The present invention relates in general to positioning of mobile equipment by use of satellites and in particular to such positioning assisted by land based communication nodes.
In recent years, determination of the geographic position of an object, equipment or a person carrying the equipment has become more and more interesting in many fields of application. One approach to solve the positioning is to use signals emitted from satellites to determine a position. Well-known examples of such systems are the Global Positioning System (GPS) and the GLObal NAvigation Satellite System (GLONASS), see e.g. [1]. The position is given with respect to a specified coordinate system as a triangulation based on a plurality of received satellite signals.
A stand-alone GPS receiver can obtain full locking to GPS satellite signals, without having any other information about the system except nominal carrier frequency and the rules by which data carried by the signals are modulated. Basically, the three-dimensional position as well as a receiver clock bias to the satellite time have to be determined in the position calculation step. However, such a start-up procedure from basically no prior information at all takes time and requires typically large computational efforts. By increasing the initial knowledge of the system, the locking procedure can be speeded up and simplified. Assisted GPS (A-GPS) technology is an enhancement of GPS, where additional information can be provided to the GPS receiver in order to facilitate the locking-on procedures. If the GPS receiver is connected to a cellular communications system, additional assistance data can be collected from the cellular communication system directly. This typically enables a rough initial estimate of the position of the receiver together with a corresponding uncertainty of the initial estimate. Furthermore, information about the approximate satellite system reference time as well as information about which satellites that are above the horizon can be provided.
When satellite signals are acquired, see e.g. [2], acquisition has to be performed in a carrier dimension, handling different Doppler shifts, as well as in a code (or range) dimension. Searching the entire carrier-code space for acquiring the satellite signal is a time-consuming process. Fine time assistance means that the GPS receiver is provided with highly accurate information related to the global GPS time and satellite positions in space. Any assistance data that might reduce the search window size will improve the process.
In the U.S. Pat. No. 6,429,815, a method and an apparatus for determining search centre and size in searches for transmissions from GPS satellites is disclosed. In a particular well defined situation, where the distribution of the position of the mobile terminal has a circular symmetry centred around a base station, a search window centre and search window size that is optimal for that particular situation can easily be determined by simple relations. The disclosure further states a wish or an assumption that further wireless communications system data can be used to further refine the definition of the search window. However, since such data removes the circular symmetry of the particular situation previously discussed, the described approach in connection with this can not be applied on cases further relying on this kind of data. Moreover, no further description of how to enable anyone skilled in the art to perform such refining of the search window based on such wireless communications system data is given.
It is thus since long known from prior art that there is a pronounced need for improving the search window position and/or size upon GPS positioning, but no general solutions are publicly available within prior art.
A general object of the present invention is to provide improved methods and devices for satellite based positioning with assistance data. A further object with the present invention is to reduce the computational efforts needed for obtaining code phases of signals transmitted from satellites. Yet a further object is to optimally reduce a search window based on available assistance data even in non-symmetry situations.
The above objects are achieved by methods and devices according to the enclosed patent claims. In general words, one upper and one lower bound on the code phase of a signal transmitted from a specific satellite can be computed for terminals that reside anywhere in a closed region, having a non-circular symmetry, obtained by an initial positioning step. A position is then determined using search windows having such upper and such lower bound for at least one satellite. The upper and lower bounds are provided using satellite position data in three dimensions, satellite time reference data as well as geometric information about the closed region of the initial positioning. If the location where the satellite time reference data is provided is located within the closed region, the search window lower limit is preferably determined to be equal to an estimated code phase shift at that location minus an uncertainty of the satellite time reference data. If the location where the satellite time reference data is provided is located outside the closed region, the search window lower limit is preferably determined to be equal to the minimum estimated code phase shift at the boundary of the closed region minus an uncertainty of the satellite time reference data. The search window upper limit is preferably determined to be equal to the maximum estimated code phase shift at the boundary of the closed region plus an uncertainty of the satellite time reference data.
