SYSTEM AND METHOD FOR ENHANCING THE PERFORMANCE OF SATELLITE NAVIGATION RECEIVERS

Abstract

A system and method for enhancing the performance of satellite navigation receivers are disclosed, which incorporate a precise frequency reference in a satellite navigation receiver that reduces the system's dependence on maintaining continuous satellite reception for RAIM availability. As one example, a system for enhancing the performance of a satellite navigation receiver is disclosed, which includes a GPS receiver and a high precision (e.g., atomic) clock incorporated into the GPS receiver. The use of the high precision clock reduces clock error and the number of satellite measurements needed to meet existing RAIM availability requirements. For example, incorporating a precision clock into a GPS receiver provides an enhanced system that meets existing RAIM availability requirements with at least one less satellite measurement than the number needed for prior systems using RAIM.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

The novel features believed characteristic of the invention are set forth in the appended claims. The invention itself, however, as well as a preferred mode of use, further objectives and advantages thereof, will best be understood by reference to the following detailed description of an illustrative embodiment when read in conjunction with the accompanying drawings, wherein:

FIG. 1 depicts a block diagram of an example system for enhancing the performance of GPS receivers, which can be used to implement a preferred embodiment of the present invention;

FIG. 2 depicts the use of three satellite measurements to derive a 2-dimensional solution, in accordance with teachings of the present invention;

FIG. 3 depicts a graphical representation of a probability function for a generalized chi-squared variable for RAIM, in accordance with a preferred embodiment of the present invention; and

FIG. 4 depicts a flow chart of an example method for enhancing the performance of a satellite navigation receiver, which can be used to implement a preferred embodiment of the present invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENT

With reference now to the figures, FIG. 1 depicts a block diagram of an example system 100 for enhancing the performance of satellite navigation receivers, which can be used to implement a preferred embodiment of the present invention. For this example embodiment, system 100 includes a GPS receiver 102, a GPS receive antenna 104, and a precision frequency generator 106 coupled to GPS receiver 102. A processing unit 108 (e.g., microprocessor) associated with GPS receiver 102 executes, among other applications, a RAIM algorithm 110. For this embodiment, precision frequency generator 106 is implemented with a Chip-Scale Atomic Clock (CSAC) integrated into GPS receiver 102. Thus, by using a precision internal clock (106) in GPS receiver 102, processing unit 108 can execute a precision clock-assisted RAIM algorithm (110) to expand RAIM availability, instead of executing a conventional RAIM algorithm using a conventional (fairly inaccurate) internal frequency source/clock. Notably, it should be understood that although system 100 is implemented using a GPS receiver 102 for this example embodiment, the present invention is not intended to be so limited. For example, another embodiment could be implemented using a navigation receiver associated with a different satellite navigation system, such as GLONASS and the like (e.g., augmented with some form of integrity monitoring algorithm, such as RAIM).

In operation, antenna 104 receives signals from a plurality of GPS satellites (not shown), which are detected and processed by receiver 102 to form suitable measurement signals (e.g., pseudorange measurements). At this point, it is useful to consider the following standard GPS position update equations:

$\begin{array}{cc}\overrightarrow{\mathrm{\Delta \rho }}={\stackrel{\rightharpoonup }{\rho }}_m-\stackrel{\bigwedge }{\stackrel{\rightharpoonup }{\rho }}=H\overrightarrow{\Delta x}+\eta & \left(1\right)\end{array}$

where {right arrow over (ρ)}_m is the vector of range measurements to the satellites (from GPS receiver 102), {circumflex over (ρ)}_i=∥{right arrow over (SV)}_i−{right arrow over (x)}_k+1∥ is the predicted range measurement to a satellite i based on the current estimated position ({right arrow over (x)}) and the satellite's position

$\left(\overrightarrow{{\mathrm{SV}}_i}\right),\stackrel{\bigwedge }{\stackrel{\rightharpoonup }{\rho }}$

is the vector of the predicted range measurement {circumflex over (ρ)}_i, H is a matrix composed of the line-of-sight vectors from the current estimated position, ({right arrow over (x)}), to the satellite and augmented with a “1” in the right-most element, {right arrow over (Δx)} is an adjustment to the current estimated position, ({right arrow over (x)}), corresponding to the range measurements, and η is the observed measurement noise.

Next, the least squares solution can be computed as

{right arrow over (Δx)}=(H^TH)⁻¹H^T{right arrow over (Δρ)}=K{right arrow over (Δρ)}  (2)

and the current position of GPS receiver 102 can be updated as

{right arrow over (x)} _k+1 ={right arrow over (x)} _k +{right arrow over (Δx )}  (3)

For this example embodiment, the element Δx in Equation (3) represents the error in the estimate of the state vector x, where x includes three position error components and one clock error component for GPS receiver 102.

