Methods and systems for mobile navigational applications using global navigation satellite systems

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
24Forward
Citations 
0
Petitions 
1
Assignment
First Claim
1. A method for facilitating vertical and horizontal guidance for mobile users using two or more independent global navigation satellite systems (GNSSs), comprising:
 detecting an occurrence of one or more satellite signal faults in either one of the two independent GNSS constellations when two or more satellites become faulty simultaneously, the detecting occurring at a substantially consistent detection probability in accordance with any specified detection probability requirement; and
deriving vertical protection levels (VPLs) and horizontal protection levels (HPLs) by trading accuracy for integrity;
wherein the detecting and deriving are (i) in position domain and (ii) determinative of safety of the mobile user.
1 Assignment
0 Petitions
Accused Products
Abstract
The present invention relates to providing navigation guidance for mobile users, using Global Navigation Satellite Systems (GNSSs) for applications where the user position needs to be determined in real time while meeting the navigation requirements, especially integrity requirement, for the given application. The present invention assists in providing guidance with a high availability of integrity function by trading accuracy for integrity for two different types of navigation applications. A first type of application requires a capability of detecting an occurrence of multiple satellite signal faults; the first embodiment of the invention provides this capability with a high availability of integrity, using two or more independent GNSSs. A second type of application requires a capability of detecting an occurrence of a single satellite signal fault; the second embodiment of the invention provides this capability with a high availability of integrity, using any one or a combination of GNSSs. The detecting and deriving of both methods are (i) in position domain and (ii) determinative of mobile user safety.
29 Citations
View as Search Results
Method for monitoring the integrity of an aircraft position computed on board  
Patent #
US 8,005,614 B2
Filed 08/20/2007

Current Assignee
Airbus Operations SL

Sponsoring Entity
Airbus Operations SL

MULTICONSTELLATION GLOBAL NAVIGATION SATELLITE SYSTEM AUGMENTATION AND ASSISTANCE  
Patent #
US 20110181465A1
Filed 01/26/2010

Current Assignee
The Boeing Co.

Sponsoring Entity
The Boeing Co.

Method for Increasing the Availability of a Global Navigation System  
Patent #
US 20100141511A1
Filed 02/01/2008

Current Assignee
Astrium GmbH

Sponsoring Entity
Astrium GmbH

Methods and apparatus for geometry extraredundant almost fixed solutions  
Patent #
US 7,768,451 B2
Filed 07/23/2007

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

POSITION DETERMINATION BASED ON CORROBORATED SIGNAL PROCESSING OF PRN CODES  
Patent #
US 20100315285A1
Filed 08/18/2008

Current Assignee
Trimble Navigation Limited

Sponsoring Entity
Trimble Navigation Limited

METHODS AND APPARATUS FOR GEOMETRY EXTRAREDUNDANT ALMOST FIXED SOLUTIONS  
Patent #
US 20090027262A1
Filed 07/23/2007

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

METHOD FOR MONITORING THE INTEGRITY OF AN AIRCRAFT POSITION COMPUTED ON BOARD  
Patent #
US 20080046141A1
Filed 08/20/2007

Current Assignee
Airbus Operations SL

Sponsoring Entity
Airbus Operations SL

METHOD AND APPARATUS FOR TRANSFORMING OVERBOUNDS  
Patent #
US 20070222669A1
Filed 03/07/2007

Current Assignee
Airbus Defence and Space GmbH

Sponsoring Entity
Airbus Defence and Space GmbH

Method for increasing the availability of a global navigation system  
Patent #
US 8,094,069 B2
Filed 02/01/2008

Current Assignee
Astrium GmbH

Sponsoring Entity
Astrium GmbH

Position determination based on corroborated signal processing of PRN codes  
Patent #
US 8,085,191 B2
Filed 08/18/2008

Current Assignee
Trimble Navigation Limited

Sponsoring Entity
Trimble Navigation Limited

METHOD AND SYSTEM FOR DETERMINING PROTECTION LIMITS WITH INTEGRATED EXTRAPOLATION OVER A GIVEN TIME HORIZON  
Patent #
US 20120215376A1
Filed 09/06/2010

Current Assignee
Safran Electronics Defense

Sponsoring Entity
Safran Electronics Defense

METHOD TO HANDLE SINGLE FAILURE GPS FAULTS IN HIGH INTEGRITY RELATIVE POSITIONING SYSTEMS  
Patent #
US 20130050020A1
Filed 08/23/2011

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

Method and system for determining protection limits with integrated extrapolation over a given time horizon  
Patent #
US 8,843,243 B2
Filed 09/06/2010

Current Assignee
Safran Electronics Defense

Sponsoring Entity
Sagem Defense Securite

Method and apparatus for transforming overbounds  
Patent #
US 8,849,569 B2
Filed 03/07/2007

Current Assignee
Airbus Defence and Space GmbH

Sponsoring Entity
Airbus DS GMBH

SYSTEM, METHOD, AND APPARATUS FOR COMPENSATING FOR IONOSPHERIC DELAY IN A MULTI CONSTELLATION SINGLE RADIO FREQUENCY PATH GNSS RECEIVER  
Patent #
US 20150054686A1
Filed 03/17/2014

Current Assignee
Samsung Electronics Co. Ltd.

Sponsoring Entity
Samsung Electronics Co. Ltd.

ARCHITECTURES FOR HIGH INTEGRITY MULTICONSTELLATION SOLUTION SEPARATION  
Patent #
US 20150145724A1
Filed 11/27/2013

Current Assignee
Honeywell International Inc.

Sponsoring Entity
Honeywell International Inc.

Selected aspects of advanced receiver autonomous integrity monitoring application to kalman filter based navigation filter  
Patent #
US 9,547,086 B2
Filed 03/26/2013

Current Assignee
Honeywell International Inc.

Sponsoring Entity
Honeywell International Inc.

Architectures for high integrity multiconstellation solution separation  
Patent #
US 9,784,844 B2
Filed 11/27/2013

Current Assignee
Honeywell International Inc.

Sponsoring Entity
Honeywell International Inc.

System, method, and apparatus for compensating for ionospheric delay in a multi constellation single radio frequency path GNSS receiver  
Patent #
US 9,784,846 B2
Filed 03/17/2014

Current Assignee
Samsung Electronics Co. Ltd.

Sponsoring Entity
Samsung Electronics Co. Ltd.

Method and apparatus for differential global positioning system (DGPS)based real time attitude determination (RTAD)  
Patent #
US 9,829,582 B2
Filed 09/19/2011

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

Selected aspects of advanced receiver autonomous integrity monitoring application to kalman filter based navigation filter  
Patent #
US 10,018,729 B2
Filed 01/13/2017

Current Assignee
Honeywell International Inc.

Sponsoring Entity
Honeywell International Inc.

Sensor installation monitoring  
Patent #
US 10,114,126 B2
Filed 04/30/2015

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

Systems and methods of tracking position and speed in GNSS applications  
Patent #
US 10,151,843 B2
Filed 12/31/2015

Current Assignee
Radio Systems Corporation

Sponsoring Entity
Radio Systems Corporation

Sensor installation monitoring  
Patent #
US 10,551,196 B2
Filed 04/30/2015

Current Assignee
Raytheon Company

Sponsoring Entity
Raytheon Company

Method and apparatus for assisted GPS  
Patent #
US 7,064,706 B2
Filed 06/03/2002

Current Assignee
Google Technology Holdings LLC

Sponsoring Entity
Motorola Inc.

