Tài liệu GPS - đường dẫn quán tính và hội nhập P8 pptx

Control of the carrier frequency and phase is also required to eliminate uplinkDoppler and to maintain coherence between code and carrier.. 9.3.1 Description of the GUS Algorithm The GUS

 Some error sources are canceled completely:

(a) selective availability and

(b) satellite ephemeris and clock errors

 With other error sources, cancelation degrades with distance:

(a) ionospheric delay error and

(b) tropospheric delay error

 Still other error sources are not canceled at all:

(a) multipath errors and

(b) receiver errors

265

Mohinder S Grewal, Lawrence R Weill, Angus P Andrews

Copyright # 2001 John Wiley & Sons, Inc Print ISBN 0-471-35032-X Electronic ISBN 0-471-20071-9

9.2 LADGPS, WADGPS, AND WAAS

9.2.1 Description of Local-Area DGPS (LADGPS)

LADGPS is a form of DGPS in which the user's GPS receiver receives real-timepseudorange and, possibly, carrier phase corrections from a reference receivergenerally located within the line of sight The corrections account for the combinedeffects of navigation message ephemeris and satellite clock errors (including theeffects of SA) and, usually, atmospheric propagation delay errors at the referencestation With the assumption that these errors are also common to the measurementsmade by the user's receiver, the application of the corrections will result in moreaccurate coordinates [81]

9.2.2 Description of Wide-Area DGPS (WADGPS)

WADGPS is a form of DGPS in which the user's GPS receiver receives correctionsdetermined from a network of reference stations distributed over a wide geographicalarea Separate corrections are usually determined for speci®c error sources, such assatellite clock, ionospheric propagation delay, and ephemeris The corrections areapplied in the user's receiver or attached computer in computing the receiver'scoordinates The corrections are typically supplied in real time by way of ageostationary communications satellite or through a network of ground-basedtransmitters Corrections may also be provided at a later date for post-processingcollected data [81]

9.2.3 Description of Wide Area Augmentation System (WAAS)WAAS enhances the GPS SPS and is available over a wide geographical area TheWAAS being developed by the Federal Aviation Administration, together with otheragencies, will provide WADGPS corrections, additional ranging signals fromgeostationary (GEO) satellites, and integrity data on the GPS and GEO satellites[81]

The GEO Uplink Subsytem includes a closed-loop control algorithm and specialsignal generator hardware These ensure that the downlink signal to the users iscontrolled adequately to be used as a ranging source to supplement the GPS satellites

in view

The primary mission of WAAS is to provide a means for air navigation for allphases of ¯ight in the National Airspace System (NAS) from departure, en route,arrival, and through approach GPS augmented by WAAS offers the capability forboth nonprecision approach (NPA) and precision approach (PA) within a speci®cservice volume A secondary mission of the WAAS is to provide a WAAS networktime (WNT) offset between the WNT and Coordinated Universal Time (UTC) fornonnavigation users

WAAS provides improved en route navigation and PA capability to WAAScerti®ed avionics The safety critical WAAS system consists of the equipment and

software necessary to augment the Department of Defense (DoD) provided GPSSPS WAAS provides a signal in space (SIS) to WAAS certi®ed aircraft avionicsusing the WAAS for any FAA-approved phase of ¯ight The SIS provides twoservices: (1) data on GPS and GEO satellites and (2) a ranging capability.

The GPS satellite data is received and processed at widely dispersed wide-areareference Stations (WRSs), which are strategically located to provide coverage overthe required WAAS service volume Data is forwarded to wide-area master stations(WMSs), which process the data from multiple WRSs to determine the integrity,differential corrections, and residual errors for each monitored satellite and for eachpredetermined ionospheric grid point (IGP) Multiple WMSs are provided toeliminate single-point failures within the WAAS network Information from allWMSs is sent to each GEO uplink subsystem (GUS) and uplinked along with theGEO navigation message to GEO satellites The GEO satellites downlink this data tothe users via the GPS SPS L-band ranging signal (L1) frequency with GPS-typemodulation Each ground-based station=subsystem communicates via a terrestrialcommunications subsystem (TCS) See Fig 9.1

In addition to providing augmented GPS data to the users, WAAS veri®es its ownintegrity and takes any necessary action to ensure that the system meets the WAASperformance requirements WAAS also has a system operation and maintenancefunction that provides status and related maintenance information to FAA airwayfacilities (AFs) NAS personnel