The invention also discloses devices and arrangements for performing the above procedures.
An advantage of the present invention is that the computational complexity in satellite-based positioning is reduced regardless of the system symmetry. The reduced complexity can be utilised to enhance the positioning sensitivity or to reduce the power consumption during the positioning or a combination thereof.
The invention, together with further objects and advantages thereof, may best be understood by making reference to the following description taken together with the accompanying drawings, in which:
a is a diagram illustrating the relation between GPS time and cellular frame times experienced at different positions in a system;
b is a diagram illustrating the relation between GPS time and GPS frame times experienced at different positions in a system;
In the detailed description below, embodiments implemented in a GPS system are illustrated. However, anyone skilled in the art realises that the corresponding principles can be applied in any satellite based positioning system, such as the GLONASS or the coming European Galileo satellite navigation system.
Likewise, in the detailed description below, WCDMA systems will be used as model systems. However, the present invention is also applicable on other wireless communications systems. Non-exclusive examples of other systems on which the present invention are e.g. the CDMA-2000 system or the GSM system. When applied to other wireless communications systems, the implementation of the different functionalities will be done in different terminals and nodes of such systems.
The term “mobile terminal” is in the present disclosure used to denote any kind of terminal that can be transported within a wireless communications system. Non-exclusive examples are telephones, personal digital assistants and portable computers.
The received data 22A-E from the satellites 20 also comprise ephemeris data, i.e. among other things a satellite orbit prediction. It is also possible to use the so-called GPS almanac, which also provides a basis for determining satellite positions. Assistance data 30, comprising satellite position data and satellite time reference data, is in this particular example sent over a reference receiver interface 36 to a Radio Network Controller (RNC) 15. A satellite positioning interface 13 receives this data and may e.g. determine which satellites might be in such positions that their ranging signals 22A-E are probable to detect.
When a positioning request occurs, e.g. in a core network 16 of the communications system 1, the positioning request 32 is provided to the RNC over a RANAP interface 34 (Radio Access Network Application Part). In an alternative embodiment, an external positioning node could be connected to the RNC, e.g. over an Iupc interface. The Iupc interface is a logical interface for the interconnection of standalone A-GPS SMLC (Serving Mobile Location Center) and RNC components of the UTRAN (Universal Terrestrial Radio Access Network) for an UMTS system, see e.g. [4]. The RNC creates control signalling ordering measurements of satellite ranging signals 22A-E and sends the control signals 12 over a RRC interface 38 (Radio Resource Control interface) to the mobile terminal 10. The measurement order is accompanied by assistance data, typically processed in the satellite positioning interface 13. The mobile terminal 10 is equipped with a receiver that is capable of detecting satellite ranging signals 22A-E and the mobile terminal 10 uses the assistance data to facilitate the locking on and measuring of the satellite ranging signals 22A-E. The measured ranging signals are then used to calculate a position of the mobile terminal 10 according to standard satellite positioning procedures. If user equipment based A-GPS is used, the processing of the ranging signals is performed in the mobile terminal. If user equipment assisted A-GPS is used, the ranging signals or representations thereof are sent to the RNC, where the processing for purposes of positioning is performed. The use of fine time assistance data allows the satellite receiver of the mobile terminal 10 to obtain the best sensitivity possible. Fine time assistance data is a relatively vague expression. The meaning of fine time assistance in the present disclosure is time reference assistance having an accuracy typically in the order of some tens of microseconds. The order of magnitude of the accuracy has to be considerably less than the GPS C/A (Coarse/Acquisition) epoque, which has a duration of 1 ms, if GPS is used. The coordinates used in satellite positioning systems, and in particular GPS, are normally based on an earth centred coordinate system.