Horizontal/vertical exclusion level determination scheme for RAIM fault detection and exclusion implementation  
Patent #
US 6,847,893 B1
Filed 01/22/2003

Current Assignee
Trimble Navigation Limited

Sponsoring Entity
Trimble Navigation Limited

Systems and methods for fault detection and exclusion in navigational systems  
Patent #
US 6,856,905 B2
Filed 04/29/2003

Current Assignee
Garmin AT Inc.

Sponsoring Entity
Garmin AT Inc.

Method for receiver autonomous integrity monitoring and fault detection and elimination  
Patent #
US 6,864,836 B1
Filed 09/05/2003

Current Assignee
Deere Company

Sponsoring Entity
Navcom Technology Inc.

Integrated navigation system  
Patent #
US 6,377,892 B1
Filed 05/02/2000

Current Assignee
Rockwell Collins Inc.

Sponsoring Entity
Rockwell Collins Inc.

28 Claims
 1. A method for facilitating vertical and horizontal guidance for mobile users using two or more independent global navigation satellite systems (GNSSs), comprising:
detecting an occurrence of one or more satellite signal faults in either one of the two independent GNSS constellations when two or more satellites become faulty simultaneously, the detecting occurring at a substantially consistent detection probability in accordance with any specified detection probability requirement; and deriving vertical protection levels (VPLs) and horizontal protection levels (HPLs) by trading accuracy for integrity; wherein the detecting and deriving are (i) in position domain and (ii) determinative of safety of the mobile user.  View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12)
 13. A method for facilitating mobile user navigation in which guidance is required, using a global navigation satellite system (GNSS), comprising:
detecting an occurrence of a satellite signal fault in a GNSS constellation; and deriving protection levels by trading navigation accuracy for higher navigation integrity when determining a navigation solution; wherein the trading includes combining two solutions with an optimized weight ratio from each pair of subset solutions.  View Dependent Claims (14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28)
1 Specification
1. Field of the Invention
The invention relates to navigational guidance for mobile users, using Global Navigation Satellite System(s) (GNSSs).
2. Background Art
GNSSs include the Global Positioning System (GPS) developed by the United States Air Force, Global Navigation Satellite System (GLONASS) developed in Russia, which is under needed improvements, Galileo under development in Europe, and possibly other systems such as Compass, currently under consideration in China. However, GPS has been the only complete system and has been widely used in aviation for a number of years. Among these other systems, the implementation of Galileo is promising in the near future. When Galileo becomes fully operational, GPS and Galileo may be used together to improve the navigational performance of mobile users anywhere in the world.
As known in the art, GPS, as will Galileo, provides highly accurate position and velocity information and precise time continuously to an unlimited number of users throughout the world. However, GPS alone, without any other augmentation, lacks a capability known as integrity. That is, if a satellite fault occurs, GPS cannot inform the user of the fault in a timely manner. The integrity issue is critical for aviation because lack of this capability can directly lead to a situation in which safety may be compromised. One important method to provide the integrity for GPS in civil aviation applications is known in the art as Receiver Autonomous Integrity Monitoring (RAIM).
A GPS user receiver only requires four satellites for its position determination; however, five or more satellites are typically visible from any location, thus making redundant GPS measurements available. If redundant measurements are available (i.e., if five or more satellites are visible), a consistency check may be performed among the different satellite range measurements to determine if any of the satellites are faulty with an unusually large ranging error. The idea is that if no satellite is faulty (i.e., has an unusually large ranging error), it is likely that the measurements are consistent. If any one of the satellites is faulty, having an unusually large ranging error, the measurements would not be consistent with each other.
RAIM is essentially an intuitive formulation and can be used for fault detection in navigation applications. The degree with which RAIM detects faults depends on the usertosatellite geometry formed by visible satellites. RAIM performance is dependent upon position error not exceeding levels for safe operation. RAIM performance is also a function of Pr{Hazardously Misleading Information (HMI)}, which is a probability that the position error exceeds the required alert limit without a timely warning to the pilot. The fraction of time in which integrity monitoring can be provided for a given alert limit, satisfying the Pr{HMI} requirements, is called the availability of RAIM. For many mobile navigation applications, especially for air navigation, RAIM availability of many 9'"'"'s (i.e., 0.999. . . ) is required. However, this requirement is often difficult to achieve for GNSS navigation applications that would provide significant operational benefits. For this reason, effective RAIM methods that provide high availability are always sought after.
RAIM typically refers to a fault detection (FD) function, which determines the presence or absence of a fault that could cause HMI. Sometimes, however, RAIM is used to refer to the fault detection and exclusion (FDE) function as well. Upon detection of a fault by the FD function, the FDE function is activated and attempts to identify and exclude the source (e.g., a satellite) of the HMI. For this reason, while RAIM FD is related to integrity, the FDE function is related to continuity of navigation. An FDE technique is derived by extending a RAIM detection method, as will be briefly described later. However, in general, the usertosatellite geometry must satisfy much more stringent criteria for the FDE function to be available than for the RAIM detection function.
What is highly desirable, therefore, are RAIM methods that provide improved availability for their respective applications over the existing RAIM methods. The present invention achieves this goal by trading accuracy for integrity.
The present invention relates to providing navigation guidance for mobile users, using GNSSs. The invention can apply to applications where the user position needs to be determined while meeting the navigation requirements for the given application, especially accuracy and integrity. A mobile user can be any system that moves, such as aircraft, ships, land vehicles and even space vehicles. While the invention can be used in a wide variety of applications, it is described in the present application within the context of air navigation.
The present invention uses GNSS(s) to assist in providing guidance with a sufficiently high level of integrity for two different types of navigation applications. The first type is one in which the probability of multiple satellite faults simultaneously occurring cannot be neglected. This type of application includes air navigation, more specifically, vertically guided aircraft approaches such as localizer performance with vertical guidance (LPV) and approach with precision vertical guidance (APV)II. It also includes the future precision approaches equivalent to that of the guidanceprovided by the Instrument Landing System (ILS).