WAAS has a functional veri®cation system (FVS) that is used for earlydevelopment test and evaluation (DT&E), re®nement of contractor site installationprocedures, system-level testing, WAAS operational testing, and long-term supportfor WAAS

GPS satellites

User’s WAAS receiver

GEO subsystem

Wide-area master station Wide-area

reference station-n

Wide-area reference station-1

GEO uplink subsystem

Fig 9.1 WAAS Top Level View

Correction and Veri®cation (C&V) processes data from all WRSs to determineintegrity, differential corrections, satellite orbits, and residual error bounds for eachmonitored satellite It also determines ionospheric vertical delays and their residualerror bounds at each of the IGPs C&V schedules and formats WAAS messages andforwards them to the GUSs for broadcast to the GEO satellites.

C&V's capabilities are as follows:

1 Control C&V Operations and Maintenance (COM) supports the transfer of

®les, performs remotely initiated software con®guration checks, and acceptsrequests to start and stop execution of the C&V application software

2 Control C&V Modes (CMD) manage mode transitions in the C&V tem while the application software is running

subsys-3 Monitor C&V (MCV) reports line replaceable unit (LRU) faults andcon®guration status In addition, it monitors software processes and providesperformance data for the local C&V subsystems

4 Process Input Data (PID) selects and monitors data from the wide-areareference equipment (WREs) Data that passes PID screening is repackagedfor other C&V capabilities PID performs clock and L1 GPS PrecisionPositioning Service L-band ranging signal (L2) receiver bias calculations,cycle slip detection, outlier detection, data smoothing, and data monitoring

In addition, PID calculates and applies the windup correction to the carrierphase, accumulates data to estimate the pseudorange to carrier phase bias,and computes the ionosphere corrected carrier phase and measured slantdelay

5 Satellite Orbit Determination (SOD) determines the GPS and GEO satelliteorbits and clock offsets, WRE receiver clock offsets, and troposphere delay

6 Ionosphere Correction Computation (ICC) determines the L1 IGP verticaldelays, grid ionosphere vertical error (GIVE) for all de®ned IGPs, and L1±L2interfrequency bias for each satellite transmitter and each WRS receiver

7 Satellite Correction Processing (SCP) determines the fast and long-termsatellite corrections, including the user differential range error (UDRE) Itdetermines the WNTand the GEO and WNTclock steering commands [99]

8 Independent Data Veri®cation (IDV) compares satellite corrections, GEOnavigation data, and ionospheric corrections from two independent computa-tional sources, and if the comparisons are within limits, one source is selectedfrom which to build the WAAS messages If the comparisons are not withinlimits, various responses may occur, depending on the data being compared,all the way from alarms being generated to the C&V being faulted

9 Message Output Processing (MOP) transmits messages containing dently veri®ed results of C&V calculations to the GUS processing (GP) forbroadcast

indepen-10 C&V Playback (PLB) processes the playback data that has been recorded bythe other C&V capabilities

11 Integrity Data Monitoring (IDM) checks both the broadcast and the broadcast UDREs and GIVEs to ensure that they are properly bounding theirerrors In addition, it monitors and validates that the broadcast messages aresent correctly It also performs the WAAS time-to-alarm validation[1, 99].

to-be-9.2.3.1 WRS Algorithms Each WRS collects raw pseudorange (PR) andaccumulated delta range (ADR) measurements from GPS and GEO satellitesselected for tracking Each WRS performs smoothing on the measurements andcorrects for atmospheric effects, that is, ionospheric and tropospheric delays Thesesmoothed and atmospherically corrected measurements are provided to the WMS.9.2.3.2 WMS Foreground (Fast) Algorithms The WMS foreground algo-rithms are applicable to real-time processing functions, speci®cally the computation

of fast correction, determination of satellite integrity status and WAAS messageformatting This processing is done at a 1-HZ rate

9.2.3.3 WMS Background (Slow) Algorithms The WMS backgroundprocessing consists of algorithms that estimate slowly varying parameters Thesealgorithms consist of WRS clock error estimation, grid ionospeci®c delay computa-tion, broadcast ephemeris computation, satellite orbit determination, satellite ephe-meris error computation, and satellite visibility computation

9.2.3.4 Independent Data Veri®cation and Validation Algorithms Thisincludes a set of WRS and at least one WMS, which enable monitoring the integritystatus of GPS and the determination of wide-area DGPS correction data Each WRShas three dual frequency GPS receivers to provide parallel sets of measurement data.The presence of parallel data streams enables Independent Data Veri®cation andValidation (IDV&V) to be employed to ensure the integrity of GPS data and theircorrections in the WAAS messages broadcast via one or more GEOs With IDV&Vactive, the WMS applies the corrections computed from one stream to the data fromthe other stream to provide veri®cation of the corrections prior to transmission Theprimary data stream is also used for the validation phase to check the active (alreadybroadcast) correction and to monitor their SIS performance These algorithms arecontinually being improved The latest versions can be found in references [48, 96,

97, 137, 99] and [98, pp 397±425]

9.3 GEO UPLINK SUBSYSTEM (GUS)

Corrections from the WMS are sent to the ground uplink subsystem (GUS) foruplink to the GEO The GUS receives integrity and correction data and WAASspeci®c messages from the WMS, adds forward error correction (FEC) encoding,and transmits the messages via a C-band uplink to the GEO satellites for broadcast

to the WAAS user The GUS signal uses the GPS standard positioning service

waveform (C=A-code, BPSK modulation); however, the data rate is higher (250bps) The 250 bps of data are encoded with a one-half rate convolutional code,resulting in a 500-symbols=s transmission rate.

Each symbol is modulated by the C=A-code, a 1:023  106-chips=s pseudorandom sequence to provide a spread-spectrum signal This signal is then BPSKmodulated by the GUS onto an IF carrier, upconverted to a C-band frequency, anduplinked to the GEO It is the C=A-code modulation that provides the rangingcapability if its phase is properly controlled

Control of the carrier frequency and phase is also required to eliminate uplinkDoppler and to maintain coherence between code and carrier The GUS monitors theC-band and L1downlinks from the GEO to provide closed-loop control of the PRNcode and L1 carrier coherency WAAS short- and long-term code carrier coherencerequirements are met

9.3.1 Description of the GUS Algorithm

The GUS control loop algorithm ``precorrects'' the code phase, carrier phase, andcarrier frequency of the GEO uplink signal to maintain GEO broadcast code±carriercoherence The uplink effects such as ionospheric code±carrier divergence, uplinkDoppler, equipment delays, and frequency offsets must be corrected in the GUScontrol loop algorithm

Figure 9.2 provides an overview of the functional elements of the GUS controlloop The control loop contains algorithm elements (shaded boxes) and hardwareelements that either provide inputs to the algorithm or are controlled or affected byoutputs from the algorithm The hardware elements include a WAAS GPS receiver,GEO satellite, and GUS signal generator

Downlink ionospheric delay is estimated in the ionospheric delay and rateestimator using pseudorange measurements from the WAAS GPS receiver on L1and L2(downconverted from the GEO C-band downlink at the GUS) This is a two-state Kalman ®lter that estimates the ionospheric delay and delay rate

At each measurement interval, a range measurement is taken and fed into therange, rate, and acceleration estimator This measurement is the average between thereference pseudorange from the GUS signal generator PRsign and the receivedpseudorange from the L1downlink as measured by the WAAS GPS Receiver …PRgeo†and adjusted for estimated ionospheric delay …PRiono† The equation for the rangemeasurement is then

z ˆ1

2‰…PRgeo PRiono† ‡ PRsignŠ TCup TL1dwnS;where TCupˆ C-band uplink delay …m†

TL1dwnS ˆ L1receiver delay of the GUS …m†

The GUS signal generator is initialized with a pseudorange value from satelliteephemeris data This is the initial reference from which corrections are made.The range, rate and acceleration estimator is a three-state Kalman ®lter that drivesthe frequency and code control loops

The code control loop is a second-order control system The error signal for thiscontrol system is the difference between the WAAS pseudorange (Prsign† and theestimated pseudorange from the Kalman ®lter The loop output is the code rateadjustments to the GUS signal generator.

The frequency control loop has two modes First, it adjusts the signal generatorfrequency to compensate for uplink Doppler effects This is accomplished using a

®rst-order control system The error signal input is the difference between the L1Doppler frequency from the WAAS GPS receiver and the estimated range rate(converted to a Doppler frequency) from the Kalman ®lter

Once the frequency error is below a threshold value, the carrier phase iscontrolled This is accomplished using a second-order control system The errorsignal input to this system is the difference between the L1carrier phase and a carrierphase estimate based on the Kalman ®lter output This estimated range is converted

to carrier cycles using the range estimate at the time carrier phase control starts as areference Fine adjustments are made to the signal generator carrier frequency tomaintain phase coherence [35, 47±49, 94]

9.3.2 In-Orbit Tests

Two separate series of in-orbit tests (IOTs) were conducted, one at the COMSATGPS Earth Station (GES) in Santa Paula, California with Paci®c Ocean Region(POR) and Atlantic Ocean Region-West (AOR-W) I-3 satellites and the other at theCOMSAT GES in Clarksburg, Maryland, using AOR-W The IOTs were conducted

Iono delay rate estimator

PRiono

PRiono

Range, rate acceleration estimator

Code control loop

Frequency control loop

GUS signal generator

Iono delay estimate

Iono delay estimate

Range & rate estimates

bit Control

Range, rate residual

Iono delay estimator

Range measurement

L1 doppler frequency L1 carrier phase GUS control loop algorithim

Σ

Fig 9.2 GUS control loop block diagram.

to validate a prototype version of the GUS control loop algorithm Data wascollected to verify the ionospheric estimation and code±carrier coherence perfor-mance capability of the control loop and the short±term carrier frequency stability ofthe I-3 satellites with a prototype ground station The test results were also used tovalidate the GUS control loop simulation.

Figure 9.3 illustrates the IOTsetup at a high level Prototype ground stationhardware and software were used to assess algorithm performance at two differentground stations with two different Inmarsat-3 satellites

9.3.3 Ionospheric Delay Estimation

The GUS control loop estimates the ionospheric delay contribution of the GEO band uplink to maintain code±carrier coherence of the broadcast SIS Figures 9.4±9.6 provide the delay estimates for POR using the Santa Paula GES and AOR-Wusing both the Santa Paula and Clarksburg GES Each plot shows the estimatedionospheric delay (output of the two-state Kalman ®lter) versus the calculated delayusing the L1and C pseudorange data from a WAAS GPS receiver Calculated delay

C-is noC-isier and varying about 1 m=s, whereas the estimated delay by the Kalman ®lter

is right in middle of the measured delay, as shown in Figures 9.4±9.6 Delaymeasurements were calculated using the equation

Ionospheric delay ˆPRL1 PRC tau L1‡ tau C

1 ‰L1freqŠ2=‰C freqŠ2

AOR-W POR HPA

Cup

L1 dwn L2

Cdwn Frequency reference

Frequency reference Frequency reference Measurement

data

Signal generator upconverterC-band

C-band to L2 downconverter L1-omni

FTS atomic clock IF

Fig 9.3 IOT test GUS setup.

where PRL1ˆ L1pseudorange …m†

PRCˆ C pseudorange …m†

tau L1ˆ L1downlink delay …m†

tau C ˆ C downlink delay …m†

L1freq ˆ L1frequency; ˆ 1575:42 MHz

Cfreq ˆ C frequency; ˆ 3630:42 MHz

The ionosphere during the IOTs was fairly benign with no high levels of solaractivity observed Table 9.1 provides the ionospheric delay statistics (in meters)

Fig 9.4 Measured and estimated ionospheric delay, POR, Santa Paula.

Fig 9.5 Measured and estimated ionospheric delay, AOR-W, Santa Paula.

between the output of the ionospheric Kalman ®lter in the control loop, and thecalculated delay from the WAAS GPS receiver's L1 and L2 pseudoranges Thestatistics show that the loop's ionospheric delay estimation is very close (low RMS)

to the ionospheric delay calculated using the measured pseudorange from the WAASGPS receiver

9.3.4 Code±Carrier Frequency Coherence

The GEO's broadcast code±carrier frequency coherence requirement is speci®ed inthe WAAS System Speci®cation and Appendix A of reference [106] It states:

The lack of coherence between the broadcast carrier phase and the code phase shall belimited The short term fractional frequency difference between the code phase rate andthe carrier frequency will be less than 5  10 11: That is,

fcode1:023 MHz

1575:42 MHz

Fig 9.6 Measured and estimated ionospheric delay, Clarksburg.

TABLE 9.1 Observed RMS WAAS Ionospheric

Correc-tion Errors

Santa Paula GES, Oct 10, 1997, POR 0.20

Santa Paula GES, Dec 1, 1997, AOR-W 0.45

Clarksburg GES, Mar 20, 1998, AOR-W 0.34

Over the long term, the difference between the broadcast code phase (1=1540) and thebroadcast carrier phase will be within one carrier cycle, one sigma This does notinclude code-carrier divergence due to inospheric refraction in the downlink propaga-tion path.

