sufficiently equal!

There are now various options that are needed Correction data to transmit, with the different radio equipment show particular advantages and disadvantages with their frequency ranges (Tab. 5.5.1):

Sending facility Frequency range advantages disadvantage
Long and medium wave transmitters 100 - 600 KHz long range (1000km),
Channel sharing
small bandwidth, low bit rates (up to 300 bps)
Marine radio beacon Europe 283-315 KHz so. so.
Aeronautical beacon Europe 255-415 KHz so. so.
Shortwave transmitter 3 - 30 MHz long range with sky waves,
Frequencies available
small bandwidth, space wave time and frequency dependent
VHF 30-300 MHz large bandwidth, 4000 bps, shared use of existing transmitters, frequencies available Range restricted by quasi-optical conditions (> 100 MHz)
Cellular networks (e.g. C, D, E ..) 450, 900, 1800 MHz Shared use of networks limited range, synchronization problems
Pseudolites (mod. GPS transmitter) 1.2-1.5 GHz Large bandwidth and range of services Limited range and interference with GPS
Satellite subsystems (e.g. EGNOS / Galileo) so. Large area coverage high costs

For civil users are two types of differential evaluation important:

  • Position correction method
  • Measured value correction procedure

in the Position correction method the reference receiver / transmitter determines the difference to the exact geodetic position of its location from the currently calculated GPS position (target coordinates of the reference point). These Coordinate difference is transmitted as a shift value to the user's mobile GPS receiver, which then sends it to its own GPS position. However, it must be noted that, strictly speaking, the satellite constellation between the reference transmitter and the terrain GPS cannot always be identical, i.e. the slightest deviations in the framework conditions of the measurement occur. To avoid extreme differences in the satellite constellation, the Horizontal areas of reception (Azimuth) matched to each other: e.g. reference transmitter with 10 ° elevation mask and mobile GPS receiver with 15 ° elevation mask, i.e. up to 300 km identical satellite constellations!

Measured value corrections refer to the differential evaluation Pseudo distance measurement of the reference receiver to all visible satellites. By estimating the receiver clock errors and calculating the pseudo distances, the target routes are determined by comparing them with the actually measured pseudo routes. These Pseudo distance correction values are communicated to the user, who can now carry out his own pseudo-route calculation more correctly. Compared to the position correction method, this type of correction data determination is more time-consuming but also more precise and flexible.

The principle of Correction by pseudorange measurement proceeds as follows (Mansfeld, 1998): The position of the reference station is determined by the geodetic coordinates x g, y g and z g given. Likewise is the Position of the satellites known by the navigation message with the coordinates x i, y i and z i. Thus, the geometric distance rgi from the reference station to the i-th satellite

r gi = [(x i -x g) 2 + (y i - y g) 2 + (z i - z g) 2] 0.5

be calculated.

The Pseudorange (ρ) is determined by the following expression:

ρ gi = r gi + εgR + ε gK + εgN + c δ t g


  • ε gR pseudorange error in the space segment
  • ε gK pseudorange error in the control segment
  • ε gN pseudorange error in the user segment
  • δ tg time difference compared to GPS time

The order of magnitude of the numerical values ​​is given in Table 5.5.2.

The Difference in the measured pseudo-distances intended for geometric distance:

δ ρ gi = ρ gi - r gi = ε gR + ε gK + ε gN + c δ t g

This Difference value is used by the broadcaster as Correction signal transmitted to the receiver, where the correction of the measured pseudo-distance is then carried out ρ i he follows:

ρ gi - δ ρ gi = r pi + ε pR + ε pK + ε pN + c δ t p - (ε gR + ε gK + ε gN + c δ t g)

Generally they are Components the recipient's pseudorange error from the user identical with those of the pseudorange error in the reference station. Exceptions are the errors caused by theMultipath propagation and through that Receiver noise arise. The corrected pseudorange can thus be expressed as follows:

(ρ pi) corr = r pi + c δ t p + ε p + c δ t a)

It contains δt p the remaining errors with the user, δt a the (time) errors due to multipath propagation and εp the receiver noise!