The second type of application is one in which the probability of multiple satellite faults simultaneously occurring can be neglected. This type of application includes air navigation in nonprecision approach (NPA), terminal, and en route phases of flight.
Consistent with the principles of the present invention, a first embodiment of the present invention includes detecting an occurrence of one or more satellite signal faults in either one of the two independent GNSSs (or two groups of independent GNSSs if more than two GNSSs are used) and deriving protection levels (PLs), trading accuracy for integrity. The detecting and deriving are (i) in the position domain and (ii) determinative of mobile user safety.
The first embodiment of the present invention can be used for the first type of application discussed above and will be referred to herein as the Optimally Weighted Average Solution (OWAS)2. OWAS2 provides integrity, using two or more independent GNSSs, for mobile user navigational applications in which the probability of multiple satellites simultaneously occurring within a constellation cannot be neglected. One example of such an application is vertically guided approaches, in which both vertical and horizontal navigation guidance must be provided with tight alert limits.
These alert limits are tight enough such that the probability of faults simultaneously occurring in two or more satellites to cause HMI with the position errors exceeding the alert limits cannot be neglected. Therefore, the OWAS2 technique requires two independent GNSSs (or two groups of independent GNSSs) and assumes that multiple satellite faults may occur at any given time within a GNSS constellation. On the other hand, it is assumed that the probability of multiple satellite faults simultaneously occurring across any two independent constellations can be neglected as will be shown later. In the description of the first embodiment of the present invention, GPS and Galileo are to be used as two independent GNSSs as an example.
A second embodiment of the present invention includes a method for facilitating mobile navigation using one or more GNSSs. This embodiment is described using an example of air navigation in NPA, terminal, and en route phases of flight, in which only horizontal guidance is required. The method includes detecting an occurrence of a satellite signal fault in a GNSS constellation and deriving crosstrack (XPL) and along track (APL) protection levels by trading accuracy for integrity. The detecting and deriving are (i) in position domain and (ii) determinative of mobile user safety.
The second embodiment of the present invention is intended to be used for the second type of application discussed above and will be referred to herein as OWAS1. OWAS1 provides integrity for mobile navigation for applications with alert limits that are large enough such that the probability of faults simultaneously occurring in two or more satellites to cause HMI with the position errors exceeding the alert limits is negligible. Therefore, the OWAS1 technique assumes that only up to a single satellite fault may occur at any given time. Additionally, the OWAS1 technique does not require more than a single GNSS. If more than one GNSS is available, they are combined and treated as a single constellation, in which occurrence of every satellite fault is independent from each other.
In short, the present invention improves the availability of the integrity function in navigation guidance for either type of the applications described above. The present application uses vertically guided approaches as an example of the first type and en route, terminal, and NPA phases of flight as an example of the second type. The invention, however, is not limited to these examples.
Further features and advantages of the present invention are provided as well as the structure and operation of various embodiments of the present invention, as described in detail below with reference to the accompanying drawings.
The accompanying drawings, which are incorporated herein and constitute part of the specification, illustrate embodiments of the present invention and, together with the general description given above and the detailed description of the embodiments given below, serve to explain the principles of the invention. In the drawings:
The following detailed description of the present invention refers to the accompanying drawings that illustrate exemplary embodiments consistent with this invention. Other embodiments are possible, and modifications may be made to the embodiments within the spirit and scope of the invention. Therefore, the following detailed description is not meant to limit the invention. Rather, the scope of the invention is defined by the appending claims.
It will be apparent to one skilled in the art that the present invention, as described below, may be implemented in many different embodiments. Any actual software code implementing the present invention is not limiting of the present invention. Thus, the operational behavior of the present invention will be described with the understanding that modifications and variations of the embodiments are possible, given the level of detail presented herein.
In the present application, RAIM is used to refer to FD, and FDE will be differentiated as such. The implementation of both RAIN FD and FDE is by means of software inside of the navigation processor that is part of the GNSS user receiver.
The first embodiment of the present invention enhances the capability to use RAIM for both horizontal and vertical guidance during instrument approach, for example. For example, the OWAS2 RAIM method of the present invention enables both GPS and Galileo to provide an extended capability to detect multiple satellite signal faults in either constellation. More specifically, the present invention brings together the best features of previous versions of RAIM methods and implements these methods in the position domain.
In the example shown assuming an air navigation application, a database control allows the user to enter user input and the navigation processor output may be output to a cockpit display 112 or to an autopilot 114
Integrity Requirement
Alert Limits. The horizontal and vertical alert limits (HAL and VAL) define maximum navigation sensor errors (NSE) allowed for safe navigation for a given type of flight. These limits apply to the total NSE, including all errors resulting from signalinspace, airborne receiver, ionosphere delay, and troposphere delay. Violation of these limits without warning to the user is considered to be HMI from the navigation system to the user.
For applications of the first embodiment of the present invention (OWAS2), user equipment desirably determines two sets of parameters. One set of parameters is for vertical and horizontal protection levels (VPL and HPL), which are highprobability upper bounds on the vertical and horizontal NSE, respectively according to the integrity requirement for a given operation. VPL and HPL are compared to VAL and HAL specified for a given operation (e.g., a vertically guided approach). If either VPL>VAL or HPL>HAL, the user is alerted that RAIM is not available. If RAIM is not available, then the user has to take an appropriate measure according to a given operational procedure. The other set of parameters is for the test statistics to detect a fault. If either of the test statistics (one for vertical and the other for horizontal) exceeds its fault detection threshold, the FDE function may be activated to attempt to exclude the bad satellite(s). However, isolation of and exclusion of multiple bad satellites is quite unreliable for tight alert limits in general, such as those for vertically guided approaches.