Next, a Post Update Measurement Residual (PUMR) is defined. When more than four satellite measurements are available (e.g., the 4-dimensional problem is over-determined), a PUMR represents the residual errors in the least-squared error solution for the available measurements (e.g., four). The least squares nature of this solution indicates that it does not match any one of the measurements exactly, but such a solution minimizes the error from each measurement. For example, the minimization of the error from each satellite measurement can be visualized (e.g., in two dimensions) by referring to FIG. 2, which depicts the use of three satellite measurements to derive a 2-dimensional solution.

As illustrated by the graphical representations depicted in FIG. 2, the “perfect world” solution 202 has a PUMR that is equal to zero. The “real world” solution 204 indicates that the least squared error solution (206) minimizes the sum of the squares of the distances from the solution to each of the measurements involved. As such, a PUMR for a single SV can be computed (e.g., by processing unit 108) as:

Δρ_i=ρ_mi−{circumflex over (ρ)}_i   (4)

where {circumflex over (ρ)}_i is the expected pseudorange measurement to a satellite i given the post-update position/clock error solution. Next, the new position estimate can be updated (e.g., by processing unit 108) based on the current measurement {right arrow over (x)}_k+1. Thus, the predicted pseudorange based on the post-update position/clock solution becomes:

{circumflex over (ρ)}_i=∥{right arrow over (SV)}_i−{right arrow over (x)}_k+1∥  (5)

and the computed PUMR (e.g., residual of the post-update pseudorange measurement associated with the post-update position/clock solution) can be represented as

$\begin{array}{cc}\overrightarrow{\mathrm{\Delta \rho }}=\left[\begin{array}{c}{\mathrm{\Delta \rho }}_1\\ {\mathrm{\Delta \rho }}_2\\ ⋮\\ {\mathrm{\Delta \rho }}_n\end{array}\right]& \left(6\right)\end{array}$

Next, a chi-squared variable can be formed (e.g., by processing unit 108). For this example embodiment, a chi-squared variable is defined as the sum of N independent (e.g., orthogonal) unit-variance random variables. As such, a set of PUMRs can be manipulated and summed to form a chi-squared variable, by normalizing the individual residuals involved. Note that the number of independent variables is equal to a number of residuals greater than the number of dimensions involved with this solution. For example, five measurements provide one independent residual (i.e., one degree of freedom), because (generally) four elements are solved for in the least squares solution (three position components and one clock error component). This is important to note, because in accordance with the principles of the clock-assisted RAIM of the present invention, only three (position) quantities are solved for directly, and the previously calibrated clock information (calibration performed while the conventional RAIM is available) is also used. Thus, the present invention provides an approach to solve for three unknowns (rather than four), which provides one degree of redundancy (providing a failure detection capability) when only four satellites are available. Also, normalization is required so that the variables become unit-variance variables. Thus, given N degrees of freedom, the chi-squared variables have a known probability distribution function. This information is used to compute a threshold above which the chi-squared variable is declared to indicate a failure in one of its components. This threshold is selected to provide a predetermined false alarm rate. The chi-squared variable will only be larger than the threshold value under normal statistical conditions a certain percentage of the time. For that percentage of the time, a failure in one of the measurements can be falsely indicated (e.g., by processing unit 108), while the threshold is being exceeded without actually experiencing a failure.

As noted earlier, it is preferable to normalize the variance of the set of PUMRs, in order to use the PUMRs to compute a true chi-squared variable. When computing a 4-dimensional GPS solution (e.g., three position errors plus one clock error) using five or more measurement values (standard RAIM availability requirement), processing unit 108 can accomplish this normalization by dividing each residual by its expected measurement noise. The expected measurement noise can be assumed to be the same for each satellite, or it can be computed (e.g., by processing unit 108) based on observed signal-to-noise and/or elevation angle information. As such, the normalization expression needed to form the chi-squared variable is represented as

χ²={right arrow over (Δρ)}R_ρ⁻¹{right arrow over (Δρ)}^T   (7)

where R_ρ represents the expected variance of the pseudorange measurements. Whereas R_ρ can take a general form, for the case where five or more measurements are being used to determine the four unknowns (e.g., three position unknowns and one clock unknown) for the GPS problem, R_ρ can be represented in the following diagonal form

$\begin{array}{cc}{R}_p=\left[\begin{array}{ccccc}{\sigma }_1^2& 0& 0& 0& 0\\ 0& {\sigma }_2^2& 0& 0& 0\\ 0& 0& \dots & 0& 0\\ 0& 0& 0& \dots & 0\\ 0& 0& 0& 0& {\sigma }_n^2\end{array}\right]& \left(8\right)\end{array}$

In Equation (8), σ_i is the measurement variance (1-sigma) for the pseudorange measurement from satellite i. Thus, the above-described equation for χ² (Equation (7)) can be rewritten as

χ²=(Δρ₁²/σ₁²+Δρ₂²/σ₂²+ . . . +Δρ_n²/σ_n²)   (9)

Note that if all of the elements σ_i are assumed to be identical, then Equation (9) can be rewritten as

χ²=(Δρ₁²+Δρ₂²+Δρ₃²+ . . . +Δρ_n²)/σ²   (10)

FIG. 3 depicts a probability density function 300 for a generalized chi-squared variable for RAIM, which illustrates certain principles of the present invention. Referring to FIG. 3, the area 302 represents the probability of the normalized chi-squared variable χ² having a value less than T, and the area 304 represents the probability of χ² having a value greater than T. In other words, the area 304 represents the probability of a false detection, P_fd. As such, the value of T can be selected based on the requirement set for P_fd. For example, if a system requirement mandates a threshold level of F_r false alarms per hour, then the number of false alarms allowed per sample can be

F=(F_r/3600)/f_s   (11)

where f_s is the sampling rate of the measurements in Hz. Thus, the probability of a false detection, T, can be determined (e.g., by lookup) from a chi-squared probability table based on the value selected for F. Assume that the value of each Δρ_i=0, except for one value at a time, then Equation (10) can be expressed as

χ₂=(0+0+ . . . +Δρ_i²/σ_i²+0+ . . . +0)=T²   (12)

Then, the corresponding values of Δ{tilde over (ρ)}_i can be computed (one-by-one) as

Δ{tilde over (ρ)}_i=T*σ_i   (13)

For a particular satellite geometry condition, K, the maximum position error that can result from a normally occurring residual is

$\begin{array}{cc}\overrightarrow{\Delta x}=K\overrightarrow{\Delta \tilde{\rho }}& \left(14\right)\\ \mathrm{where}\overrightarrow{\Delta \tilde{\rho }}=\left[\begin{array}{c}0\\ ⋮\\ \Delta {\tilde{\rho }}_i\\ ⋮\\ 0\end{array}\right]& \left(15\right)\end{array}$

Therefore, for any given satellite geometry condition, K, a protection limit can be determined, which is equal to the largest position error ({right arrow over (Δx)}) introduced by an undetected failure in a single satellite, over the set of satellites being used. For this example embodiment, this protection limit is defined to be the “RAIM protection limit”.

As discussed earlier, the conventional RAIM approach used for checking the integrity of GPS navigation solutions is currently used for those situations where five or more satellite measurements are available. However, in accordance with the present invention, RAIM coverage for GPS navigation can be extended to those situations where only four satellites are available, if the GPS receiver (e.g., receiver 102) includes a frequency oscillator/clock (e.g., atomic clock 106), which is stable enough so that its drift error can be accurately modeled. For this example embodiment, a very high precision frequency source/clock (e.g., atomic clock 106) is incorporated into GPS receiver 102 and used instead of a conventional (fairly inaccurate) frequency source/internal clock. In this regard, a method for enhancing the performance of a satellite navigation receiver (e.g., augmented with RAIM) using a very high precision internal clock is shown in FIG. 4.

FIG. 4 depicts an example method 400 for extending RAIM availability for a satellite navigation receiver, which can be used to implement a preferred embodiment of the present invention. Referring to FIGS. 1 and 4 for this example embodiment, processing unit 108 determines if measurements from five or more (e.g., GPS) satellites are available (step 402). If five or more satellite measurements are available, processing unit 108 models the clock phase offset and frequency offset error parameters (step 404), which are respectively referred to as ĉ and

However, if (at step 402) only four satellite measurements are available, processing unit 108 computes only three unknowns (e.g., position errors) instead of the standard four unknowns (step 406). In this case, method 400 still provides RAIM-like integrity, because three unknowns (e.g., three position errors) are being computed and four satellite measurements are being used. However, in this case, instead of processing unit 108 computing a fourth unknown (clock error) during each measurement epoch, processing unit 108 uses the above-described clock error model to estimate the current clock phase offset values (step 408). In this regard, for 3-dimensional positioning, the following expression can be used:

$\begin{array}{cc}\overrightarrow{\mathrm{\Delta \rho }}={\stackrel{\rightharpoonup }{\rho }}_m-\stackrel{\bigwedge }{\stackrel{\rightharpoonup }{\rho }}=H\overrightarrow{\Delta x}+\eta & \left(15\right)\end{array}$

where {right arrow over (ρ)}_m the vector of the range measurements from GPS receiver 102 to the satellites involved, and

$\begin{array}{cc}\widehat{\stackrel{\rightharpoonup }{\rho }}=\overrightarrow{\mathrm{SV}}-\stackrel{\rightharpoonup }{x}+\hat{c}+\widehat{\stackrel{.}{c}}*\Delta t& \left(16\right)\end{array}$

is the predicted range measurement to a satellite, which is based on the current estimated position ({right arrow over (x)}) and the satellite position ({right arrow over (SV)}) adjusted for the estimated clock error using the modeled clock error parameters ĉ and

Next, in order to perform RAIM computations using less than five (e.g., four) measurements and modeled clock errors, method 400 normalizes the PUMR (e.g., least-squares solution) for this situation (step 410). For this case, the normalization matrix can now be expressed as:

χ²={right arrow over (Δρ)}R_ρ⁻¹{right arrow over (Δρ)}^T   (17)

which is similar to Equation (12) above, except the element R in Equation (17) now takes on the form

$\begin{array}{cc}{R}_{\rho }=\left[\begin{array}{cccc}{\sigma }_1^2+{\sigma }_c^2& {\sigma }_c^2& \dots & {\sigma }_c^2\\ {\sigma }_c^2& {\sigma }_2^2+{\sigma }_c^2& \dots & {\sigma }_c^2\\ ⋮& ⋮& ⋮& ⋮\\ {\sigma }_c^2& {\sigma }_c^2& \dots & {\sigma }_n^2+{\sigma }_c^2\end{array}\right]& \left(18\right)\end{array}$

In Equation (18), σ_i again represents the measurement variance (e.g. 1-sigma) for the pseudorange measurement from satellite i. However, since the modeled clock adds an uncertainty to the value of each element a clock uncertainty element ν_c² is provided in Equation (18). Note, for this example embodiment, that processing unit 108 executes a full matrix inversion computation for Equation (18), because in this case the normalization matrix for the PUMR is not a simple diagonal matrix. Also note, for this example embodiment, that although the value of σ_c will increase during the clock error intervals involved, this element is only intended to be modeled and not observed.

In summary, during the time intervals when five or more satellite measurements are available, processing unit 108 observes the clock error values and develops an estimate (model) of the clock phase errors and clock phase rate errors (e.g., can also develop an estimate of the clock phase acceleration errors, if this information is desired). Processing unit 108 uses these modeled parameters to compute predicted range measurements, which are used if the satellite measurement availability is reduced (e.g., only four satellite measurements available for processing instead of five) and RAIM-like (e.g., integrity checking) protection is to be used. Notably, in this case, the modeled clock error values are known to increase over time. Nevertheless, in accordance with the present invention, this growth in the clock error can be suitably modeled by using the specified stability of the high precision frequency source (oscillator) being used. Thus, for this example embodiment, the high precision frequency source used is an atomic clock, which has a typical stated accuracy of 10e⁻¹¹. Therefore, for this example embodiment (at step 412), the clock uncertainty can be modeled (e.g., by processing unit 108) with the expression

σ_c=S_c*Δt   (19)

where the parameter, S_c, represents the specified clock stability, and the parameter, Δt, represents the time elapsed since the last actual clock value was observed. The computed position error data and the modeled clock error data are thus used by processing unit 108 for RAIM-like integrity checking of the navigation solutions obtained (step 414).

It is important to note that while the present invention has been described in the context of a fully functioning system for enhancing the performance of satellite navigation receivers, those of ordinary skill in the art will appreciate that the processes of the present invention are capable of being distributed in the form of a computer readable medium of instructions and a variety of forms and that the present invention applies equally regardless of the particular type of signal bearing media actually used to carry out the distribution. Examples of computer readable media include recordable-type media, such as a floppy disk, a hard disk drive, a RAM, CD-ROMs, DVD-ROMs, and transmission-type media, such as digital and analog communications links, wired or wireless communications links using transmission forms, such as, for example, radio frequency and light wave transmissions. The computer readable media may take the form of coded formats that are decoded for actual use in a particular system for enhancing the performance of satellite navigation receivers.

The description of the present invention has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the invention in the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art. These embodiments were chosen and described in order to best explain the principles of the invention, the practical application, and to enable others of ordinary skill in the art to understand the invention for various embodiments with various modifications as are suited to the particular use contemplated.

Claims

1. A system for enhancing the performance of a satellite navigation receiver, comprising: a satellite navigation receiver, the satellite navigation receiver adapted to receive and process a plurality of measurement signals from a plurality of space-based satellite transmitters;a precision frequency source associated with said satellite navigation receiver; anda processing unit coupled to said satellite navigation receiver, the processing unit adapted to model a first plurality of frequency errors associated with said precision frequency source if a predetermined number of measurement signals are received and processed by the satellite navigation receiver, and estimate a second plurality of frequency errors associated with said precision frequency source if less than the predetermined number of measurement signals is received and processed by the satellite navigation receiver.

2. The system of claim 1, wherein the processing unit is further adapted to: compute a plurality of position error values;estimate a plurality of current clock phase offset values;define a post update measurement residual associated with the computed plurality of position error values and a clock phase error value;normalize the post update measurement residual;model an increase in the values of the modeled first plurality of frequency errors; andoutput the plurality of position error values and modeled increase in the values of the modeled first plurality of frequency errors, if less than the predetermined number of measurement signals is received and processed by the satellite navigation receiver.

3. The system of claim 1, wherein the satellite navigation receiver is a GPS receiver.

4. The system of claim 1, wherein the precision frequency source is an atomic clock.

5. The system of claim 1, wherein the precision frequency source is a Chip-Scale Atomic Clock incorporated in the satellite navigation receiver.

6. The system of claim 1, wherein the satellite navigation receiver comprises a GPS receiver augmented with RAIM.

7. The system of claim 1, wherein the precision frequency source comprises a frequency generator with a precision substantially equal to that of an atomic clock.

8. The system of claim 1, wherein the predetermined number of measurement signals comprises at least five measurement signals.

9. The system of claim 1, wherein the plurality of measurement signals comprises a plurality of pseudorange measurement signals.

10. The system of claim 1, wherein the processing unit is further adapted to: normalize a variance of a set of the post update measurement residuals; andcompute a chi-squared variable for comparison to a RAIM detection threshold value.

11. A system for enhancing the performance of a satellite navigation receiver, comprising: means for receiving and processing a plurality of measurement signals from a plurality of space-based satellite transmitters;means for generating a precision frequency associated with the means for receiving; andmeans, coupled to the means for receiving, for modeling a first plurality of frequency errors associated with said means for generating if a predetermined number of measurement signals are received and processed by the means for receiving, and estimating a second plurality of frequency errors associated with said means for generating if less than the predetermined number of measurement signals is received and processed by the means for receiving.

12. The system of claim 11, further comprising: means for computing a plurality of position error values, estimating a plurality of current clock phase offset values, defining a post update measurement residual associated with the computed plurality of position error values, normalizing the post update measurement residual, modeling an increase in the values of the modeled first plurality of frequency errors, and outputting the plurality of position error values and modeled increase in the values of the modeled first plurality of frequency errors, if less than the predetermined number of measurement signals is received and processed by the means for receiving.

13. The system of claim 11, wherein the means for receiving is a GPS receiver.

14. The system of claim 11, wherein the means for generating comprises an atomic clock.

15. The system of claim 11, wherein the means for generating comprises a Chip-Scale Atomic Clock integrated into the means for receiving.

16. A method for enhancing the performance of a satellite navigation receiver, the method comprising the steps of: receiving and processing a plurality of measurement signals from a plurality of space-based satellite transmitters;generating a precision frequency;modeling a first plurality of frequency errors associated with said precision frequency if a predetermined number of measurement signals are received and processed; andestimating a second plurality of frequency errors associated with said precision frequency if less than the predetermined number of measurement signals is received and processed.

17. The method of claim 16, further comprising the steps of: computing a plurality of position error values;estimating a plurality of current clock phase offset values;defining a post update measurement residual associated with the computed plurality of position error values;normalizing the post update measurement residual;modeling an increase in the values of the modeled plurality of frequency errors; andoutputting the plurality of position error values and modeled increases in values of the modeled first plurality of frequency errors, if less than the predetermined number of measurement signals is received and processed.

18. The method of claim 16, wherein the receiving and processing step is performed by a satellite navigation receiver.

19. The method of claim 16, wherein the receiving and processing step is performed by a GPS receiver.

20. The method of claim 16, wherein the predetermined number of measurement signals comprises five measurement signals.