The position of the user is determined by measuring the pseudoranges to four or more satellites. As the correction data of the pseudorange increases discrete times are sent by the reference station and the satellite movement causes a change in the errors with regard to the pseudo-distances, the correction data also change constantly and are therefore provided with an additional, time-dependent correction Mistake!

Tab. 5.5.2: Error balance of the pseudorange measurement between GPS and DGPS with activated SA; Values ​​as standard deviation in meters (Mansfeld, 1998)

GPS segment Source of error GPS (SA) DGPS (SA)
Space segment Satellite orbit disruption
Satellite clock error
Control segment Ephemeris error
User segment Air time delays
- ionosphere
- troposphere
Multipath propagation
Overall system root mean square value = equivalent distance error 33,3 3,3

According to the two methods described above for calculating correction values, the necessary differential compensation be carried out in different forms of operation (Brinkkötter-Runde, 1995):

  • Up-link procedure
  • Down-Link Procedure
  • Post processing method

in the Up-link procedure the correction values ​​determined by the reference station are in Real time sent to the mobile station via radio data transmission (e.g. RDS). The differentially corrected data is available to the userright away to disposal; this process is also called real-time DGPS and is used wherever Real-time position information is necessary in the error range of approx. 1-5 m.

The Down-Link Procedure uses the opposite data path; i.e. the mobile receivers send their raw data to the reference station, where the differential equalization then takes place stationary is calculated. The corrected position data are then sent back to the user via a control center or held in a database for queries.

The versatile Post processing method are used wherever more precise position data are not required immediately in real time, or where no suitable correction data radio links are available. They only become mobile after all raw position data has been collected collected coordinates converted into corrected GPS coordinates by means of correction data recorded at the same time elsewhere (temporally separated work step!). This procedure has a mostly higher accuracy on (approx. 1m) and allows post-processing of GPS locations even with real-time positioning.

Format of the correction data

The Transfer protocol the correction data for the DGPS operation was replaced by the Radio Technical Commission for Maritime Service Study Committee 104 (RTCM SC-104 or short RTCM) Are defined. As Figure 5.5.2 shows, it consists of one Message frame with a varying number of 30-bit words. The first two expressions make up the Header (Header information). The first word contains the 8-bit preamble, the frame ID, the 10-bit station information and 6 check bits. The frame IDs identify one of the 64 possible message types (see below) and the 10-bit station information identifies the reference transmitter. The second word consists of the modified Z-Count, the three bits for the sequence number, the 3-bit station status information and the 6 check bits.

Fig. 5.5.2: Message frame and header information of the RTCM signal for the DGPS (Mansfeld, 1998)

The following is a brief overview of the 63 possible message formats:

Type 1Words 3-7; contains the Correction data for all satellites which can be 'seen' by the reference station. In addition, the User Differential Distance Error (UDRE) given as a 2-bit code (estimated uncertainty factor over the standard deviation of the pseudo-distance, see Fig. 5.5.3); in addition, the satellite identification values ​​for which the correction data apply.
Type 2 Current data to be used if the user Path and time parameters used that are older than that of the reference station.
Type 3 Fixed 6-word format with which the Exact position (cm) the Reference station is communicated.
Type 4 Carrier phase measurement for the Land survey (Geodesy), in the future via type 19-21.
Type 5 Operating condition the satellite.
Type 6 Fill data (no information!).
Type 7 For Marine beacons as a reference station with switching mechanism with regard to the closest beacon for navigation.
Type 8 Almanac data for pseudolites.
Type 9 Like type 1, but only for optimal satellite constellations (speed gain!).
Type 10 Only for P code!
Type 11Just C / A code about L< class="Stil4">2 !
Type 12 For pseudolites, especially for Time difference and phase location the antenna.
Type 13Gives the exact location of the transmitter.
Type 14Only for Land survey !
Type 15To transfer real ionospheric and tropospheric measurement data for correction (instead of model specification!).
Type 16Graphic Message (for computers with display).
Type 17Total Satellite orbit data for correcting satellite data without orbital parameters.
Type 18-21 Correction data for Pseudorange measurement and carrier phase for the kinematic DGPS- Application.
Type 22 - 58 Not yet occupied.
Type 59 Property data (-right).
Type 60 - 63 Different variable parameters to the transmitter.

Fig. 5.5.3: Message format type 1, word 3 to 7 of the RTCM signal for the DGPS (Mansfeld, 1998)

Distance and Accuracy

The influence of the Distance to the reference transmitter The accuracy of the DGPS measurement has to be taken into account for many applications (e.g. data acquisition, air traffic, etc.). The estimation of the DGPS pseudoranges with high accuracy almost always only applies to users in relative, finite proximity to the reference station. If the user is now in large vertical or horizontal distance and also moves with it high speed, are the influence of different parameters (Electron densities in the ionosphere and the Troposphere, Doppler effects etc.) comparatively large.

One therefore has for them Flight navigation special models to contain this error contribution are implemented using model calculations, so that the differences in transit time between the reference station and the actually measured transit time of the signals on board an aircraft can be matched to one another. The essential parameters are height and the Refractive index the Tropos or ionosphere(Fig. 5.5.4 and 5.5.5).

Fig. 5.5.4: System components of the DGPS in flight navigation (Mansfeld, 1998)

Fig. 5.5.5: Changes in running times depending on the amount (Mansfeld, 1998)

Reference station networks

The requirements for the DGPS are very different and therefore become technical user-specific implemented. This is especially true for the Navigation (sea, air and land transport), the Land surveying (geodesy) but also increasingly for the digital Geospatial data acquisition (sampling, mapping, field inspections)(-> see also internship 10th session!).

in the Navigation area need powerful DGPS traffic control systems be set up so that correct position determinations can be made in real time when approaching the landing or when approaching port entrances. The highest accuracy requirements can only be achieved using the staticDifferential method be solved (geodesy).

The application of the Differential method assumes that appropriate DGPS reference stations be available. Due to the different requirements are therefore in the course of time, partly independent of each other, both individual DGPS stations as well as complete station networks has been put into operation. The technology is being developed simultaneously in all countries in which GPS is used (Mansfeld, 1998).

Both geodetic method the large-area measurement takes place within the framework of a reference system, which is given by defined reference points. In modern geodesy, these are points which, in addition to the exact coordinate of the reference system, also perform the task of the DGPS reference station. Mobile transmitter / receiver systems are therefore already calibrated for every geodetic campaignTopographic points (TP) set up; in the case of stationary systems are yours exact coordinates known. The high accuracy is now over a simultaneous Carrier phase measurement of the signal of several satellites with mostly two reference receivers initiated.

There are three variants the measurement, which in principle provide eight distance values ​​from which differences can then be formed. Although the pseudo-distances can also be determined via the code phase measurement, the more precise carrier phase measurement is used in geodetic practice:

  • Single differences: Carrier phase measurement between two satellites and a recipient. In this way, receiver-dependent errors (Clock failure) turn off. For measurements between a satellite and two recipients can make the satellite dependent errors like Time and Pahn parameter fluctuations to be compensated. Also Ionospheric and tropospheric influences can be eliminated if the distance between the two receivers is not too great.
  • Double differences: combination of two single differences; i.e. carrier phase measurement between two satellites and two recipients. Bug fixes as above!
  • Triple differences: Combination of each two double differences todifferent times. This method is used to eliminate the ambiguity used by carrier phase measurements and is also used in geodesy as Geometric method designated.

In order to have one in geodesy at every location Regional area To be able to carry out GPS surveys nationwide reference stations be available. Such a station can if necessarytemporarily as a mobile system be erected, but mostly will be on fixed stationary systems resorted to. As a rule, the user then only needs a (D) GPS receiver and access to the data of the reference stations (Correction data). Since the effectiveness with the distance to the stations decreases, must for a sufficiently dense network at reference stations (Mansfeld, 1998).

With this in mind, the Federal Republic since 1991 the Satellite positioning service (SAPOS) set up, which from a network of permanently registered geodeticGPS reference stations the national survey is built. This network provides the correction data for the entire area of ​​the respective federal state (mainly for geodetic purposes). The operation of the respective stations is the responsibility of the federal states, although since 1998 around 150 reference stations have been set up and put into operation.

The correction data is also used for Navigation tasks offered. In the final phase of the realization should200 fixed stations take care of the flow of correction data. The locations of the stations are within theGerman reference network (DREF 91) coordinated and are an integral part of the European Reference Network (EUREF 93) at. To ensure interference-free measurements, the locations are chosen so that no topographical or other type of obstacle stands in the way of wave propagation.

The Correction data the user receives about Radio services or the mobile Telephone network. For real-time applications, the data can also be accessed via Radio station the ARD in the VHF area (RDS system or RASANT) or (unfortunately only until 2005) via the Mainflingen LW transmitter sent out (see Tab. 5.5.3). A is required for DGPS-compatible GPS receivers Auxiliary reception modem (RDS modem), which is connected to the GPS to feed in the RTCM correction data. In some federal states there is also a special radio service on 160 MHz (2m band). Cellular networks (GSM) or the telephone line system is used to provide the correction data for post-processing. This applies to all data RTCM (SC-104) Standard format (Vers. 2.0 / 2.1)!

Tab. 5.5.3: Characteristic values ​​of SAPOS (Mansfeld, 1998)

Operating mode SAPOS transmission distance Error-
Real-time code measurement EPS VHF, MW, LW, geostationary satellites, internet up to 500km 10cm - 10km 1 to 3m 1 min
Real-time carrier
phase measurement
HEPS 160 MHz, cellular network up to 25km 2cm - 10km 1 - 5cm 2 min
Post processing
Carrier phase measurement
GPPS Telephone, ISDN, mobile communications up to 10km 1cm - 10km 1 cm 10 min
Post processing
Carrier phase measurement
GHPS Telephone, ISDN, mobile communications, data carriers up to 10km 1cm -10km <1cm > 45 min

In the Maritime shipping the reference stations have the task of enabling precise positioning by means of DGPS for shipping. For this purpose, one has the construction and operation of the plants in GermanyHelgoland (North Sea) and Wustrow (Baltic Sea) placed in the hands of the state. The two stations are part of one European network of approx. 50 reference stations in the coastal area (Fig. 5.5.6). The Correction data signal will be in RTCM standard above Marine radio fire between 283.5 and 315 KHz (Europe) or 285 and 325 KHz (USA). The accuracy of the DGPS location is around 1m at a short distance from the transmitter and 3m at a greater distance (approx. 400 km)! The maximum range is several 100 km. The position is calculated using the Differential method (see above) and can also be used for terrestrial DGPS positioning in coastal areas! For the area of Inland shipping Usually combined VHF and LW reference stations (so-called MF beacons) are used!

Fig. 5.5.6: Network of DGPS reference stations on the coasts of Western Europe (Mansfeld, 1998)

In the aviation the high-resolution DGPS is used as a supporting tool for theLanding approach so far only experimental used at Altenburg / Thuringia airport. The reference data have a local character and are transferred via VHF (108 - 118 MHz) the Instrument Landing System (ILS) of the aircraft on approach. Using the VHF has the advantage high transfer rates (4800 bit / sec), but with a quasi-optical connection must be guaranteed between the GPS receiver and the reference transmitter (this is always the case in the late phase of the landing approach!). The position accuracy in the endurance test is between 10 - 15 cm.

In addition, so-called Pseudolites (a made-up word Pseudo and satellite ) used which terrestrial broadcasting stations represent, but one Functionality like a GPS satellite have. The pseudolite sends the C / A code on L1 its signals are now additionally processed by the mobile GPS receiver. Due to the earth-bound position of the pseudolite, mainly more favorable ones result DOP factors. The geometrically better sending and receiving conditions optimize the time and location-dependent parameters as well as VDOP and HDOP factors. The result is a more precise position calculation. In the future it will be the Microwave landing system in connection with DGPS are used internationally. Here the correction signal is transmitted in the C-band (5 GHz). Classic DGPS methods are used to determine the position during long-haul flights.

For terrestrial traffic control systems the DGPS plays a special role because it has a certainBroad impact is given in civil use and the requirements are high. At the same time that isNetwork of reference stations heterogeneous and it avails itself different data transmission paths.

The VHF area, in which the correction data inRTCM standard about the Radio Data System (RDS) e.g. in Radio (FM, 88 - 140MHz) be sent out. An application example are the RDS corrections via WDR 5. The mean achieved range of each transmitter is 120 km, the Location error at approx. 2m! The RDS is also used in some neighboring European countries to transmit the correction data via radio stations. In many of the car manufacturers' car models, the permanently installed pilot systems are based on RDS-supported DGPS technology (car radio often as a receiver)!

Until the end of 2005, there was another possibility of receiving correction data for the broad civilian user community of the DGPS. The approach was to send the data through a Long or medium wave transmitter, Mainflingen with large area coverage (> 500 km radius). The latter was from theGerman Telecom operated and had a transmission frequency at 122.5 KHz with 50 KW power. The data was saved in RTCM format 2.0 with 300 bit / sec and F1 modulation. The one achievable with it accuracy (depending on the distance) was im Medium at 2m (see Fig. 5.5.7). After the shutdown of ALF, many civil users switched to the RTCM system RAPID changed, which is assigned to the nationwide SAPOS operation. Here, however, the RDS modem must be reset to the appropriate frequency depending on the transmitter location. However, the RASANT system was completely discontinued in 2012.

Fig. 5.5.7: Three-dimensional positioning error and the resulting accuracy of the RTCM signal as a function of the distance to the telecom transmitter Mainflingen (Mansfeld, 1998)

Note: The correction signal that is used in the internship for DGPS commissioning of the Garmin-GPS-12 no longer comes from the Mainflingen long-wave transmitter. SAPOS is used!

International reference services

The DGPS described above largely possesses national character and is therefore limited in its range to a few 100 to 1000 km. A DGPS solution that can be used worldwide consists in one network together connected reference stations, which exchange their correction data with each other and make it accessible over great distances. Such a system is called Wide Area DGPS (WADGPS) (Mansfeld, 1998). Only a small number of reference stations is necessary to cover a large region, since the WADGPS was designed with the Pseudorange error in their regional components be decomposed, through whose sectors e.g. an airplane moves. Each component is calculated for a regional section and therefore the calculations for each individual station within the relevant region are not required. The correction data are only used by selected onesRegional Control Stations (RCS) passed per sector.

According to the basic concept, this includes WADGPS network different reference stations, which an exact determination of the ephemeris, the atmospheric delays and the time shifts per sector takes place. In the Main Control Station (MCS) with subordinates Reference stations (RS) the signals are coordinated (Fig. 5.5.8). One thing is important absolute time synchronization of the stations and a working one communication between them, which by special geostationary satellites is achieved. EGNOS and the European satellite navigation system GALILEO represent a first step towards a European section of this WADGPS.

Fig. 5.5.8: WADGPS network using North America as an example (a) and in use for flight navigation (b) (Mansfeld, 1998)