If the fault(s) cannot be excluded and/or navigation cannot continue with the remaining satellites, the user is alerted that a fault has been detected. When a fault is detected, then the user is required to take an appropriate action according to a given operational procedure, which could be either executing a missed approach or, if possible, reverting to an instrument approach with greater alert limits to the same runway.
For applications of the second embodiment of the present invention (OWAS1), user equipment desirably determines the same sets of parameters as with OWAS2, but for the example of air navigation in NPA, terminal, and en route phases of flight, the parameters are determined only with respect to the horizontal position. The pilot has to take an appropriate action according to a given operational procedure where the pilot is alerted with either a “no RAIM” or a “fault detection” message. As stated earlier, for applications in which OWAS1 is used, only up to a single satellite fault is assumed at any given time. With this assumption and with comparatively large alert limits for en route through NPA phases of flight, availability of the FDE function associated with OWAS1 is typically high such that it would be worthwhile to execute the FDE function.
The first embodiment of the present invention can be used for the vertically guided approaches. The performance of this invention can be measured by the availability achieved for a particular value of VAL. Desirable ranges of interest are 20 m≦VAL≦35 m. The minimum value of VAL is specified for APVII. The maximum VAL of 35 m is for LPV presently being considered for the Wide Area Augmentation System (WAAS).
Integrity Risk. The International Civil Aviation Organization (ICAO) Annex 10 states that the Category (CAT) I integrity risk, defined as the probability of HMI resulting from undetected signalinspace errors, is Pr{HMI} less than or equal to 2×10^{−7 }per approach. For a future system, the target for Pr{HMI} is assumed to be less than or equal to 1×10^{−7 }per approach.
The fault scenarios considered in
In
The fault case 204 includes a single fault or a commonmode fault affecting multiple faults in only GPS or only Galileo, but not in both. The capability of detecting simultaneous errors resulting from a commonmode fault has been considered unnecessary until now because RAIM has been used thus far only for the en route, terminal, and NPA phases of flight, for which the alert limits are relatively large. The probability of having large errors on multiple satellites simultaneously, such that it would cause an HMI event against the relatively large alert limits, is considered negligible.
However, for applications with tight alert limits such as those for vertically guided approaches, even when the ranging errors for individual satellites are not unusually large, an HMI event may result when multiple satellites are involved. (For example, if all the range measurements are consistent with an erroneous earth orientation, then the resulting errors exceeding the tight alert limits for vertically guided approaches are undetectable by the conventional RAIM techniques.) Therefore, the probability of a commonmode fault within one constellation that may affect multiple satellites 204 cannot be neglected. For this reason, consideration of the possibility of HMI results involving multiple satellites simultaneously is required for such applications as the vertically guided approaches. While conventional RAIM techniques cannot be assured to detect simultaneous errors, the improved RAIM technique of the first embodiment of the present invention can accommodate multiple faults by working with two independent constellations.
The fault case 206 includes simultaneous faults in GPS and Galileo. The faults 206 that occur in both GPS and Galileo must be independent events, and therefore, one can show as follows that the probability of such faults occurring simultaneously is sufficiently small to meet the allocated integrity risk requirement. In both constellations, it is assumed that the probability of a fault in each constellation is 10^{−4}. The probability of simultaneous faults in both GPS and Galileo is taken as (10^{−4})^{2}=10^{−8 }since the constellations (e.g, GPS and Galileo) are operated independently. This probability meets the allocated integrity risk of 10^{−8}/approach even if it is conservatively assumed that the conditional probability of not detecting simultaneous faults in both constellations can be equal to one.
Availability
In estimating the performance of RAIM, availability is defined as the average fraction of time that VPL≦VAL and HPL≦HAL subject to meeting the accuracy requirement as well. ICAO Annex 10 recommends availability to be in the range of 0.99 to 0.99999. A particular value of availability depends upon operational need.
Enhanced RAIM
An enhanced version of RAIM is provided in a first embodiment of the present invention. The enhanced RAIM will be referred herein to as the Optimally Weighted Average Solution (OWAS), more specifically OWAS2, to distinguish this first embodiment of the present invention from the second embodiment of the invention called OWAS1. OWAS2 was primarily developed for application to a combined GPS/Galileo position solution that would be used for vertically guided approaches; however, the technique is not limited to these applications. OWAS2 is based on the salient features of two recently developed RAIM methods known by those of skill in the art as the Group Separation (GS) method and NIORAIM.
By way of background, the GS method was devised to provide an improved RAIM when GPS and Galileo would be used in a combined position solution for vertically guided approaches. The advantage of the GS method over the conventional RAIM methods is the ability to handle multiple faults within either the individual GPS or Galileo constellations. Therefore, the GS method can be used for vertically guided approaches that require a capability to detect simultaneous multiple faults.
When determining a protection level, error boundaries a_{1 }and a_{2 }are added to protect against any masking of fault detection by the faultfree errors (e.g., receiver noise and multipath). That is, a_{1 }and a_{2 }bound the respective position solution errors with a probability that, in combination with an a priori rate of occurrence of a fault, would meet the integrity requirement. The probability is determined as follows. For example, if the integrity requirement is 8×10^{−8 }(see
The GS technique 300 works regardless of how many satellites are faulty in either the GPS or Galileo at any given time. While the GS technique 300, unlike other RAIM methods, provides a measure of multiple fault capability, which is needed, for example, for vertically guided approaches, it does not deliver sufficiently high availability.