For the WAAS program, short term is de®ned as less than 10 s and long term lessthan 100 s

Pseudorange minus the ionospheric estimates averaged over t seconds isexpressed as

For long term code±carrier coherence calculations, the difference between thepseudorange and the phase measurements is given by

DPR PHˆ ‰FPR=lL1Š FPH cycles/s,where lL1is the wavelength of the L1carrier frequency and ``long term coherence''equals jDPR PH…t ‡ 100† DPR PH…t†j cycles

For short-term code±carrier coherence calculations, the difference between thepseudorange and the phase measurements is given by

10  c (speed of light)and ``short term coherence'' is jdPR PH…t ‡ 10† dPR PH…t†j

The IOT long- and short-term code±carrier results from Santa Paula andClarksburg are shown in Table 2 The results indicate that the control loop algorithmperformance meets the long- and short-term code±carrier requirements of WAASwith the I-3 satellites

9.3.5 Carrier Frequency Stability

Carrier frequency stability is a function of both the uplink frequency standard, GUSsignal generator, and I-3 transponder The GEO's short-term carrier frequencystability requirement is speci®ed in the WAAS System Speci®cation and Appendix

A of reference [106] It states:

The short term stability of the carrier frequency (square root of the Allan variance) atthe input of the user's receiver antenna shall be better than 5  10 11 over 1 to 10seconds, excluding the effects of the ionosphere and Doppler

The Allan variance [2] is calculated on the second difference of L1 phase datadivided by the center frequency over 1±10 s Effects of smoothed ionosphere andDoppler are compensated for in the data prior to this calculation Test results in Table9.3 show that the POR and AOR-W I-3 GEOs, in conjunction with WAAS groundstation equipment, meet the short-term carrier frequency stability requirement ofWAAS

9.4 GEO UPLINK SUBSYSTEM (GUS) CLOCK STEERING

ALGORITHMS

The local oscillator (cesium frequency standard) at the GUS is not perfectly stablewith respect to WAAS network time (WNT) Even though the cesium frequencystandard is very stable, it has inherent drift Over a long period of operation, as in the

TABLE 9.2 Code Carrier Coherence

a Data averaging 30 s for short term.

b Data averaging 60 s for long term.

TABLE 9.3 Carrier Frequency Stability Requirements Satis®ed

WAAS scenario, this slow drift will accumulate and result in an offset so large thatthe value will not ®t in the associated data ®elds in the WAAS Type 9 message This

is why a clock steering algorithm is necessary at the GUS This drifting effect willcause GUS time and WNTto slowly diverge The GUS can compensate for this drift

by periodically re-synchronizing the receiver time with the WNTusing the estimatedreceiver clock offset ‰a0…tk†Š This clock offset is provided by the WMS in WAASType 9 messages (See Fig 9.7.)

GUS steering algorithms for the primary and backup GEO uplink subsystems[103, 46] are discussed in the next section

The primary GUS clock steering is closed loop via the signal generator, GEO,WRS, WMS, to the GUS processor The backup GUS clock steering is an open-loopsystem, because the backup does not uplink to the GEO The clock offset iscalculated using the estimated range and the range calculated from the C&Vprovided GEO positions

The GUS also contains the WAAS clock steering algorithm This algorithm usesthe WAAS Type 9 messages from the WMS to align the GEO's epoch with the GPSepoch The WAAS Type 9 message contains a term referred to as a0, or clock offset.This offset represents a correction, or time difference, between the GEOs epoch andWNT WNT is the internal time reference scale of WAAS and is required to track theGPS time scale, while at the same time providing the users with the translation toUTC Since GPS master time is not directly obtainable, the WAAS architecturerequires that WNTbe computed at multiple WMSs using potentially differing sets ofmeasurements from potentially differing sets of receivers and clocks (WAAS

GPS antenna

GPS receiver

WAAS message processor (steering algorithim)

GUS processor GUS

Type 9

message

Type 9 message

RS-232 data

10 MHz

1 PPS

Steering commands WNT

Cesium frequency standard

Fig 9.7 WMS to GUS clock steering.

9.4... within one carrier cycle, one sigma This does notinclude code-carrier divergence due to inospheric refraction in the downlink propaga-tion path.

For the WAAS program, short term is de®ned