GPS is a code division multiple access (CDMA) system. The GPS signal from each satellite is hence associated with a specific code. The chip rate of this code being 1.023 MHz for the civil coarse acquisition (C/A) signal. The signal from each satellite is retrieved by correlation against the unique code of each satellite. This code has a duration of 1023 chips (exactly 1 millisecond). A further complication is now that a 50 Hz bit stream is superimposed on the GPS ranging signals from the satellites. These GPS message bits contain information that the GPS receiver would have needed in order to calculate its position in case that assistance data would not be available from the cellular communications system. The bit edges complicate ranging correlations since the unknown switches of sign at the bit edges deteriorate correlation receiver performance in case the exact time instances of the bit edges are not known. Until accurate synchronisation to GPS time has been established in the GPS receiver, coherent correlation over more than 10 milliseconds is hence not possible. This fact reduces performance significantly when the first satellite is acquired since the assisted GPS receiver sensitivity is reduced with 5-10 dB since incoherent correlation needs to be used. The remaining satellites do not suffer from this sensitivity loss since they can exploit the synchronisation to GPS time obtained as a consequence of the detection of the first satellite. To conclude, the first and most important benefit of fine time assistance is that it allows the assisted GPS receiver to apply coherent correlation detection also for the first satellite it acquires.
Other advantages associated with fine time assistance is that it allows the correlation search window to be reduced in the code dimension more than a factor of 10 as compared to the complete 1023 chips code epoch of the GPS ranging signal. GPS correlation receivers search a two-dimensional code and Doppler space due to the large variation of the relative speeds of the satellites. This search window reduction results in an additional assisted GPS sensitivity improvement since there are less code and Doppler search bins that can result in false alarms of the receiver. This gain is however relatively small. Calculations indicate that it is of the order of 0.1-0.5 dB depending on the assumptions. More importantly, the reduced search window sizes reduce the computational complexity of the GPS receiver proportionally, a fact that translates into the possibility to correlate for longer periods of time to enhance sensitivity, or to reduce the computation time, thereby also reducing the power consumption. The latter benefit may be substantial in cases where the assisted GPS receiver is used for satellite acquisition purposes during extended periods of time. Note that the benefit of a reduced search window is always present when new and undetected satellites are searched for.
The present invention relates to the determination of the search window used in the code and Doppler correlation search step in order to achieve an always optimised window alignment so that search windows of minimal size can be used in the GPS signal acquisition. This information can also be used to select the first satellite to search for when establishing GPS time, so that the best achievable GPS receiver sensitivity is obtained for this satellite.
In order to determine a distance between a receiver and a satellite, the receiver has to have knowledge about the time instant when the transmitter transmitted the signal. In a system having access to assistance data, an approximate system time can be provided. However, since the mobile terminal to be positioned typically is placed at a distance from the node providing the time difference, durations for transferring time references have to be compensated.
In
The receiving terminal time scale is as seen in
There are also other alternatives than time stamping. One such alternative that is under discussion is to use the terminals to determine the relation between GPS time (code phase) and a defined, periodically repeated, transmission instant of the cellular communication systems ordinary transmission. Terminals of opportunity that perform assisted GPS positioning would then report this information to the cellular communications system for further distribution to other users.
The principles above are intended to allow a GPS receiver of a mobile terminal to make an alignment of correlation search windows and measured GPS signals in the best way possible. The satellite signals of each GPS satellite are retrieved by correlation against a unique code. Since the position of the mobile terminal is not known exactly to the GPS receiver, an additional effect affects the search window alignment with respect to the received signal from each GPS satellite. Briefly, the unknown location of the terminal implies that the GPS code phase received in the GPS receiver of the terminal may be early or late with respect to what is experienced in the reference site, e.g. the radio base station.
It is now clear that the distribution of fine GPS time assistance, e.g. by using the frame structure of the cellular communications system, will introduce variations when aligning GPS code phase search windows of terminals to the cellular communication system. It is a request to reduce the size of the search window as much as possible, since the computational effort scale proportionally to the search window size.