NIORAIM was devised by modifying one of the conventional RAIM methods called a leastsquares residual (LSR) method to improve RAIM availability. The LSR method determines a protection level (e.g., HPL) on the basis of the maximum slope in a plot 400 of position error 402 versus a test statistic 404, as illustrated in
In the plot 400 of
OWAS2 Technique
With this formulation, the next step is to find an optimal weight ratio that satisfies the criteria for RAIM to be available as much as possible while taking into account the accuracy requirements for both vertical and horizontal positions. The accuracy is expressed by the standard deviation (σ_{V}) for the vertical position and the rootmeansquare error (d_{rms}) for the horizontal position. In our example, the horizontal position error is assumed to have a twodimensional Gaussian distribution, and as a conservative estimate of d_{rms}, the standard deviation of the horizontal position error along the major axis of the assumed Gaussian bivariate error ellipse (d_{major}) is used.
The criteria for σ_{V }and d_{major }are derived from the operational requirements:
(a) 95^{th }percentile NSE bound for vertical and horizontal position error
(b) NSE bound to meet integrity in a faultfree case against vertical and horizontal alert limits, respectively.
From these two requirements, it follows that the maximum allowable values of σ_{V }and d_{major}, (σ_{Vmax}, d_{majormax}) are determined.
By adding the integrity availability requirements for protection levels, OWAS2 RAIM is available if all of the following four conditions are satisfied:
σ_{V}≦σ_{Vmax}, d_{major}≦d_{majormax }(accuracy) and VPL≦VAL, HPL≦HAL (integrity availability).
The OWAS2 process is described in greater detail below. Because the processes between the vertical and horizontal positions are quite similar, they are described primarily for the vertical position. Any difference for the horizontal position will be noted. As a preliminary matter, OWAS2 is a technique that can be used to determine an optimal weight ratio that would satisfy all of the above criteria to the extent possible.
An initial step in the OWAS2 process is determining a weight ratio (r) that minimizes VPL. Just as in the GS method above,
VPL=max {VPL_{1}, VPL_{2}} (3)
VPL_{1 }and VPL_{2 }can be expressed as linear functions of the weight ratio r as illustrated in
The tradeoff between accuracy and integrity is illustrated in a graph 600 in
Because a smaller VPL and thus a higher availability of the integrity function are obtained at the expense of a degraded accuracy of the navigation solution, it is desirable to ensure that the accuracy requirement is also met when weighting a position solution to decrease a protection level. It is shown later that both VPL and variance, σ_{v}^{2}, representing the accuracy of the vertical position error (in the absence of a fault condition) are functions of the weigh ratio.
In general, the weight ratios that minimize σ_{v }or VPL are not the same as illustrated in
(i) r_{bestint }minimizes VPL to the extent possible while still meeting the accuracy requirement σ_{v}≦σ_{vmax }
(ii) r_{bestacc }minimizes σ_{v }to the extent possible while still meeting the integrity availability requirement VPL≦VAL.
Detailed expressions for OWAS2 are provided below. First, expressions for covariance matrices used in the subsequent equations are derived. Second, equations for VPL are derived as a function of their weight ratios. Third, equations for σ_{V }are derived as functions of their weight ratios. Finally, it is shown how to determine an optimal weight ratio that satisfies both the integrity availability and accuracy requirements for RAIM according to the selection between minimizing VPL and minimizing σ_{V}.
Derivation of Covariance Matrices for OWAS2
As described earlier in conjunction with
Note that from the definition of r in (6), the following must hold:
0≦r≦1 (7)
The position error associated with the weighted average solution is written as
Δx_{A}=r Δx_{1}+(1−r)Δx_{2} (8)
The general expression for the positionerror for a weighted leastsquares solution is given by
Δx=(G^{T}WG)^{−1}G^{T}WΔR (9a)
ΔR: Ranging error
 W: Diagonal matrix whose i^{th }element is the inverse of the variance of the i^{th }ranging error
 G: Matrix that connects ranging errors into position and receiver clock errors
Therefore,
Δx_{1}=(G_{1}^{T}W_{1}G_{1})^{−1}G_{1}^{T}W_{1}ΔR_{1} (9b)
Δx_{2}=(G_{2}^{T}W_{2}G_{2})^{−1}G_{2}^{T}W_{2}ΔR_{2} (9c)
The covariance matrix of the error in the measured separation of the GPS solution from the weighted average solution is defined by
Equation (10) is evaluated termbyterm as follows.
P_{1}≡E[Δx_{1}Δx_{1}^{T}]=(G_{1}^{T}W_{1}G_{1})^{−1} (11)
E[Δx_{A}Δx_{1}^{T}]=rP_{1}+(1−r)E[Δx_{2}Δx_{1}^{T}] (12)
Because GPS and Galileo are independent systems, one may assume that the GPS and Galileo ranging errors are uncorrelated, i.e.,
E[ΔR_{2}ΔR_{1}^{T}]=0 (13)
Then,
E[Δx_{2}Δx_{1}^{T}]=(G_{2}^{T}W_{2}G_{2})^{−1}G_{2}^{T}W_{2}E[ΔR_{2}ΔR_{1}^{T}]W_{1}G_{1}(G_{1}^{T}W_{1}G_{1})^{−1}=0 (14)
Therefore, (12) is simplified to
E[Δx_{A}Δx_{1}^{T}]=rP_{1} (15a)
Also,
E[Δx_{1}Δx_{A}^{T}]=rP_{1} (15b)
Similarly,
P_{A}=E[Δx_{A}Δx_{A}^{T}]=r^{2}P_{1}+(1−r)^{2}P_{2} (16)
Inserting (11), (15) and (16) into (10) gives
dP_{1}=(1−r)^{2}(P_{1}+P_{2})=(1−r)^{2}P_{S} (17)
where P_{S}≡P_{1}+P_{2} (18)
Similarly, one can derive
dP_{2}≡E[(Δx_{2}−Δx_{A})(Δx_{2}−Δx_{A})^{T}]=r^{2}P_{S} (19)
Derivation of VPL and HPL as a Function of the Weight Ratio for OWAS2:
First, the expression for VPL is derived.
Define
where
Q^{−1}: Inverse of the complement of the onesided standard normal cumulative distribution function (CDF) (tail probability)
P_{ffd}: Maximum allowable false alert probability
P_{md}: Maximum allowable missed detection probability
Note: P_{ffd }is divided by 4 because the total allowable probability is evenly allocated between vertical and horizontal, each for a twosided faultfree error distribution
Further define,
σ_{V,1}≡√{square root over (P_{1}(3,3))} (21a)
σ_{V,2}≡√{square root over (P_{2}(3,3))} (21b)
σ_{V,S}≡√{square root over (P_{S}(3,3))}=√{square root over (P_{1}(3,3)+P_{2}(3,3))}{square root over (P_{1}(3,3)+P_{2}(3,3))} (21c)
Then based on (17) and (19),
σ_{dV,1}≡√{square root over (dP_{1}(3,3))}=(1−r)√{square root over (P_{S}(3,3))}=(1−r)σ_{V,S} (21d)
σ_{dV,2}≡√{square root over (dP_{2}(3,3))}=r√{square root over (P_{S}(3,3))}=rσ_{V,S} (21e)
With these definitions, and noting that the VPL_{1 }and VPL_{2 }are the sums of the respective thresholds and misseddetection buffers:
VPL_{1}=T_{1}+a_{1}=K_{ffd}σ_{dV,1}+K_{md}σ_{V,1}=(K_{ffd}σ_{V,S})(1−r)+K_{md}σ_{V,1} (22a)
VPL_{2}=T_{2}+a_{2}=K_{ffd}σ_{dV,2}+K_{md}σ_{V,2}=(K_{ffd}σ_{V,S})r+K_{md}σ_{V,2} (22b)
VPL=max {VPL_{1}, VPL_{2}} (23)
To obtain the minimum VPL, set
VPL_{1}=VPL_{2} (24)
From (22a), (22b), and (24),
Substituting (25) into (22) gives
The weight ratio in (25) gives the smallest VPL. A weight ratio that minimizes HPL can also be derived in a similar manner. In order to satisfy the integrity requirement, recall that VPL and HPL so derived must satisfy both of the following conditions:
VPL≦VAL (27a)
HPL≦HAL (27b)
Derivation of σ_{V }and d_{major }as a Function of the Weight Ratio for OWAS2:
The covariance matrix for the weighted average position error, P_{A}, was given earlier by (16):
P_{A}=E[Δx_{A}Δx_{A}^{T}]=r^{2}P_{1}+(1−r)^{2}P_{2} (16)
σ_{V }and d_{major }are given, respectively, by
σ_{V}=√{square root over (P_{A}(3,3))} (28a
and
