`
`EX. PGS 1042
`
`
`
`
`
`
`
`PCT
`WORLD INTELLBCfUAL PROPERTY ORGANIZA 110N
`International Bureau
`INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION 1REATY (PCT)
`WO 96/21163
`
`(11) International PubUcation Number:
`
`(51) International Patent Classification 6:
`GOlS M4
`
`Al
`
`(43) International PubHcation Date:
`
`11 July 1996 (11.07.96)
`
`(Zl) International AppHcation Number:
`
`PCTIN095/00239
`
`(22) International Filing Date:
`
`20 December 1995 (20.12.95)
`
`(30) Priority Data:
`944954
`
`20 December 1994 (20.12.94) NO
`
`(71) AppHcant (jor all designated States except US): GECO A.S.
`[NO/NO]; Bjergstedveien 1, N-4000 Stavanger (NO).
`
`(72) Inventor; and
`(75) Inventor/Applicant (jor US only): VIGEN, Erik [NO/NO];
`Bergstien 21, N-3016 Drammen (NO).
`
`(74) Agent: ONSAGERS PATENTKONTOR A/S; P.O. Box 265
`Sentrum, N-0103 Oslo (NO).
`
`(81) Designated States: AM, AT, AU, BB, BG, BR, BY, CA, CH,
`CN, CZ, DE, DK, EE, ES, Fl, GB, GE, HU, IS, JP, KE,
`KG,KP,KR,KZ,LK,LR,LT, LU,LV,MD,MG,MN,
`MW, MX, NO, NZ, PL. PT, RO, RU, SD, SE, SG, Sl, SK,
`TJ, TM, TT, UA, UG, US, UZ, VN, ARIPO patent (KE,
`LS, MW, SD, SZ, UG), European patent (AT, BE, CH, DE,
`DK, ES, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE), OAPI
`patent (BF, BJ, CF, CG, CI, CM, GA, GN, ML, MR, NE,
`SN, TD, TG).
`
`Published
`With international search report.
`Before the expiration of the time limit for amending the
`clllims and to be republished in the event of the receipt of
`amendments.
`In English transllltion (filed in Norwegian).
`
`(54) Title: A METHOD FOR IN1EGRITY MONITORING IN POSffiON DETERMINATION
`
`(57) Abstract
`
`In a method for integrity monitorin& in position determination by means of the Global Positioning System (GPS), especially by means
`of differential GPS (OOPS) or multi-station OOPS, wherein there is employed a monitor station (M) with a first GPS receiver, provided
`at a location with a known geographical position, GPS-based position measurement results are transferred on a communication line from
`the monitor station (M) to a user location (F) the integrity of whose geographical position has to be monitored by means of a second GPS
`receiver provided at the location (F). A space vector between the location of the monitor station (M) and the location (F) is determined
`by means of the position measurement results for the location (F) and the monitor station (M), whereupon the dearee of agreement is
`determined between the position for the location (F), the space vector and the known position of the monitor station (M), the degree of
`agreement being required to correspond to a maximum error in the position for the user location (F).
`
`Ex. PGS 1042
`
`
`
`FOR THE PURPOSES OF INFORMATION ONLY
`
`Codes used to identify States party to the Per on the front pages of pamphlets publishing international
`applications under the PCf.
`
`AM
`AT
`AU
`BB
`BE
`BF
`BG
`BJ
`BR
`BY
`CA
`CF
`CG
`CH
`Cl
`CM
`CN
`cs
`cz
`DE
`DK
`EE
`ES
`Fl
`FR
`GA
`
`Annenia
`Austtia
`Australia
`Barbados
`Belgium
`BwtiDa Fuo
`Bulgaria
`BeniD
`Bnzil
`Belanu
`Canada
`Central African Republic
`Congo
`Swiuerland
`COte d'lvoire
`Cameroon
`ChiDa
`Cz.ecboalovakia
`Czech Republic
`Germany
`Denmarlr:
`Bllonia
`Spain
`F"mland
`Prince
`Gabon
`
`GB
`GE
`GN
`GR
`HU
`IE
`IT
`JP
`KE
`KG
`KP
`
`KR
`KZ
`u
`LK
`LR
`LT
`LU
`LV
`MC
`MD
`MG
`ML
`MN
`MR
`
`United Kingdom
`Georgia
`Guinea
`GTeece
`HungiiiY
`Ireland
`Italy
`Japan
`Kenya
`Kyrgystan
`Democratic People's Republic
`of Korea
`Republic of Korea
`Kazakhstan
`LiechtensteiD
`Sri Lanka
`Liberia
`Lithuania
`LuxembourJ
`Latvia
`Monaco
`Republic of Moldova
`Madaguc:ar
`Mali
`Mongolia
`Mauritania
`
`MW
`MX
`NE
`NL
`NO
`NZ
`PL
`PT
`RO
`RU
`SD
`SE
`SG
`Sl
`SK
`SN
`sz
`TD
`TG
`TJ
`TT
`VA
`VG
`us
`vz
`VN
`
`Malawi
`Mexico
`Niger
`Netherlands
`Norway
`New Zealand
`Poland
`Portugal
`Romania
`Russian Federation
`Sudan
`Sweden
`Singapore
`Slovenia
`Slovakia
`Senegal
`Swaziland
`Chad
`Togo
`Tajikistan
`Trinidad and Tobago
`Ukraine
`Uganda
`United Swes of America
`Uzbekistan
`VietNam
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCf/N095/00239
`
`A method for integrity monitoring in position determination
`
`1
`
`5
`
`10
`
`15
`
`20
`
`25
`
`The invention concerns a method for integrity monitoring in position determi(cid:173)
`nation by means of the Global Positioning System (GPS), especially by
`means of differential GPS (DGPS) or multi-station DGPS, wherein a monitor
`station is employed with a first GPS receiver provided at a location with a
`known geographical position.
`
`It is normally desirable to have a high degree of certainty that a navigation
`aid really provides a performance which lies within the expected range of
`accuracy. A standard method is to employ a surplus of measurements in order
`to verify this, e.g. by means of statistic test methods. When using GPS this
`excess is dependent on the number of satellites which are visible to the user
`at any time, which at times means that the control of surplus measurements is
`reduced. The typical navigation user will have to compensate for this by
`employing greater safety margins which can lead to disadvantages and
`increased costs.
`
`In the context of entrepreneur activity where fixed measured points do not
`exist within reasonable proximity such as offshore, for activities such as, e.g.,
`the acquisition of seismic data or pipe laying, it will be of great economic
`importance to be able to verify the reliability of the positioning system within
`the same tolerance independent of the number of available satellites.
`
`A method is proposed herein which will make this kind of integrity moni(cid:173)
`toring possible.
`
`At present GPS is in the process of developing into a universal navigation
`system for use onshore, at sea and in the air. The system is owned by the US
`authorities and is placed at the disposal of all users who acquire a GPS
`receiver. The system is therefore very inexpensive for the user. Moreover the
`system has a global coverage which is quite unique in comparison with
`previous systems, while at the same time it offers the possibility of per(cid:173)
`forming extremely accurate position determination.
`
`30
`
`GPS will therefore be the natural choice of navigation system in the future
`for most people who require to navigate.
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCI'/N095/00239
`
`2
`
`It is highly desirable that the necessary control of the GPS application can be
`verified without incorporating extra systems since these will usually have a
`limited range of coverage as well as increasing the costs and thereby losing
`many of the advantages of using GPS.
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`A typical demanding application of GPS position determination is in con(cid:173)
`nection with the acquisition of marine seismic data. In this field so-called
`differential GPS is used in which correction data is broadcast from measuring
`stations located at known points. By utilizing these corrections the accuracy
`can be substantially improved and this kind of application is typical for
`accuracy requirements in the range 0.5 - 10 metres. In seismic data
`acquisition it is also normal practice to have buoys deployed in the water and
`these may also by towed by a vessel. The position of the buoys is also
`required to be known with a high degree of accuracy, and GPS is also a
`suitable system for performing such a determination. In this case a relative
`determination is made between, e.g., the boat and the buoy concerned.
`
`The fact that both a high degree of accuracy is required here and simul(cid:173)
`taneously a high degree of certainty that the actual performance should agree
`with the estimated or expected performance makes integrity monitoring
`extremely demanding. There is a requirement that errors which are not
`appreciably greater than the noise in the system should be detected and
`rendered harmless. The economic risk involved in faulty positioning of
`geological information and constructions is formidable. The proposed method
`will be an aid to solving this problem.
`
`The use of a navigation system will always be encumbered with errors. The
`coincidental errors, i.e. random noise, are relatively harmless in this con(cid:173)
`nection. The level of this noise is generally used to describe the accuracy of a
`system, and in the case of GPS the level of this noise is low. Such errors are
`easily dealt with by recognized calculation algorithms based on the least
`squares method or Kalman filtering.
`
`Since the noise level in GPS is low the error picture is dominated to a greater
`extent than for many other systems by model errors which usually take a
`systematic course. Systematic errors of this kind can have a very unfortunate
`effect on the position determination performed by algorithms of the above(cid:173)
`mentioned type. Great emphasis is therefore placed on detecting and
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCI'/N095/00239
`
`3
`
`removing such errors prior to the position determination. Model errors should
`be understood here to mean deviations between the mathematic model, on
`which the performance is based, and the real physical conditions which
`prevail during use.
`
`5
`
`GPS is a comprehensive system and the possible sources of error are there(cid:173)
`fore also many. The system is often considered on the basis of 3 main
`components:
`
`•
`
`*
`
`*
`
`a space segment, i.e. satellites
`
`a control segment, i.e. ground stations and control centres which are
`responsible for daily operation of satellites and signals
`
`a user segment, i.e. the user's receiver system with antenna and software.
`
`Errors can occur in the space segment in connection with, e.g., power supply,
`antennae, clocks, stability, etc. Such conditions will normally result in the
`satellite becoming unusable rather than introducing small errors which the
`user has trouble in detecting. The control segment monitors the satellites with
`regard to such errors and by using the satellite signals will warn the user
`about the condition. Of course it cannot be ruled out that an error with
`relatively little effect can occur here, with the result that the warning to the
`user arrives too late or not at all. The reporting mechanism which is
`employed can also cause it to take many minutes, and perhaps as much as an
`hour, between the detection of the error and notification of the users.
`
`In addition to monitoring of the satellites the control segment will perform
`measurements which will form the basis for calculating updated satellite
`orbits and clock operation. These data are programmed into each individual
`satellite of the control segment. There are, of course, many details which
`potentially can go wrong during this process, e.g. operator error, software
`error, error in connection with uploading of data and programmes to the
`satellites, measurement error on the ground stations as well as during transfer
`of these data to the control centre etc. The control segment works under
`stringent requirements with regard to procedures with continuous quality
`improvement. Even though it cannot be ruled out that errors can also occur in
`
`10
`
`15
`
`20
`
`25
`
`30
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCTIN095/00239
`
`this part of the process it is probably not the most conspicuous source of
`error.
`
`4
`
`In the user segment on the other hand the possibilities for error are many and
`the possibilities for detecting them correspondingly limited. Examples of
`errors which may arise are as follows:
`
`5
`
`•
`•
`•
`
`•
`
`receiver error
`
`software error
`
`measurement error as a result of phase error on account of multiple-path
`interference
`
`deviation in signal path as a result of ionospheric and tropospheric
`refraction.
`
`If differential GPS is employed the effects will also be felt of corresponding
`errors in the reference stations which are in use and the transfer of correction
`data therefrom.
`
`The first three types of error are cumulative, while the last is reduced to some
`extent by the fact that the error is common to the reference stations and the
`user, but this will be dependent on the distance between them.
`
`Thus it is the duty of the user to protect himself against any remaining errors
`from the space segment and the control segment in addition to the user
`segment-oriented error sources. In the case of differential use combatting the
`so-called .. Selective Availability .. (SA) will also take a similar form to the
`normal error sources. Intentional errors are built into SA in GPS in order to
`prevent unauthorized users from achieving ultimate accuracy. The
`unauthorized user naturally obtains no information on what the errors are, but
`information is supplied on what he can expect. The integrity control in
`relation to this will thereby take the form of an investigation into whether one
`is more influenced by SA than by 11What can be expected ...
`
`To sum up briefly it can be said that if the user does not check the system in
`relation to these and other possible error sources he does not know what
`
`10
`
`15
`
`20
`
`25
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCI'IN09S/00239
`
`5
`
`performance it gives, and the result is that he is using an unreliable
`navigation system. The conscientious user must therefore have a method
`available in order to perform such a check.
`
`5
`
`10
`
`The most common method for integrity monitoring is based on the use of
`statistical hypothesis testing. The initial hypothesis selected is normally that
`all measurements are distributed according to a known statistical distribution,
`usually normal distribution. It is !Q.en tested against a number of alternative
`hypotheses, in each of which there is assumed a specific error pattern.
`Standard alternative hypotheses are the occurrence of only one error at a time
`which leads to testing of as many alternative hypotheses as there are
`measurements.
`
`The weaknesses of the method consist in the fact that if the error pattern
`which actually occurs in a given error situation does not form part of any of
`the alternative hypotheses, it is not certain that the error can be detected.
`15 Moreover it is a purely statistical method where a given error only has a
`probability of being detected and identified. In other words there is also a
`certain probability that an error will avoid detection. These probabilities are
`influenced to a great extent by the ratio between measuring accuracy and the
`size of the error against which protection is required. The method begins to
`be highly impractical for most applications when the ratio approaches 1:5, but
`there is no absolute limit. This practical limit is in tum very dependent on
`geometrical strength in the position determination, since the poorer the
`geometry the less the probability of detecting a given error.
`
`20
`
`25
`
`The principle is based on control of the agreement of the measurements. If
`there is only one surplus measurement it will only be possible to detect that
`one error exists, but it is impossible to isolate it by means of this method.
`Thus it also becomes clear that if there is no surplus measurement, the entire
`agreement will become meaningless and errors cannot be detected.
`
`30
`
`The problem with an insufficient surplus of measurements or poor geometry
`can be remedied by combining GPS with other systems.
`
`GLONASS is a Russian satellite navigation system which works on the same
`principle as GPS. This will double the number of satellites available and is
`thereby expected to remedy this detail. The drawbacks with this method are
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCf/N095/00239
`
`6
`
`associated in the short term with the availability of receiver equipment for
`GLONASS signals. Present day receivers are expensive compared with GPS
`receivers. In the longer term there is a degree of uncertainty with regard to
`the availability of GLONASS signals.
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`INMARSA T is in the process of launching new versions of its com(cid:173)
`munication satellites. These have an inbuilt function enabling them to trans(cid:173)
`mit signals in approximate GPS format. These signals will carry additional
`information concerning the condition of the GPS satellites, but are also
`intended to be used to make additional measurements, and thus the effect will
`be the same as if there were an extra GPS satellite available. This will make
`1-3 extra measurements available, thus ensuring a higher minimum number of
`measurements. The disadvantage is that these signals are not yet available.
`Moreover it is worth noting that the satellites travel in geostationary orbits,
`which leads to a gradual deterioration in the coverage at higher latitudes and
`it becomes unavailable in the Arctic and Antarctic. Signals from satellites
`with a low angle of elevation are generally of poor quality due to the effects
`of refraction. Another phenomenon which can occur due to the fact that all
`these satellites lie on the same plane is that if the available GPS satellites
`should also happen to lie close to the same plane, the extra measurements
`will not provide any significant contribution to the geometry.
`
`Inertial navigation systems will also contribute with additional measurements,
`but of a different kind. These systems are not sensitive to fixed or slowly
`varying position errors, and are therefore unable to assist in exposing them.
`Moreover there is a substantial cost consideration to be taken into account
`when using inertial platforms.
`
`The above methods have the advantage that they all focus on the integrity of
`the user's position estimate which is the important factor in this context, but it
`can be problematic to strike a suitable balance between low probability of
`false alarm and high probability of detection. Low probability of false alarm
`also affords low probability of detection and vice versa.
`
`By providing a GPS receiver at a known point it will be possible to express
`with much more certainty the reliability of the system, and any errors can be
`demonstrated with a greater degree of certainty. This is a recognized principle
`for a so-called monitor station. If an application of differential GPS requires
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCI'JN095/00239
`
`7
`
`to be checked, this station can be supplied with the same differential
`corrections as the user receives. As long as this monitor determines its own
`position with a degree of error which is less than that against which pro(cid:173)
`tection is desired, it will be assumed that there is no error in the system. As
`soon as the error limit is exceeded the monitor will transmit a warning to the
`user via a communication channel suited to the purpose. This can often be
`arranged in such a manner that the warning is transmitted to the closest
`reference station which then passes it on to the user together with the broad(cid:173)
`cast of differential corrections. The user will thereby be able to receive the
`warning within a few seconds.
`
`However, this system has a substantial weakness in that the user's own
`measurements do not form part of the control. As indicated in the intro(cid:173)
`duction it is at this point that there is the greatest probability of errors
`occurring.
`
`The object of the invention is to provide a method which eliminates the
`disadvantages of known methods of integrity monitoring, this object being
`achieved by transferring GPS-based position measuring results on a com(cid:173)
`munication line from the monitor station to a user location whose integrity
`will be monitored by means of a second GPS receiver provided at the
`location, to determine a space vector between the location of the monitor
`station and the location by means of the position measuring result for the
`location and the known position of the monitor station, and to determine the
`degree of agreement between the position of the location, the space vector
`and the known position of the monitor station, the degree of agreement being
`required to correspond to a maximum error in the position of the user
`location.
`
`It is proposed to employ a GPS receiver positioned at a known point in the
`same manner as for the position monitor described above, but the measure(cid:173)
`ments which this performs are used in a different way.
`
`It is recommended that integrity monitoring by means of hypothesis testing
`should always be used, since the only drawback this entails is that sufficient
`calculating power must be made available in order to perform the calculations
`continuously. The proposed method will constitute a more stringent extra
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCf/N095/00239
`
`8
`
`control, and its greatest importance will naturally lie in filling in the gaps in
`the first method.
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`Quite simply, the method is based on the fact that the monitor station trans(cid:173)
`mits its measurement results to the user via a communication line suited to
`the purpose. The user can then combine these measurements with his own in
`order to determine a space vector between himself and the monitor station.
`· The user will then be able to chec~ that there is agreement between his own
`position, which requires to be controlled, the determined space vector and the
`monitor's known position. The user has previous knowledge of the monitor's
`exact position. The discrepancy which can be tolerated can be set as being
`equal to the position error against which protection is desired. The position
`gap can be determined by means of a simple solid geometric observation or
`by geodetic calculations.
`
`The method is particularly well suited to a vessel for marine seismic data
`acquisition which also uses GPS as a navigation aid. This will be shown in
`the following description, but the method is not restricted to this area of
`application. This special application, however, is the basis for the further
`detailed description.
`
`The invention will now be explained in more detail with reference to the
`drawing, wherein fig. 1 illustrates a relevant navigation situation with
`position determination, and fig. 2 an example of the siting of a monitor
`station in relation to user location and one or more reference stations.
`
`Fig. 1 illustrates how a vessel F navigates by means of signals from GPS
`satellites G combined with differential corrections broadcast via a broad-
`casting medium K from one or more reference stations R which are arranged
`in known position(s). The vessel F which is equipped with a GPS receiver
`together with a receiver for the broadcasts from the reference station( s) R
`performs the navigation by means of so-called differential GPS (DGPS)
`which is a known method.
`
`The vessel F tows one or more buoys B whose position is determined relative
`to the vessel F by a suitable GPS-based technique. For this purpose the
`buoy(s) B is equipped with a GPS receiver and the equipment required for
`telemetering measurement data for the vessel F.
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCTJN095/00239
`
`9
`
`As shown in fig 1, a GPS receiver has been provided in a monitor station M
`at a co-ordinate-determined point. In addition the monitor is equipped with
`the necessary equipment for communicating via a suitable medium T with the
`vessel F. It is also desirable for the vessel F to be equipped with an extra
`GPS receiver in a secondary measuring station I on board. The reason for this
`is described below in connection with the method.
`
`The vessel determines its own position by means of signals from the satellites
`G and the reference station(s) R according to the method for differential GPS
`(DGPS). The space vector from the vessel F to (each of) the buoy(s) B is
`determined by means of the same signals from the satellites G observed from
`F and from B. The measurement results from the observation of the signals
`on board B are transferred to F through the established telemetry channel.
`Both sets of measurements are transferred to a computer on board the vessel
`F. The measurement differences between F and B for the same satellites
`provide a data set which indicates distance and clock error differences
`between F and B in relation to each of the satellites G concerned. On the
`basis of known GPS technique it is then a commonplace task to derive the
`space vector and the clock difference between F and B.
`
`The problem now is how the integrity of this position determination can be
`monitored if none of the known methods mentioned above is adequate. First
`and foremost the problem is relevant to the position determination ofF since
`
`*
`
`*
`
`the use of buoys is restricted to special operations such as seismic data
`acquisition;
`
`in the case of seismic data acquisition the integrity of the buoy(s) will
`also be controlled by means of other necessary (underwater) positioning
`systems such as, e.g., hydroacoustic measurements and magnetic com(cid:173)
`passes.
`
`The integrity monitoring of the position determination for F can be performed
`by transferring measurements from the monitor station M to the vessel F by
`means of the communication connection T. The information which is trans(cid:173)
`mitted from M to F is of exactly the same type as that which is normally
`transmitted from B to F. The measurements F received from M are entered
`into the same computer as mentioned above. Thereupon exactly the same
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCI'IN095/00239
`
`10
`
`calculation is performed for the monitor M as that which is performed for the
`buoy(s) B. This means that a space vector is determined between F and M. If
`all measurements, differential corrections, data transfers, calculation models
`and software programmes are correct, the vessel's F position together with the
`space vector will be able to generate a position for the monitor M which
`agrees with the previously known coordinates for M. This position can be
`calculated by means of a solid geometric observation or by a geodetic
`calculation.
`
`It should be noted here that the determination of the space vector between F
`and M is performed completely independently of the reference station(s)' R
`measurements and its (their) calculated differential corrections. A complete
`control is thereby obtained of the DGPS infrastructure as it appears to the
`user. This is a distinct improvement in relation to the above-mentioned
`position monitor which is installed at a different location from the user and
`operates under independent conditions. The proposed method does not appear
`to have any weaknesses in relation to the prior art.
`
`There is, however, a weakness in relation to the problem of monitoring the
`integrity of the vessel's F position. Theoretically it is quite possible to
`generate a correct position for M according to the described method even
`though the position for F has a moderate error. This situation can arise if
`there is a measurement error in the observation of the satellite signals from
`the satellites G on board F. These measurements are used both to obtain the
`position for F and to determine the space vector between F and M. On closer
`consideration it will be revealed that these two elements will be encumbered
`with equally substantial errors from any measurement error of this kind. The
`effect on position and space vector respectively will have opposite signs,
`thereby cancelling the error when the position forM is derived. However, this
`problem is easily remedied by arranging an additional GPS receiver I on
`board F. Duplication of the receiver and antenna system will provide direct
`protection against errors in this equipment, but if the antennae are located at a
`distance from each other which is sufficient to achieve reasonably indepen(cid:173)
`dent observation conditions (e.g. 5 metres or more), this will also provide an
`efficient protection against multiple-path interference which is considered to
`be the most frequently occurring source of error for a GPS user. Like the
`measurement from M the measurements performed by I are entered into the
`computer, and a space vector is calculated in a similar fashion. This space
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCI'/N095/00239
`
`11
`
`vector can then be compared with the known vector which can be measured
`between the antennae on board by other means, e.g. by a tape measure and
`use of the vessel's F navigation compass.
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`One of the greatest sources of error in connection with DGPS is refraction of
`the GPS signals through the ionosphere. The variations in this effect are great
`enough to introduce substantial errors in relation to generally known models
`of this effect. Potentially this could lead to an integrity problem if there are
`stringent requirements. The simplest way of avoiding this is to use two(cid:173)
`frequency GPS equipment. Equipment of this kind is much more expensive
`for the user. Moreover all publicly available reference station services of any
`extent are based only on single-frequency equipment, thus eliminating much
`of the benefit. There is of course nothing to prevent both the monitor station
`M and the vessel F from being equipped with two-frequency equipment, thus
`permitting a more accurate vector to be determined virtually unaffected by
`ionospheric errors. In this case an even more efficient integrity monitoring
`would be achieved. However, another proposal is presented here which
`enables this increase in costs to be avoided.
`
`The influence of the ionosphere on the GPS signals is correlated over
`relatively great distances (> 1000 km). This is exploited to the advantage of
`DGPS as any remaining errors are largely cancelled between reference station
`and user. However, it is the integrity of this detail which has to be monitored
`since it is based on a statistical assumption. By positioning the monitor
`station Min relation to the user F and the reference station(s) R in such a
`manner that the distances F - M and F - R are less than M - R, the advantage
`is achieved that the expected correlation between F and M is greater than
`between R and M. This fact makes the vector F - M suitable for checking the
`accuracy of the ratio between R and F. See fig. 2 of the drawing.
`
`The chosen position of the monitor M in relation to the reference station R
`provides a favourable operational area for the vessel F which is defined by
`the two circular arcs with centres in the points M and R respectively and with
`radius equal to the distance between M and R.
`
`If a network of reference stations R is employed, in order to make optimum
`use of DGPS with regard to accuracy it will normally be desirable to arrange
`the reference stations R in a circle around the operational area for the vessel
`
`Ex. PGS 1042
`
`
`
`W096/21163
`
`PCI'IN095/00239
`
`12
`
`F. Unless the extent of the network of R is extremely large (e.g. con(cid:173)
`siderably larger than 1000 km), ionospheric refraction will no longer be a
`great threat. In the case of such a network of R, within which F operates, and
`regardless of size, the optimum position for M will be located within the
`same network, but then the condition illustrated above naturally cannot be
`fulfilled for all reference stations at the same time. On the basis of a pure
`consideration of integrity it will be best to have M as close to F as possible,
`but this is not really critical.
`
`The best way of performing the control is to have the position of the monitor
`station continuously estimated by the computer which also compares the
`estimate with the known coordinates. In its simplest form it can be performed
`in such a manner that the computer transmits an alarm to the user when a
`predetermined limit for position error is exceeded. It should thereby be
`possible to expose errors which are less significant than those detected by
`utilizing the statistic hypothesis testing alone.
`
`Nor is there any reason why the method of hypothesis testing should not be
`extended to also help to find the error source after a significant error vector
`has been detected by means of the monitor's measurement data. The actual
`hypothesis formulation will then have to be modified in order to utilize the
`knowledge obtained by including the monitor, but the object will be the same,
`viz. to find the most probable error measurement in order to remove it. A
`major advantage of this method in relation to that described above is that
`identification is possible even though the vessel F does not have access to
`any surplus measurements apart from those the monitor has performed. A
`natural consequence if F is already operating with a minimum number of
`measurements is that it is only possible to expose and correct the error in the
`measurement value, but there is then no longer a suffici