In the U.S. Pat. No. 6,429,815, a particular situation is considered in detail, where there is additional information available about the distance between the mobile terminal and the radio base station. In other words, the time difference Δ1 is known, and the mobile terminal is situated somewhere at a circle centred at the base station. With such geometry, also the estimation of the possible extremes of Δ2 becomes simple. Since an entire circle is considered, there are always two points at the circle that are situated in the same vertical plane as the satellite and the base station. These two points correspond to the two extreme cases of Δ2, and can easily be calculated being dependent on the cosine of the satellite elevation.
However, when the circular symmetry is broken and/or the uncertainty of the distance between the mobile terminal and the base station is relatively large, such reasoning is not applicable. It can be shown that with non-circular symmetry of the area in which the mobile terminal can be located, the necessary minimum search window can vary considerably. Examples are shown in Appendix 1. It is not obvious from any prior art and in particular not from U.S. Pat. No. 6,429,815 how a general minimisation valid for any shape or size of the area in which the mobile terminal can be located is to be carried out.
In the present invention, the search window used in the code and Doppler correlation search step for the registered satellite ranging signal is determined by using information of e.g. cell geometry or other initial position information together with calculated satellite positions. This allows an optimised search window alignment so that the search window of minimal size can be used in the GPS signal acquisition. The optimised search window is achieved by finding a search window lower limit that is as high as possible, but still ensured to be less than or equal to the actual code phase shift for the registered satellite ranging signal. Similarly, a search window upper limit is found, which is as low as possible, but still ensured to be larger than or equal to the actual code phase shift for the registered satellite ranging signal.
Additional assistance data is collected from the cellular communication system directly, typically to obtain a rough initial estimate of the position of the terminal together with a corresponding uncertainty of the initial estimate. This position is often given by a so-called cell identity positioning step, i.e. the position of the terminal is determined with cell granularity. This is schematically illustrated in
Alternatively, a more accurate position can be obtained by measuring the travel time of radio waves from the serving radio base station 14 to the terminal 10 and back, thereby establishing a region 42 at a certain approximate distance from the serving radio base station 14 where the mobile terminal 10 must be located. In WCDMA this is denoted round trip time (RTT) positioning. The result of the positioning is reported in terms of an arc 42 with the centre in the serving radio station 14 site coordinates. The thickness of the arc 42 is due to measurement uncertainties. If the thickness of the arc 42 is large compared to the required final positioning accuracy or if the arc 42 is smaller than 360 degrees, prior art methods for determining search windows can not be applied to provide an optimum search window.
In
As mentioned above, an initial determination of the closed area within which the mobile terminal is situated is performed. In a particular embodiment, the closed area is a cell polygon that describes the extension of the cell. The coordinate system is normally based on the WGS84 earth model and polygon corners are usually given as a list of latitude, longitude values that comprise the coordinates of each corner of the polygon.
Satellite ephemeris data and satellite time information are then collected from a reference node. Ephemeris data for the GPS system is described in e.g. [3]. Using ephemeris information, the position of all satellites can be computed in the WGS 84 earth centred coordinates, using the present updated satellite system time. The corners of the cell polygon and the position of the satellites may be transferred to an earth tangential coordinate system, typically centred somewhere in the cell in question, as discussed earlier in connection with
In a particular embodiment, a number of test points to be used for the calculation of the search windows are spread out in the closed area where the mobile terminal is known to be located initially. In case the initial positioning step resulted in a cell polygon, the test points are selected on the cell polygon boundary, including the corner points. This is due to the fact that only points on the polygon boundary or at the radio base station site are relevant in the determination of the search windows. This is formally proven in Appendix 2. In practice a finite number of test points may be spread out along the boundary of the region. An important consequence of this is, however, that the complexity of the calculations is reduced significantly, as compared to a search extending also over the interior of the closed area. These test points represent tentative terminal positions that are to be tested for the satellite ranging signal arrival time from each satellite, as discussed further below. The set of test points is denoted {riTEST}i⊃IN. Note that the above is true for all geometries, i.e. also for the circular arc, the test points need only be spread out on the actual boundary. One may conjecture that the number of possible points could be refined further to only include the corners of the polygon.
The next step comprises a calculation of lower and upper limits on the satellite code phase experienced by terminals in the closed area, and in the present embodiment, these limits are calculated using the test points. Towards that end it is noted that the total code phase variation that needs to be accounted for is the sum of three terms as follows:
ΔΦ=ΔΦTimeStamp+ΔΦCellular Propagation+ΔΦGPS Propagation.
Here the first term represents the uncertainty caused by the time stamping of the (future) cellular frame event in the serving radio base station. The first term has a size limited as follows (c.f.
|ΔΦTimeStamp|≦δ
as expressed in GPS C/A code chips. The second term affects the uncertainty in frame start as explained by
where c denotes the speed of light, ΦGPS denotes the GPS C/A code chip rate, ri denotes the vector pointing to the terminal location, rs denotes the vector pointing to the radio base station site and where ∥ ∥ denotes the Euclidian length of a vector (i.e. normal distance). Note that ri is not known, the procedure of the invention rather aims at minimising search windows using the fact that ri is somewhere within a pre-determined area. The third term reflects the effect of
Here ri denotes the vector that points to the i:th satellite position in the earth tangential coordinate system.
The objective of the invention is now to compute minimal search windows that still guarantee that the actual code phase of the GPS satellites can be found somewhere in the search window. This requires that the following two quantities are determined:
when ri varies.
The quantity
is determined by insertion of all test points {riTEST}i=IN in the equations for the terms of ΔΦ above, followed by a selection of the point and value that renders the highest value. The test point selection rests on the understanding that the maximum code phase difference is attained on the boundary of the cell polygon, which is valid in all cases where the initial area is a closed polygon area, and in case the distance to the satellites is much larger than the extension of the initial area where the terminal is known to be located. Mathematically, this can be expressed as:
The proof behind this is found in Appendix 2.
Note that in the case the closed area is limited by circular arc sections, this alternative can be seen as a limiting case as being defined by a polygon with an infinite number of corners. Hence the result for that case is that
is attained on the circular arc boundary.
The quantity
is attained in the serving radio base station site coordinates in case these coordinates are in the interior or on the boundary of the cell polygon. Mathematically, this is expressed as:
In case the serving radio base station coordinates are outside of the cell polygon,
is attained in a point on the boundary of the cell polygon, i.e.:
A proof for this is also given in Appendix 2.
After testing of all boundary test points {riTEST}i=IN, the following maximising and minimising points
Note that in case the radio base station site is in the interior of the initial region, then
min ΔΦ=−δ
The resulting code phase search window then becomes:
[min ΔΦ, max ΔΦ]
as expressed with respect to the code phase corresponding to tGPS
At this point in time it is suitable to mention that there are two types of A-GPS positioning. One type, mobile terminal based A-GPS, performs the positioning calculation in the mobile terminal. The other type, mobile terminal assisted A-GPS performs only ranging measurements in the mobile terminal. The position is calculated in a node of the cellular communication system using the code phases measured in the mobile terminal. In WCDMA, these are denoted UE based A-GPS and UE assisted A-GPS, respectively. The procedure discussed in this disclosure is applicable to both types of A-GPS. The main difference is if the search window alignment is performed in the cellular communications system positioning node or in the mobile terminal. Embodiments of both cases are presented further below. Note that alignment in the terminal can be achieved in case it is provided with fine time assistance as well as in situations where fine time assistance data is not available. In the latter case the mobile terminal has acquired a first GPS satellite and is hence synchronised to the GPS time.
It is possible to comprise an initial positioning unit 62 in the RNC 15, providing a coarse mobile terminal position in the form of a closed area within which the mobile terminal 10 is known to be present. In one embodiment, this is a cell identity positioning unit, providing the definition of the cell to which the mobile terminal 10 is associated. Such closed area data is provided to a coarse position receiver unit 64 in the mobile terminal 10. Such an embodiment is, however, at the moment not supported by the present WCDMA standard, but is nevertheless easy to implement if necessary.
In an alternative particular embodiment, the initial positioning unit 62 is a unit separated from the RNC 15. The coarse mobile terminal position is then provided to the coarse position receiver unit 64 e.g. comprised in general control signalling data if the initial positioning unit 62 still resides within the communications system itself. The coarse mobile terminal position could also be provided as a data packet sent to the mobile terminal over the data plane. This could e.g. be convenient if the initial positioning unit 62 is not controlled by the communications system operator.
The mobile terminal 10 is now provided with all data necessary for making an optimisation of the search window. This data comprises three-dimensional satellite position data, satellite time reference data and data defining the closed area. The adaptation of the search window to a specific satellite is performed in a processor 60 connected to the means for providing assistance data 56 and coarse terminal position 64. The processor 60, the means for providing assistance data 56 and the coarse position receiver unit together constitute an arrangement 63 for assisting in determining a position for a mobile terminal 10. The adapted search window is then used by a satellite ranging signal registering unit 54, connected to a GPS receiving antenna 52, for obtaining the ranging information from the satellite with minimum efforts. The satellite ranging signal is then utilised for determining a mobile terminal position in a positioning unit 70. Such determining is described in e.g. [5].
The result of the positioning is then typically sent via the RNC to the core network of the communications system. The satellite ranging signal can be combined with other satellite ranging signals or any other positioning information, such as e.g. measured ranges to different radio base stations within the mobile communications network. Such position determination is known as such in prior art and will not be discussed in any details in this disclosure.
From
In
In another embodiment, also the fine time assistance part 21 may be separated from the RBS 19 position. In such a case, the fine time assistance part 21 has to be provided with an antenna system that can be listening on the radio signals of the communications system and thereby determine the cellular frame time reference. If the separation between the RBS 19 and the fine time assistance part 21 is significant, such a measured cellular frame time reference has to be compensated for the travelling time between the RBS 19 and the fine time assistance part 21.
It is even possible to use another mobile terminal as the fine time assistance part 21 of the satellite reference node 18. If this mobile terminal is locked to the satellite positioning system and has a well established position as well as a correct satellite reference time, GPS time is readily available and can be distributed to other mobile terminals as assistance data. However, if the satellite reference node 18 is mobile, particular care has to be taken to correct for any distance offsets regarding the position of the satellite reference node 18 relative the radio base station 19 site.
In
The split configuration of the reference satellite node 18 as described in connection with
In
The split configuration of the reference satellite node 18 as described in connection with
It is of course also possible to have parts of the arrangement for position determination situated within other nodes of the mobile communications system, either entirely or in a distributed manner. The RNC implementation in the embodiments described above should only be regarded as a non-limiting example of where the parts could be arranged.
In the embodiments above, it is implicitly assumed that the data that is transferred forth and back between the communications network and the mobile terminal utilises different types of control signalling, i.e. the data is transferred at a control plane of the communications network. However, there are also alternative ways for communicating data. The data may e.g. be communicated as data packets, i.e. as unspecified bit streams, at a user plane of the wireless communications system. This may be even more attractive if the satellite reference node and/or parts of the positioning system are more separated from the actual communications network.
The main steps of an embodiment of a method according to the present invention is illustrated in a flow diagram in
The basic idea of the invention is to compute optimally small satellite code search windows, for use in the code and Doppler search step of the detection of satellite signals in satellite ranging signal receivers. This is achieved by accounting for the detailed geometry, e.g. the cell polygon, of the region were the terminal is known to be located when positioning is started. Furthermore, the exact 3D locations of all satellites are accounted for. The result is an optimally small code search window, for each individual satellite.
More specifically, the invention relates to the determination of assistance data in the cellular communication system, that is required to provide the satellite signal receivers in mobile terminals with so called fine time assistance. Briefly, fine time assistance means that the satellite signal receiver is provided with highly accurate information related to the global satellite system time and satellite positions in space. Together with the assistance data, upper and lower bounds on the code phases of signals transmitted from all satellites can be computed for terminals that reside anywhere in the region obtained by the initial positioning step. This follows since the times of transmission of the signals from the satellites are synchronised with extreme precision, and since the orbits of these satellites can be calculated in the cellular communication system using other types of assistance data obtained from reference receivers.
The embodiments described above are to be understood as a few illustrative examples of the present invention. It will be understood by those skilled in the art that various modifications, combinations and changes may be made to the embodiments without departing from the scope of the present invention. In particular, different part solutions in the different embodiments can be combined in other configurations, where technically possible. The scope of the present invention is, however, defined by the appended claims.
The purpose of the example below is to illustrate the gains that may be achieved by the present invention. The calculations of this example are based on the geometry of
The intention is to illustrate the variation of the search window size as a function of both the azimuth and elevation of the satellite, for a specific cell polygon and for one interior and one exterior site location. Noting that the distance from the origin of the earth tangential coordinate system to the satellite is the only unknown distance it needs to be solved for. This can be done starting with the vector relation
R
1
=R
ε
+R
S−1
Taking the dot product of this equation with itself and exploiting the geometry results in
R
1
2
=R
E
2
+R
S−1
2+2RRS−1 sin(α).
Solving for the unknown results in
R
S−1
=−R
E sin(α)±√{square root over (R12−R2 cos2(α))}
where only the positive sign applies. Using RS−1 the following vector to the satellite results in the earth tangential coordinate system
r
i=(RS−1 cos(α)cos(β)RS−1 cos(α)sin(β)RS−1 sin(α))r,
where β denotes the azimuth.
Note: This corresponds to an east-north-up coordinate system.
The corresponding site coordinates are
r
s=(xs ys 0)r,
while the cell polygon coordinates are
r
ei=(xci yci 0)T, i=1, . . . , N.
All needed quantities are now at hand for the evaluation.
A rural cell is treated. The test points are selected as the corners of the rural cell polygon in this part of the example. The numerical quantities of Table 1 were used:
The cell polygon and the site positions are plotted in
Some comments are in order.
When the elevation approaches 90 degrees the search window size becomes constant as a function of the azimuth as it should.
The maximum search window size occurs when the main cell area is between the site and the satellite in azimuth. This follows since then the radio signals of the cellular communication system and the radio signals from the GPS satellite meet, thereby maximising the code phase mismatch within the cell area. The GPS reference time is taken in the radio base station site.
The minimum search window size occurs when the site is between the GPS satellite and the main cell area. This follows since then the radio signals of the cellular communication system and the radio signals from the GPS satellite travel in approximately the same direction, thereby minimising the code phase mismatch within the cell area.
The maximum and minimum search windows occur for low elevations. The reason is that the GPS radio signal in such a case travels almost parallel to the surface of the earth.
The behaviour is similar for interior as well as exterior sites.
From the figures above it is clear that the required search window size for large ranges of satellite azimuth and elevation allow far smaller search windows than what is required when existing technology is used. With most prior-art methods, the maximum search window size needs to be used for all satellite positions. In order to assess the gains, the average search window size computed from the
dA(α)=diameter×height=2πr cos(α)×rdα.
Dividing with the area of a half sphere, it is clear that the distribution can be written as:
fα,β(α,β)=C cos(α).
The constant can be determined by the normalising relation:
The formula for calculation of the expectation of the search window size hence becomes:
Here Window(α,β) is the quantity displayed in
Using the formula for the expectation the following values were calculated for each of the figures and they are displayed in Table 3.
Obviously, A-GPS complexity can be reduced by more than ⅓ by the procedure of the invention. This translates into an extended battery life and/or a reduced computation time. Equivalently, for constant correlation resources, the correlation time can be increased by a factor of 1.5, this being equivalent to an A-GPS sensitivity gain of 1010log(1.5)≈2 dB.
This is a proof of the fact that only points on the polygon boundary are relevant in the determination of the maximum limit of the search window.
First note that the first term of ΔΦ is independent of the terminal position. It is constant in each case since a time stamp is only determined once for each positioning. The consequence is that only the second and third terms need to be considered in the maximisation and minimisation.
Now assume the contrary to the results, i.e. that the maximum value is attained for an interior point of the polygon. Then, by the topological definition of an interior point, there is a neighbourhood around this point that is also interior to the cell polygon. The maximum value of the phase difference can then be made larger than the assumed maximum, by moving in a suitable direction within the neighbourhood. All directions are possible since movement in an open neighbourhood is considered. First, movement along a circle of constant distance to the site, in the direction that increases the value of the third term of ΔΦ is performed, noting that the second term remains constant on the circle, and a contradiction is obtained.
In case the terminal position would be exactly on the line, projected onto the horizontal plane of the earth tangential coordinate system, between the site and the satellite, the assumed maximum value can instead be increased by moving straight towards the satellite. This follows since the radio signals from the GPS satellites and the serving radio base station site both travel with the same speed c. Furthermore, the elevation angle of the GPS satellite is strictly larger than zero. Hence the difference in travel distance of GPS signals to the interior point on one hand and the boundary of the neighbourhood towards which movement is considered on the other hand, must be smaller than the corresponding travel distance along the surface of the earth that is experienced by the radio signals from the serving radio base station. Hence the experienced code phase advance due to the second term of ΔΦ will be larger than the code phase reduction due to the third term of ΔΦ. The overall effect is a code phase advance and a contradiction is again obtained.
In case the site is between the terminal and the satellite, the maximum value also increases when the terminal moves away from the site along the projected line between the site and the satellite. Both the second and the third terms then contribute to the code phase advance with the same sign. A contradiction is obtained again. Obviously the above argument still hold in case the radio base station site is located outside the cell polygon. It can hence be concluded that the assumption that the maximum code phase is attained in an interior point is false. Hence
is always attained at the boundary of the cell polygon.
This is a proof of the fact that only points on the polygon boundary or at the radio base station site are relevant in the determination of the minimum limit of the search window.
Since the elevation angles of the GPS satellites are strictly greater than zero, it follows that the phase advance introduced by the second term of ΔΦ is greater than any phase retardation caused by the third term, for all tentative terminal positions r1. Hence, in case the serving radio base station site is located in the interior of the cell polygon, the minimum phase difference is attained when the terminal is located in the same coordinates as the serving radio base station site.
In case the serving radio base station site is located outside the cell polygon, then there exist a point on the boundary where ΔΦ attains a minimum value. The boundary is a compact set and ΔΦ is a continuous function. This can as above be proved by assuming the contrary, i.e. that the minimum value of ΔΦ is attained in the interior of the cell polygon. Then by following a circle around the site, ΔΦ can be reduced by moving in one of the two possible directions unless the interior point is located exactly on the projected line segment between the serving radio base station site and the satellite. Since the travel distance differences for GPS signals between points on the surface are less than for radio signals that travel along the surface, it follows that ΔΦ can be reduced by moving towards the radio base station site. A contradiction has thus been obtained and it is clear that minΔΦ is located on the boundary of the cell polygon in case the serving radio base station site is located outside the cell polygon.
0[2] E. D. Kaplan (ed.), Understanding GPS—Principles and Applications. Norwood, Mass.: Artech House, 1996, pp. 119-120.
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/SE04/01054 | 6/29/2004 | WO | 00 | 11/13/2007 |