In order for OWAS2 RAIM to be available, σ_{V }and d_{major }should satisfy the following conditions:
σ_{v}≦σ_{vmax} (29a)
d_{major}≦d_{majormax} (29b)
σ_{Vmax }and d_{majormax }are derived so as to meet the accuracy and faultfree integrity requirements. Note that in the calculations of P_{A}, σ_{V }and d_{major}, optimum weight ratios are different in general for vertical and horizontal.
Determination of an Optimal Weight Ratio
As was discussed earlier, there are two ways of selecting a weight ratio: one minimizing VPL, the other maximizing accuracy. If the former choice is selected, the process to determine an optimal weight ratio, takes the following steps. Step 1: If the four inequalities given in (27) and (29) are satisfied, then r_{minVPL }is selected as the optimal ratio. OWAS2 RAIM is declared available. Step 2: If either or both of the two integrity availability inequalities in (27) are not satisfied, then OWAS2 RAIM is declared unavailable. Step 3: If both integrity availability inequalities in (27) are satisfied, but either or both of the accuracy inequalities in (29) are not satisfied, then one may search for a new weight ratio to satisfy that inequality. A more precise process to determine the new optimal ratio is described below. Note that the optimal ratio needs to be changed only for the vertical or horizontal accuracy requirement that is not attained. Note that if the accuracy requirement is satisfied for only one but not both of vertical and horizontal positions, a new optimal ratio needs to be determined only for the position not yet satisfied.
First, an inequality that a weight ratio must satisfy to meet the accuracy requirement for vertical position is derived.
σ_{V}≦σ_{Vmax}→P_{A}(3,3)≦(σ_{Vmax)}^{2}→r^{2}P_{1}(3,3)+(1−r)^{2}P_{2}(3,3)−(σ_{Vmax})^{2}≦0 (30)
Replacing P_{1}(3,3) and P_{2}(3,3) with the notations defined earlier, we obtain
(σ_{V,1}^{2}+σ_{V,2}^{2})r^{2}−2σ_{V,2}^{2}r−[(σ_{Vmax})^{2}−σ_{V,2}^{2}]≦0 (31)
Therefore,
r_{1}≦r≦r_{2} (32a)
where
In addition, as presented earlier in Eq. (7), the weight ratio must also satisfy
0≦r≦1 (7)
Therefore, the accuracy requirement can be met only for r that satisfies the inequality
minr_{Vacc}≦r≦maxr_{Vacc} (33)
where
minr_{Vacc}=max(r_{1},0)
maxr_{Vacc}=min(r_{2},1) (34)
For the option of choosing a weight ratio that would minimize VPL to the extent possible, the weight ratio for vertical is determined, as follows:
If minr_{Vacc}≦r_{minVPL}≦maxr_{Vacc}, then
Opt_r_{V}=r_{minVPL} (35a)
If r_{minVPL}<minr_{Vacc}, then
Opt_r_{V}=minr_{Vacc} (35b)
If r_{minVPL}>maxr_{Vacc}, then
Opt_r_{V}=maxr_{Vacc} (35c)
If the inequality in (35a) holds, VPL does not need to be redetermined. It is given in (26). However, if the inequality in (35a) does not hold, then VPL needs to be redetermined using Opt_r_{V }from either (35b) or (35c), whichever is applicable. Similar steps are taken to revise the optimal weight ratio for the horizontal position solution (Opt_r_{H}).
Then, if both conditions (VPL≦VAL) and (HPL≦HAL) are met for VPL and HPL determined above using Opt_r_{V }and Opt_r_{H}, respectively, RAIM is declared available. Otherwise, it is declared unavailable. An optimal weight ratio that maximizes accuracy while still satisfying VPL and HPL criteria can be derived in a similar manner.
OWAS1 Technique
OWAS1 is an extension of the OWAS2 technique, discussed above, but used for different applications. First, OWAS1 does not require two independent constellations as OWAS2 does. That is, OWAS1 works with a single (or multiple constellations combined into one) constellation. Secondly, OWAS1 provides integrity under the assumption of up to a single satellite fault. For this reason, OWAS1 is intended to be used, just like all conventional RAIM methods and NIORAIM, for en route, terminal, and NPA phases of flight, for which it may be assumed for the reason discussed earlier that only up to a single satellite fault may occur at any given time. OWAS1 is similar to NIORAIM in that both OWAS1 and NIORAIM achieve higher RAIM availability at the expense of less than optimal position accuracy, that is, by trading accuracy for integrity (or equivalently, availability of integrity). However, OWAS1 differs from NIORAIM in that while NIORAIM derives weights numerically and applies them to the range measurements, OWAS1 derives weights analytically and applies them to the position estimates. Like all conventional methods, both OWAS1 and NIORAIM can provide both the RAIM (detection) function and the fault detection and exclusion (FDE) function for en route through NPA phases of flight.
While OWAS2 deals only with one pair of independent subset solutions, OWAS1 deals with N (≧5) correlated subset solutions. However, in order to avoid becoming unwieldy, OWAS1 considers a horizontal position error along different axes of interest separately. Specifically two orthogonal axes are chosen for OWAS1 to consider the horizontal position error components: cross track (XTK) and alongtrack (ATK) as illustrated in
The first step of OWAS1 is to identify which pair of subset position solutions is the pair that constrains the protection level (called the “PLconstraining pair”) along the selected axis (i.e., XTK or ATK).
As shown in
With the above formulation, the optimal weights are derived the same way as for OWAS2 except for some modifications necessary to take into account the following two factors: (i) x_{1 }and x_{2 }are correlated, and (ii) the horizontal position is considered along the two orthogonal axes, and thus both the protection level and accuracy are likewise considered along those axes: HPL is replaced by the XTK protection level (XPL) and the ATK protection level (APL), and d_{major }is replaced by σ_{X }and σ_{A }representing the standard deviations of the XTK and ATK position errors, respectively. That is, in order for OWAS1 RAIM to be available for the horizontal position, the following conditions must be satisfied:
XPL≦HAL and APL≦HAL (integrity) (36a, b)
σ_{X}≦σ_{Xmax }and σ_{A}≦σ_{Amax }(accuracy) (37a, b)
Note that the vertical position is of no concern for en route through NPA phases of flight, for which OWAS1 is intended. Also note that the alert limits for position errors along the XTK and ATK are assumed to be identical. The detailed derivations of equations for OWAS1 are presented below.
Determination of Optimum Weight Ratio for OWAS1
Tradeoff Between Accuracy and Integrity Availability
As in the case with OWAS2, it is desirable with OWAS1 to ensure that the accuracy requirement is met along with the integrity requirement when weights are determined. Two options are possible for OWAS1 in determining the weight (i.e., weight ratio): (1) minimizing HPL (i.e., XPL and APL) while still meeting the accuracy requirement (max(σ_{X}, σ_{A})≦σ_{Hmax}); or (2) minimizing max(σ_{X}, σ_{A}) while still meeting the integrity availability criterion (XPL≦HAL and APL≦HAL). This tradeoff is possible for any pair of subset solutions determined to be the PLconstraining pair.
Evaluation of Availability of Fault Detection and Exclusion (FDE) Function
Just like the conventional RAIM methods, the OWAS1 Fault Detection and Exclusion (FDE) function is defined as available if the OWAS1 RAIM detection function is available for every subset of (N−1) satellites where N is the number of satellites being tracked. Therefore, availability of OWAS1 FDE is evaluated in the following steps:
Step 1: Take one satellite at a time out of N satellites being tracked and create a subset of (N−1) satellites.
Step 2: For the given set obtained in Step 1, determine the PLconstraining pairs along the XTK and ATK axes for a given direction of flight and calculate the optimal weight and the protection level for the selected PLconstraining pair. The protection level and OWAS1 RAIM detection availability are calculated as described below.
Step 3: Declare OWAS1 FDE is available if and only if the RAIM detection function is available for every subset of (N−1) satellites.
Selection of Axes for OWAS1
As described earlier, the first step of OWAS1 is to determine the PLconstraining subset solution pairs along the XTK and the ATK axes. How to determine the PLconstraining pair was described above with reference to
Derivation of Covariance Matrices for OWAS1
The subset solutions in the PLconstraining pair are labeled as x_{1 }and x_{2}. With this notation, many equations for OWAS1 will be in the same form as those presented earlier for OWAS2. The difference comes from the fact that, while the two subset solutions in OWAS2 are independent, they are correlated in OWAS1. The procedure to determine an optimal weight that minimizes a protection level and thus maximizes the RAIM availability for the selected pair is described below.
The position solution is a weighted average of the two PLconstraining subset solutions, x_{1 }and x_{2}:
From the definition of r in (38), the following must hold:
0≦r≦1 (40)
The position error associated with the weighted average solution is written as
Δx_{A}=rΔx_{1}+(1−r)Δx_{2} (41)
The general expression for the position error for a weighted leastsquares solution is given by
Δx=(G^{T}WG)^{−1}G^{T}WΔR (42a)
where ΔR is a ranging error vector and W is a diagonal matrix whose i^{th }element is the inverse of the variance of the i^{th }ranging error.
Assume that x_{1 }and x_{2 }are the subset solutions based on all but the i^{th }and j^{th }satellite range measurements, respectively.
Then,
Δx_{1}=(G_{1}^{T}W_{1}G_{1})^{−1}G_{1}^{T}W_{1}ΔR_{1} (42b)
Δx_{2}=(G_{2}^{T}W_{2}G_{2})^{−1}G_{2}^{T}W_{2}ΔR_{2} (42c)
where G_{1 }and ΔR_{1 }and G_{2 }and ΔR_{2 }are G and ΔR with their i^{th }and j^{th }rows removed, respectively. G_{1 }and G_{2 }are (m−1)×4 matrices, and ΔR_{1 }and ΔR_{2 }are (m−1)×1 matrices.
Note that Eqs. (38) through (42c) above for OWAS1 are expressed in identical forms as Eqs. (6) through (9c) for OWAS2. The difference is that while Δx_{1 }and Δx_{2 }for OWAS2 are independent, those for OWAS1 are correlated.
Δx_{1 }and Δx_{2 }can be reexpressed in terms of G and ΔR (instead of G_{1 }and ΔR_{1 }or G_{2 }and ΔR_{2}) as follows by defining matrix M_{i }and M_{j }to be an (m×m) identity matrix with the i^{th }and j^{th }diagonal elements zeroed out, respectively. In this case, Δx_{1 }can be expressed as follows:
Likewise,
Δx_{2}=(G^{T}M_{j}WG)^{−1}G^{T}M_{j}WΔR (43b)
The covariance matrix of the error in the measured separation of the solution x_{1 }from the weighted average solution is given by
Eq. (44) is evaluated term by term, as follows.
Define,
P_{11}≡E[Δx_{1}Δx_{1}^{T}] (45a)
P_{12}≡E[Δx_{1}Δx_{2}^{T}] (45b)
P_{21}≡E[Δx_{2}Δx_{1}^{T}] (45c)
Note that
E[ΔRΔR^{T}]=W^{−1} (46)
Using (43) and (45),
P_{11}=(G^{T}M_{i}WG)^{−1} (47)
P_{12}=(G^{T}M_{i}WG)^{−1}(G^{T}M_{i}WM_{j}G)(G^{T}M_{j}WG)^{−1} (48)
Likewise,
P_{21}=(G^{T}M_{j}WG)^{−1}(G^{T}M_{j}WM_{i}G)(G^{T}M_{i}WG)^{−1} (49)
Because there are common range measurements used for the two subset solutions, x_{1 }and x_{2}, they are correlated. That is,
P_{12}≠0 (50)
In general, P_{21}≠P_{12} (51)
However, because P_{21}=P_{12}^{T}
P_{21}(k,k)=P_{12}(k,k) for any k (52)
From (41) and (45),
E[ΔX_{1}Δx_{A}^{T}]=rP_{11}+(1−r)P_{12} (53a)
E[Δx_{A}Δx_{1}^{T}]=rP_{11}+(1−r)P_{21} (53b)
Similarly,
P_{A}=E[Δx_{A}Δx_{A}^{T}]=r^{2}(P_{11}+P_{22}−P_{12}−P_{21})−(2P_{22}−P_{12}−P_{21})r+P_{22} (54)
Inserting Eqs. (45a), (53a), (53b), and (54) into Eq. (44) gives
dP_{1}=(1−r)^{2}P_{S} (55)
where
P_{S}≡P_{11}+P_{22}−P_{12}−P_{21} (56)
Similarly, one can derive
dP_{2}=r^{2}P_{S} (57)
Derivation of XTK Protection Level (XPL) as a Function of the Weight Ratio
Define,
σ_{X,1}≡√{square root over (P_{11}(1,1))} (58a)
σ_{X,2}≡√{square root over (P_{22}(1,1))} (58b)
σ_{X,S}≡√{square root over (P_{S}(1,1))}=√{square root over (P_{11}(1,1)+P_{22}(1,1)−2P_{12}(1,1))}{square root over (P_{11}(1,1)+P_{22}(1,1)−2P_{12}(1,1))}{square root over (P_{11}(1,1)+P_{22}(1,1)−2P_{12}(1,1))} (58c)
Then from (55) and (57),
σ_{dX,1}≡√{square root over (dP_{1}(1,1))}=(1−r)σ_{X,S} (59a)
σ_{dX,2}≡√{square root over (dP_{2}(1,1))}=rσ_{X,S} (59b)
With these definitions, and noting that the XPL_{1 }and XPL_{2 }are the sums of the respective thresholds and missed detection buffers,
XPL_{1}=D_{1}+a_{1}=K_{ffd}σ_{dX,1}+K_{md}σ_{X,1}=(K_{ffd}σ_{X,S})(1−r)+K_{md}σ_{X,1} (60a)
XPL_{2}=D_{2}+a_{2}=K_{ffd}σ_{dX,2}+K_{md}σ_{X,2}=(K_{ffd}σ_{X,S})r+K_{md}σ_{X,2} (60b)
where K_{ffd }and were defined earlier in (20a) and (20b). Note that in (20a), P_{ffd }is again divided by 4 in OWAS1 because the total allowable probability is evenly allocated between XTK and ATK, each for a twosided faultfree error distribution.
Note that XPL_{1 }and XPL_{2 }above are linearly decreasing and increasing, respectively, at the same rate as a function of the weight ratio r.
XPL is selected as
XPL=max {XPL_{1}, XPL_{2}} (61)
To obtain the minimum XPL, set
XPL_{1}=XPL_{2} (62)
From Eqs. (60a), (60b), and (62),
Substituting Eq. (63) into Eqs. (60a) and (60b) and then (61),
The weight ratio in (64) provides the smallest XPL. A weight ratio that minimizes APL can also be derived in a similar manner.
In order to satisfy the integrity requirement, recall that XPL and APL must satisfy both of the following conditions:
XPL≦HAL (65a)
APL≦HAL (65b)
Derivations of Navigation Solution Accuracy as a Function of Weight Ratio for OWAS1
The standard deviations of the XTK and ATK position errors can be denoted as σ_{X }and σ_{A}. They are derived from the covariance matrix for the weighted average position error, P_{A }in Eq. (54).
From P_{A},
σ_{X=}√{square root over (P_{A}(1,1))} (66a)
σ_{A=}√{square root over (P_{A}(2,2))} (66b)
In order for RAIM to be available for horizontal position, σ_{X }and σ_{A }are required to satisfy the following conditions:
σ_{X}≦σ_{Xmax} (67a)
σ_{A}≦σ_{Amax} (67b)
where σ_{Xmax }and σ_{Amax }are derived so as to meet the accuracy and faultfree integrity requirements:
σ_{Xmax=σ}_{Amax}=min {95% horizontal accuracy/1.96,HAL/K_{H}} (68)
Determination of an Optimum Weight Ratio for OWAS1
As for the case with OWAS2, there are two ways of selecting a weight ratio: one minimizing XPL (or APL) to the extent possible while still meeting the XTK (or ATK) accuracy requirement, and the other, minimizing σ_{X }(or σ_{A}) to the extent possible while still meeting the integrity availability requirement XPL≦HAL (or APL≦HAL). The processes to determine the optimal weight ratios with the two options are similar. In the following, the process to determine the weight ratios with the former option is described below as an example:
Step 1: If the four inequalities given in Eqs. (65) and (67) are satisfied, then RAIM is declared available.
Step 2: If either or both of the two integrity availability inequalities in Eq. (65) are not satisfied, then RAIM is declared unavailable.
Step 3: If both integrity availability inequalities in Eq. (65) are satisfied, but either or both of the accuracy inequalities in Eq. (67) are not satisfied, then one may search for a new weight ratio to satisfy that inequality. A more precise process to determine the new optimal ratio is described below. Note that if the accuracy requirement is satisfied for only one axis but not both XTK and ATK, a new optimal ratio needs to be determined only for the axis not yet satisfied.
First, an inequality that a weight ratio must satisfy to meet the accuracy requirement for XTK position is derived.
From Eqs. (66a) and (67a)
P_{A}(1,1)<(σ_{Xmax)}^{2} (69)
From Eqs. (54), (56), and (52),
P_{A}(1,1)=P_{S}(1,1)r^{2}−2[P_{22}(1,1)−P_{12}(1,1)]r+P_{22}(1,1) (70)
Define
B_{X}≡P_{22}(1,1)−P_{12}(1,1) (71)
Then, replacing P_{S}(1,1) and P_{22}(1,1) with the notations defined earlier in (58), we obtain
P_{A}(1,1)=σ_{X,S}^{2}r^{2}−2B_{X}r+σ_{X,2}^{2} (72)
Inserting (72) into (69), we obtain
σ_{X,S}^{2}r^{2}−2B_{X}r−[(σ_{Xmax})^{2}−σ_{X,2}^{2}]<0 (73)
Therefore,
r_{1}<r<r_{2} (74)
where
In addition, as presented earlier in (40), the weight ratio must also satisfy
0≦r≦1
Therefore, the XTK accuracy requirement can be met only for weight ratio r that satisfies the inequality
minr_{Xacc}<r<maxr_{Xacc} (76a)
where
minr_{Xacc}=max(r_{1},0)
maxr_{Xacc}=min(r_{2},1) (77b)
For the option of choosing a weight ratio that would minimize XPL to the extent possible, the weight ratio for XTK is determined as follows:
If minr_{Xacc}≦r_{minXPL}≦maxr_{Xacc}, then Opt_r_{X}=r_{minVPL} (78a)
If r_{minXPL}<minr_{Xacc}, then Opt_r_{X}=minr_{Xacc} (78b)
If r_{minXPL}>maxr_{Xacc}, then Opt_r_{X}=maxr_{Xacc} (78c)
If r_{minVPL }satisfies the inequality in (78a), XPL does not need to be recalculated; it is given in Eq. (64). However, if the inequality in (78a) does not hold, then XPL is recalculated using Opt_rX in either Eq. (78b) or Eq. (78c), whichever is applicable, using Eqs. (61a), (60b), and (61). For ATK, both Opt_r_{A}, and APL are also recalculated in a similar manner if necessary.
If both XPL and APL recalculated above satisfy the conditions (XPL≦HAL) and (APL≦HAL) respectively, RAIM is declared available. Otherwise, it is declared not available.
More specifically, in
If either σ_{V }or d_{major }exceeds its limit, then a new set of weight ratios is calculated and the following parameters are also calculated with these new weight ratios: the weighted average solution, VPL, HPL, standard deviations for vertical and horizontal (σ_{V }and d_{major}) and the test statistics (for vertical and horizontal) for fault detection 1120. If either test statistic exceeds its detection threshold 1122, a “fault detected” flag is raised, as indicated in step 1130. If neither test statistic of step 1124 exceeds its detection threshold, σ_{V }and d_{major }are compared with their maximum allowable values. If both σ_{V }and d_{major }are within their respective limits, then RAIM is available and navigation continues 1126; otherwise, a “RAIM not available” flag is raised 1128.
Next, in step 1206, from the N subset position estimates, a protection level constraining pair is identified along each of the XTK and ATK axes. For the two protection level constraining pairs identified, a pair of optimal weights is determined as described earlier. With these weights, the following parameters are calculated: the weighted average solutions (XTK and ATK), XPL and APL, the standard deviations for vertical and horizontal (σ_{X }and σ_{A}), and the test statistics 1208. If either of the two test statistics exceeds its respective fault detection threshold, an attempt to isolate and exclude the bad satellite is made if the FDE function is available. If the FDE attempt is not successful 1210, a fault detection is declared and the user must take an appropriate action according to an operational procedure; if the FDE attempt is successful, then the navigation may continue with the remaining visible satellites 1212.
If neither test statistic exceeds its threshold, the XPL and APL are compared with HAL. If both XPL and APL are less than or equal to HAL and also both σ_{X}≦σ_{Xmax }and σ_{A}≦σ_{Amax}, then RAIM is declared available and the navigation continues using the weighted average solution 1222. If either XPL or APL exceeds HAL, then the user is alerted that RAIN is not available. If both XPL≦HAL and APL≦HAL, but either σ_{X }or σ_{A }exceeds its maximum allowable value, then a new weight ratio is calculated as described earlier, and with the new weight ratios, the following parameters are recalculated: the weighted average solutions (XTK and ATK), XPL and APL, the standard deviations for vertical and horizontal, and the test statistics 1214. If either one of the test statistics newly calculated exceeds its detection threshold, an attempt to isolate and exclude the bad satellite may be made if the FDE function is available.
If the FDE attempt is not successful, a fault detection is declared 1210 and the user must take an appropriate action according to an operational procedure. If the FDE attempt is successful, then the navigation may continue with the remaining visible satellites 1224. If neither test statistic exceeds its threshold, the newly calculated σ_{X }and σ_{A }are compared with their respective maximum allowable values. If σ_{X}≦σ_{Xmax }and σ_{A}≦σ_{Amax}, then RAIN is declared available and the navigation continues 1220. Otherwise RAIM is declared unavailable 1222.
The present invention can apply to a broad range of users. Exemplary users, to name a few, can include aircraft, ships, land vehicles, or even space vehicles. For aircraft navigation, this includes localizer performance with vertical guidance (LPV), approach with precision vertical guidance (APV)II. Other types of aircraft navigation include future precision approaches equivalent to that of the guidance provided by the Instrument Landing System (ILS) for CAT I, II and III. It even includes Real Time Kinematic (RTK) navigation with a differential system for land survey.
The present invention can also work with different types of satellite ranging signal sources including code phase measurements and carrier phase measurements, each with or without differential corrections.
The present invention has been described above with the aid of functional building blocks illustrating the implementation of specified functions and relationships thereof. The boundaries of these functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternate boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed.
The foregoing description of the specific embodiments will so fully reveal the general nature of the invention that others can, by applying knowledge within the skill of the art, readily modify and/or adapt for various applications such specific embodiments, without undue experimentation, without departing from the general concept of the present invention. Therefore, such adaptations and modifications are intended to be within the meaning and range of equivalents of the disclosed embodiments, based on the teaching and guidance presented herein. It is to be understood that the phraseology or terminology herein is for the purpose of description and not of limitation, such that the terminology or phraseology of the present specification is to be interpreted by the skilled artisan in light of the teachings and guidance.
The breadth and scope of the present invention should not be limited by any of the abovedescribed exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents.