GPSDP Gage

research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain
GPS Data processing: Code and Phase

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., and Ramos P. gAGE/UPC, Barcelona, Spain
gAGE
http://gage1.upc.es
JEAGAL, 2004-2005 Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.
"This document has been produced with the financial assistance of the European Union. The contents of this document are the sole responsibility of the authors and can under no circumstances be regarded as reflecting the position of the European Union".
gAGE
JEAGAL, 2004-2005
Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.
This course is based on the book edited by gAGE/UPC: GPS data processing: code and phase. Algorithms, techniques and recipes. (available at http://gage1.upc.es) These slides show some exercises and examples whose data files (with actual GPS data) and corresponding software (with source code) are provided in the book.
gAGE
JEAGAL, 2004-2005
Summary
Introduction Introduction GPS measurements and their combinations GPS measurements and their combinations
The RINEX files The RINEX files Ionospheric combination Ionospheric combination Ionosphere-Free combination Ionosphere-Free combination Wide-lane and Narrow-lane combinations Wide-lane and Narrow-lane combinations
Satellite coordinates Satellite coordinates The model The model Navigation equations Navigation equations Code and phase differential positioning. Code and phase differential positioning. Floating versus fixing ambiguities Floating versus fixing ambiguities
gAGE
gAGE
Specific Objectives:
To learn about GPS observables (code and phase), their characteristics, properties, combinations and applications. To learn how to calculate satellites orbits and clocks from navigation message. To know the achievable precision. To learn how to model pseudodistance for code and phase measurements. This includes calculation of: 1) Coordinates at emission epoch, 2) Ionospheric delay (Klobuchar model), 3) Tropospheric delay, 4) relativistic correction, 5) clocks offsets and satellite instrumental delays, 6) phase wind-up, etc. To learn how to set and solve the navigation equation system using least-squares or Kalman filter (algorithm level). To know how to use phase differential positioning: Floating and fixing ambiguities.
gAGE
To get tools and skills to process and analize GPS To get tools and skills to process and analize GPS data. Implement algorithms defined in GPS/SPS-SS data. Implement algorithms defined in GPS/SPS-SS
gAGE
JEAGAL, 2004-2005
Introduction
A ship determines its location from a set on lighthouses that send an acoustic signal at a known time.
Knowing the emission time Knowing the emission time t0 in the lighthouse and t0 in the lighthouse and the reception time t1 in the reception time t1 in the ship, the traveling time the ship, the traveling time t1-t0, and the geometric t1-t0, and the geometric range =v(t1-t0) may be range =v(t1-t0) may be computed. computed. With only one lighthouse With only one lighthouse there is a whole circumference there is a whole circumference of possible locations of possible locations
=v(t1-t0)
gAGE
JEAGAL, 2004-2005
gAGE
With two lighthouses there With two lighthouses there are two possible solutions. are two possible solutions. But, one of them is not on But, one of them is not on the sea! the sea!
JEAGAL, 2004-2005
With three lighthouses a With three lighthouses a single solution is found single solution is found
gAGE
JEAGAL, 2004-2005
10
Errors in the clocks (lighthouses and ship) synchronism affects the accuracy
True range or
Geometric range
The ranges are measured by means the traveling time of the acoustic signal from the lighthouses to the ship. Thence, the synchronism errors between the lighthouses and ship clocks will degrade the positioning accuracy. JEAGAL, 2004-2005
gAGE
Pseudorange or
apparent distance
11 due to the error clocks
SUMMARY: The positioning system is based on: SUMMARY: The positioning system is based on: To know the coordinates of the lighthouses To know the coordinates of the lighthouses To know the ranges from the ship to the lighthouses To know the ranges from the ship to the lighthouses To solve a geometric problem. To solve a geometric problem.
True range or
Geometric range
NOTE: the ranges are measured by means NOTE: the ranges are measured by means the traveling time of the acoustic signal the traveling time of the acoustic signal from the lighthouses to the ship. from the lighthouses to the ship. Thence, the synchronism errors between Thence, the synchronism errors between Pseudorange or the lighthouses and ship clocks will degrade the lighthouses and ship clocks will degrade apparent distance the positioning accuracy. Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P. JEAGAL, 2004-2005 the positioning accuracy. 12
due to the error clock.
gAGE
How GPS Works

Satellites broadcast orbit and clock data Satellite coordinates and clock offset
Lighthouses coordinates
P
Receiver measures traveling time from satellite to receiver Pseudorange (P)
Lighthouses-ship ranges.
gAGE
Thence, the receiver coordinates are found solving a Thence, the receiver coordinates are found solving a geometrical problem: from sat. coordinatesSalazar D., Ramos P. 13 Hernndez-Pajares M., Juan M., Sanz J, JEAGAL, 2004-2005 geometrical problem: from sat. coordinates and ranges and ranges
How GPS Works

gAGE
One of the solutions is not on the Earth surface.
JEAGAL, 2004-2005
14
How GPS Works

gAGE
Receiver location
JEAGAL, 2004-2005
15
How GPS Works Lesson 3:

GPS measurements and its combinations

Measurements: Ranges
Pseudoranges are computed from the traveling time sat-rec Several error sources affect these measurements
gAGE
Receiver location
JEAGAL, 2004-2005
16
How GPS Works

Satellite location
Satellite coordinates and clock offsets are computed from the navigation message: (orbit.f)
Lesson 4:
GPS Orbits and clocks
Measurements: Ranges
Pseudoranges are computed from the traveling time sat-rec Several error sources affect these measurements
gAGE
Receiver location
JEAGAL, 2004-2005
17
How GPS Works

Satellite location research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain
Measurements: Ranges Atmospheric propagation: IONO, TROPO

Pseudoranges are computed from the traveling time sat-rec Several error sources GPS measurements affect these measurements modeling (code) Relativistic effects
Lesson 5:
MODEL:
Atmospheric propag., relativistic effects, clocks and instrum. delays are modeled and removed. And the navigation equations are built
gAGE
Receiver location
JEAGAL, 2004-2005
18
How GPS Works

Satellite location research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain
1 xio x1 1 yio yy1 xio x yio Measurements: 1 1 io 1 11 1 io 1 Pi1111 io11+dt 1111 kk 111 Rangesio P + + dt P io ioio dt x2 2 Pi 2io + dt2 k2k2 xio x yio y 2 2 2 2 2 xio yio are ++dt 2 k PP 2io 22+dt 2 2k 22 Pseudoranges 2 y P P 2 +dt io 2 i i dt kk = computed from 2 the io = io io 2 io io io traveling time sat-rec ........ ........ ........ ........ .......... .......... P nnn nnnn+dt nnn nnn Several nerror sources Pi n Pi n io++dt nkkk io P io io +dt k n dt io x n xx x yio these affect yio yy n io n n measurements n n io ioio io
zio 1 zio z 1 1 z
2 zio z 2 zio z 2 2 io io
1 io io
Lesson 6:
Solving the navigation Equations
n zio z n zio z n n io io
11 11 11
x iixi i y ii iiy z iiz ii cdtii cdti i
MODEL:
Atmospheric propag., relativistic effects, clocks and instrum. delays are modeled and removed. And the navigation equations are built
Navigation equations
The geometric problem is linearized, and Weighted Least Mean Squares or Kalman filter are used to compute the solution.
JEAGAL, 2004-2005
gAGE
19
How GPS Works

Lesson 7:
Code and phase differential positioning. Floating/fixing ambiguities
gAGE
Receiver location
JEAGAL, 2004-2005
20
Lesson 3
GPS measurements and their combinations: The RINEX files
gAGE
JEAGAL, 2004-2005
21
GPS SIGNAL STRUCTURE

Two carriers in L-band: L1=154 fo=1575.42 Mhz L2=120 fo=1227.60 Mhz where fo=10.23 Mhz
C/A-code for civilian users [C1(t)] P-code only for military and authorized users [P(t)] Navigation message with satellite ephemeris and clock corrections
[D(t)]
P(Y)
C/A P(Y)
1227.6 MHz
1575.42 MHz
L2
L1
gAGE
L1 (t ) = a1 P (t ) D(t ) sin( f1t ) + a1C (t ) D(t ) cos( f1t ) L2 (t ) = a2 P(t ) D(t ) sin( f 2t )
22
GPS Pseudorange Measurements

L1 (t ) = a1 P(t ) D(t ) sin( f1t ) + a1C (t ) D(t ) cos( f1t ) L2 (t ) = a2 P (t ) D(t ) sin( f 2t )
binary code P
C1,P1, P2
P1= c t= c [trec(TR)-tsat(TS)]
From hereafter we will call: From hereafter we will call: C1 pseudorange computed from C binary code (on frequency 1) C1 pseudorange computed from C binary code (on frequency 1)

gAGE
P1 pseudorange computed from P binary code (on frequency 1) P1 pseudorange computed from P binary code (on frequency 1) P2 pseudorange computed from P binary code (on frequency 2) P2 pseudorange computed from P binary code (on frequency 2)
JEAGAL, 2004-2005
23
L1 ( t ) = a1 P ( t ) D ( t ) sin( f1t ) + a1C ( t ) D ( t ) cos( f1t )

GPS Phase Measurements f

=
= fr fe =
& c
fe
L2 ( t ) = a 2 P ( t ) D ( t ) sin( f 2 t )
Carrier phase L
& = c
f + c tt fe
pseudoranges L1, L2 (containing unknown bias) can be also measured from the carrier phases
C1,P1, P2
L1(t), L2(t) (integrated Doppler)

From hereafter we will call:
L1, L2
gAGE
L1 phase measur. computed from the carrier phase on frequency 1 L2 phase measur. computed from the carrier phase on frequency 2
C1 pseudorange computed from C binary code (on frequency 1)

P1 pseudorange computed from P binary code (on frequency 1) P2 pseudorange computed from P binary code (on frequency 2)
24
Phase and Code pseudorange measurements

P1
P1
P1 + clock offset 20.000 Km

P1 is basically the geometric range () between satellite and receiver, plus the relative clock offset. The range varies in time due to the satellite motion relative to the receiver.
gAGE
JEAGAL, 2004-2005
Hernndez-Pajares P1 is an absolute measurement M., Juan M., Sanz J, Salazar D., Ramos P. (unambiguous)
25
Phase and Code pseudorange measurements

Relative measurement (shifted by the unknown ambiguity N)

Each time that the receiver loose the phase lock, the unknown ambiguity changes by an integer number of
gAGE
L1 + clock offset + N
JEAGAL, 2004-2005
26
Code and phase measurements

Code (unambiguous but noisier)
Ambiguity
gAGE
Carrier Phase (ambiguous but precise)
JEAGAL, 2004-2005
27
GPS measurements: Code and Phase pseudoranges

L1 (t ) = a1 P (t ) D(t ) sin( f1t ) + a1C (t ) D(t ) cos( f1t ) L2 (t ) = a2 P (t ) D(t ) sin( f 2t )

Wavelength (chip)
Antispoofing (A/S): Antispoofing (A/S): The code P is encrypted to Y. The code P is encrypted to Y. signal Only theGPSC at Only thecode C at code
frequency L1 is available. frequency L1 is available. C1 P1 (Y1): encrypted P2 (Y2): encrypted L1 L2
noise (1% of ) [*]
Main characteristics
Code measurements
300 m 30 m 30 m 19.05 cm 24.45 cm 3m 30 cm 30 cm 2 mm 2 mm Unambiguous but noisier
gAGE
Phase measurements Precise but ambiguous
[*] the codes can be smoothed with the phases in order to reduce noise Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P. JEAGAL, 2004-2005 (i.e, C1 smoothed with L1 50 cm noise)
28
RINEX measurement file RINEX measurement file

HEADER
gAGE
MEASUREMENTS
JEAGAL, 2004-2005
29

gAGE
JEAGAL, 2004-2005
30

Measurement time (receive time tags) Number of tracked satellites
gAGE
Epoch flag 0: OK
JEAGAL, 2004-2005
One satellite per row

31

Synthetic P2 (A/S=on)
gAGE
JEAGAL, 2004-2005
S/N indicatorsLoss M., Juan M., Sanz J, Salazar D., Ramos P. Hernndez-Pajares of lock indicator
32
Pseudorange modeling
P= c t= c [trec(TR)-tsat(TS)]
sat sat rec rec
sat rec
+ c ( dt rec dt
Geometric range
sat
) +
Clock offsets
gAGE
sat sa sat = rel rec + T ro p rect + Io n rec + K rec + K sat +
Ionospheric delay Relativistic effects noise Juan M., Sanz J, Salazar JEAGAL, 2004-2005 33 Tropospheric delay Hernndez-Pajares M.,Instrumental D., Ramos P. delays
gAGE
34
gAGE
35
Exercise 1:
a) Using Exercise 1: the file 95oct18casa___r0.rnx, generate the txt file 95oct18casa.a (with data ordered in columns). b) Plot code and phase measurements for satellite PRN28 and discuss the results.
Resolution: a) cat 95oct18casa___r0.rnx| rnx2txt > 95oct18casa.a b) See next plots:
gAGE
JEAGAL, 2004-2005
36
An example of program to read the RINEX: rnx2txt

RINEX file
rnx2txt
sta Doy sec PRN L1
txt file
L2
C1/P1
P2
cambiar
gAGE
The RINEX file is convert to a columnar format to easily plot its contents and to analyze the measurements (the public domain free JEAGAL, 2004-2005 tool gnuplot is used in the book to Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P. make the plots).
37
Code measurements
P1
The geometry is The geometry is the dominant term in the dominant term in the plot. The pattern the plot. The pattern in the figures is due to in the figures is due to the variation of the variation of
P1
gAGE
sat 1sta
JEAGAL, 2004-2005
sat sta
+ c (dtsta dt ) + rel + Trop + Ion

sat sat sta sat sta
sat 1sta
38 + K1sta + K1 +
sat
Code measurements
P2
Similar plot for code Similar plot for code measurements at f2. measurements at f2. Notice that Notice that Ionosphere (Ion) and Ionosphere (Ion) and Instrumental delays (K) Instrumental delays (K) depend on frequency. depend on frequency.
sat sat sat 39 P2sat = sta + c (dtsta dt sat ) + relsta + Tropsta + Ion2sat + K2sta + K2sat + sta sta
gAGE
Code and Phase measurements Code and Phase measurements

Code measurements: C1,P1,P2
C1sta ; P sat = sat + c (dt dt sat ) + rel sat + Tropsat + Ion sat + K + K sat + 1sta 1sta 1sta 1 sta sta sta sta
sat sat sat P2 sta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ion2 sta + K2 sta + K2 + sat sat sat
sat
Phase measurements: L1,L2
Frequency dependent
sat sat sat sat L1sat = sta + c (dtsta dt sat ) + relsta + Tropsta Ion1sta + k1sta + k1sat + 1 N1 + w1 + sta
gAGE
phase Ambiguities
N1, N2 are integers
Wind Up
JEAGAL, 2004-2005
40
Phase measurements
The geometry is The geometry is the dominant term in the dominant term in the plot. The pattern the plot. The pattern in the figures is due to in the figures is due to the variation of . the variation of . The curves are broken The curves are broken when the receiver loss when the receiver loss the lock (cycle-slip). the lock (cycle-slip).
sat sat sat Hernndez-Pajares + k M., Sanz J, Salazar D., N P. L1sat = JEAGAL, 2004-2005 sta dt sat ) + relsta + Tropsta Ion1sat M., Juansta + k1sat + 1Ramos+ w1 + + c (dt 41 1 1 sta sta sta
gAGE
When a cycle-slip happens, the phase measurement L changes by un unknown integer number of cycles (N)
Phase measurements
The geometry is The geometry is the dominant term in the dominant term in the plot. The pattern the plot. The pattern in the figures is due to in the figures is due to the variation of . the variation of . The curves are broken The curves are broken when the receiver loss when the receiver loss the lock (cycle-slip). the lock (cycle-slip).
sat sat sat sat L2sat = JEAGAL, 2004-2005sta dt sat ) + relsta + TropstaHernndez-Pajares+ k2staM.,+ k2J, Salazar2 N2 + w2 + + c (dt Ion2sta M., Juan Sanz sat + D., Ramos P. 42 sta sta
gAGE
When a cycle-slip happens, the phase measurement L changes by un unknown integer number of cycles (N)
Combination of measurements:
Ionospheric combination Ionosphere-Free combination Wide-lane and Narrow-lane combinations
gAGE
JEAGAL, 2004-2005
43
1. IONOSPHERIC COMBINATION 1. IONOSPHERIC COMBINATION
L1-L2 P2 -P1 P2 group L1-L2 research-P1 of Astronomy and Geomatics gAGE
PI= P2-P1=Iono+ctt PI= P2-P1=Iono+ctt LI= L1-L2= Iono+ctt+Ambig LI= L1-L2= Iono+ctt+Ambig
Ambiguity
Technical University of Catalonia, Barcelona, Spain

sat sat sat
sat sat sat C1sta ; P sta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ion1sta + K1sta + K1 + 1 sat sat sat sat P2 sta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ion2 sta + K2 sta + K2 + sat sat sat
sat sat sat L2sat = sta + c (dtsta dt sat ) + relsta + Tropsta Ion2sat + k2sta + k2sat + 2 N2 + w2 + sta sta
phase Ambiguities 44
GPS observables PI= P2-P1=Iono+ctt PI= P2-P1=Iono+ctt


The pattern corresponds to The pattern corresponds to the ionospheric refraction the ionospheric refraction ((Ion),because the other Ion), because the other terms ((K)are constant. terms K) are constant. Notice that code Notice that code measurements are noisier. measurements are noisier.
LI= L1-L2= Iono+ctt+Ambig LI= L1-L2= Iono+ctt+Ambig
Pij= c t= c [trec(TR)-tems(TS)]
Ambiguity
gAGE
JEAGAL, 2004-2005
45

The ionospheric delay (Ion) is proportional to the electron density integrated along the ray path (STEC).
40.3 Ion = 2 STEC f

r r [ GPSreceiver ]
STEC =
r r [ GPStransmitter ]
r N e (r , t )dr
The iono refraction The iono refraction depends on: depends on: Geographic location Geographic location Time of day Time of day Time with respect to Time with respect to solar cycle (11y). solar cycle (11y).
gAGE
JEAGAL, 2004-2005
46

The ionospheric delay (Ion) is proportional to the electron density integrated along the ray path (STEC).
40.3 Ion = 2 STEC f

r r [ GPSreceiver ]
STEC =
r r [ GPStransmitter ]
r N e (r , t )dr
The iono refraction depends on: The iono refraction depends on: Geographic location Geographic location Time of day Time of day Time with respect to solar Time with respect to solar cycle (11y). cycle (11y).
gAGE
Ambiguity
JEAGAL, 2004-2005
47

PI= P2-P1=Iono+ctt PI= P2-P1=Iono+ctt LI= L1-L2= Iono+ctt+Ambig LI= L1-L2= Iono+ctt+Ambig
NOTE: Ionosphere delays code Ionosphere advances phase
Ambiguity

sat sat sat
sat sat sat C1sta ; P sta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ion1sta + K1sta + K1 + 1 sat sat sat sat P2 sta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ion2 sta + K2 sta + K2 + sat sat sat
gAGE
sat sat sat L2sat = sta + c (dtsta dt sat ) + relsta + Tropsta Ion2sat + k2sta + k2sat + 2 N2 + w2 + sta sta
48
2. IONOSPHERE-FREE COMBINATION (Pc,Lc) Ionospheric-Free Combination

and can be removed up to 99.9% combining f1 and f2 signals:
The ionospheric refraction depends on the inverse of the squared frequency
40.3 Ion = 2 STEC f

2 2
f 1 P1 f 2 P 2 Pc = f 1 2 f 22
2 2
f1 L1 f 2 L 2 Lc = f 1 2 f 22
sat sat sat sat Pcsta = sta + c (dtsta dt sat ) + relsta + Tropsta + Kc sta +
sat sat sat sat Lcsta = sta + c (dtsta dt sat ) + relsta + Tropsta + kcsta + kc sat + c Rc + wc +
The ionospheric refraction has been removed in Lc and Pc c = 10.7 cm The Rc ambiguities are NOT integers!! Rc = W N1 W N 2 1 2
gAGE
49
Comments: Comments: Two-frequency receivers are needed to apply the Two-frequency receivers are needed to apply the ionosphere-free combination. ionosphere-free combination. If a one-frequency receiver is used, a ionospheric If a one-frequency receiver is used, a ionospheric model must be applied to remove the ionospheric model must be applied to remove the ionospheric refraction. The GPS navigation message provides the refraction. The GPS navigation message provides the parameters of the Klobuchar model which accounts parameters of the Klobuchar model which accounts for more than 60% of the ionospheric delay. for more than 60% of the ionospheric delay.
gAGE
JEAGAL, 2004-2005
50
Narrow-lane (Pw) and Wide-lane Combination Lw Ionospheric-Free Combination

The wide-lane combination Lw provides a signal with a large wavelength (=86.2cm~4*1). This makes it very useful for detecting cycle-slips through the Melbourne-Wbbena combination: Lw-Pw
f 1 P1 + f 2 P 2 Pw = f1 + f 2
f1 L1 f 2 L 2 Lw = f1 f 2
The same sign
sat sat sat sat Pwsta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ionw sat + Kw sta + Kw sat + sta sat sat sat sat Lwsta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ionw sat + kw sta + kw sat + w Nw + sta
gAGE
The ambiguities Nw are INTEGERS!

No wind-up
51
Detecting cycle-slips
This cycle-slip involves millions of cycles it is easy to detect!!
There is a cycleslip of only one cycle (~20cm) How to detect it?

gAGE
52
Exercise 2:
a) Using the file 95oct18casa___r0.rnx, generate the txt file 95oct18casa.a (with data ordered in columns). b) Insert a cycle-slip of one wavelength (19cm) in L1 measurement at t=5000 s (and no cycle-slip in L2). c) Plot the measurements L1, L1-P1, LC-PC, Lw-Pw and LI and discuss which combination/s should be used to detect the cycle-slip.
Resolution: a) cat 95oct18casa___r0.rnx| rnx2txt > 95oct18casa.a b) cat 95oct18casa.a | gawk {if ($4==18)
print $3,$5,$6,$7,$8} > s18.org cat s18.org | gawk {if ($1>=5000) $2=$2+0.19; printf %s %f %f %f %f \n, $1,$2,$3,$4,$5} > s18.cl
gAGE
c) See next plots:

53
The geometry is the dominant term in the plot. The variation The geometry is the dominant term in the plot. The variation of in 1 sec may be hundreds of meters, many times greater than of in 1 sec may be hundreds of meters, many times greater than the cycle-slip (19 cm) the cycle-slip (19 cm) the variation of shadows the cycle-slip! the variation of shadows the cycle-slip!
L1 (without the cycle-slip) L1 (with the cycle-slip)

1 unit = 19 cm (L1 cycles)
A jump of =19 cm (one cycle in L1) has been introduced in L1 at t=5000s
sat sat sat L1sat = JEAGAL, 2004-2005 sta dt sat ) + relsta + Tropsta Ion1sat M., Juansta + k1sat + 1Ramos+ w1 + + c (dt Hernndez-Pajares + k M., Sanz J, Salazar D., N 54 1 1 P. sta sta sta
gAGE
The geometry and clock offsets have been removed. The geometry and clock offsets have been removed. The trend is due to the Ionosphere. The P1 code noise The trend is due to the Ionosphere. The P1 code noise shadows the cycle-slip, and without the reference (in blue), the shadows the cycle-slip, and without the reference (in blue), the time where the cycle-slip happens could not be identified. time where the cycle-slip happens could not be identified.
L1-P1 (without the cycle-slip) L1-P1 (with the cycle-slip)
1 unit = 19cm 1 unit = 19 cm (L1 cycles)
A jump of =19 cm (one cycle in L1) has been introduced in L1 at t=5000s
L1sat P sat = 2Ion1sat + ctt + ambig + sta sta 1sta
sat sat sat P sat = sta + c (dtsta dt sat ) + relsta + Tropsta + Ion1sat + K1sta + K1sat + 1sta sta sat sat sat sat L1sat = sta +2004-2005 sta dt sat ) + relsta + Tropsta Ion1sta M., Juansta Sanz J,sat + 1Ramos+ w1 + Hernndez-Pajares + k M., + k Salazar D., N JEAGAL, c (dt 55 1 1 1 P. sta
gAGE
The geometry, clock offsets and iono have been removed. The geometry, clock offsets and iono have been removed. There is a constant pattern plus noise. The P1 code noise also There is a constant pattern plus noise. The P1 code noise also shadows the cycle-slip, and without the reference (in blue), the shadows the cycle-slip, and without the reference (in blue), the time where the cycle-slip happens could not be identified. time where the cycle-slip happens could not be identified.
LC-PC (without the cycle-slip) LC-PC (with the cycle-slip)
A jump of =19 cm (one cycle in L1) has been introduced in L1 at t=5000s 1 unit = 10.7 cm (Lc cycles)
sat sat Lcsta Pcsta = ctt + ambig +
gAGE
sat sat sat sat Pcsta = sta + c (dtsta dt sat ) + relsta + Tropsta + Kcsta +
sat sat sat sat sat Lcsta = sta2004-2005 (dtsta dt sat ) + relsta + Tropsta + kcsta +M., Sanz J,+ c Rc + wc + +c Hernndez-Pajares M., Juan kc Salazar D., Ramos P. JEAGAL, 56
The geometry, clock offsets and iono have been removed. The geometry, clock offsets and iono have been removed. There is a constant pattern plus noise. The P1 code noise also There is a constant pattern plus noise. The P1 code noise also shadows the cycle-slip, and without the reference (in blue), the shadows the cycle-slip, and without the reference (in blue), the time where the cycle-slip happens could not be identified. time where the cycle-slip happens could not be identified.
LI-PI (without the cycle-slip) LI-PI (with the cycle-slip)
A jump of =19 cm (one cycle in L1) has been introduced in L1 at t=5000s 1 unit = 5.4 cm
sat sat LI sta PI sta = ctt + ambig +
gAGE
sat PI sta = IonI + K I sta + K I sat +

JEAGAL, 2004-2005
sat LI sta = IonI + kHernndez-Pajares M., 1 N1M., Sanz2J,N2 + D., Ramos + kIsat + Juan Salazar wI + P. I sta
57
The geometry ,,clock offsets and iono have been removed. The geometry clock offsets and iono have been removed. There is a constant pattern plus noise. The Pw code noise is There is a constant pattern plus noise. The Pw code noise is under one cycle of Lw..Thence, the cycle-slip is clearly detected under one cycle of Lw Thence, the cycle-slip is clearly detected
Lw-Pw (without the cycle-slip) Lw-Pw (with the cycle-slip)
A jump of =19 cm (one cycle in L1) has been introduced in L1 at t=5000s 1 unit = 86.2cm 1 unit = 86.2 cm (Lw cycles)
sat sat Lwsta Pwsta = ctt + ambig +
sat sat sat sat Pwsta = sta + c (dtsta dt sat ) + relsta + Tropsta + Ionw sat + Kw sta + Kw sat + sta sat sat sat sat sat Lwsta = JEAGAL,+ c (dtsta dt sat ) + relsta + Tropsta + Ionw M., Juan kwSanz +Salazar D.,+ wP. w + sta 2004-2005 + M., sta J, kw sat Ramos N 58 Hernndez-Pajares sta
gAGE
The geometry and clock offsets have been removed. The geometry and clock offsets have been removed. The trend is due to the Iono. The L1 code noise is few mm, and The trend is due to the Iono. The L1 code noise is few mm, and the variation of the ionosphere in 1 second is lower than 1 =19 cm the variation of the ionosphere in 1 second is lower than 1 =19 cm Thence, the cycle-slip is detected. Thence, the cycle-slip is detected.
LI (without the cycle-slip) LI (with the cycle-slip)
A jump of =19 cm (one cycle in L1) has been introduced in L1 at t=5000s 1 unit = 5.4 cm
mm wI <<
sat LI sta IonI + ctt + ambig
sat sat sat L1sat = sta + c (dtsta dt sat ) + relsta + Tropsta Ion1sat + k1sta + k1sat + 1 N1 + w1 + sta sta sat sat sat L2sat = JEAGAL, 2004-2005sta dt sat ) + relsta + TropstaHernndez-Pajares+ k2staM.,+ k2J, Salazar2 N2 + w2 + + c (dt Ion2sat M., Juan Sanz sat + D., Ramos P. 59 sta sta sta
gAGE
Summary L1
L1-P1
LI-PI
LC-PC
gAGE
JEAGAL, 2004-2005
60
The cycle-slips are detected by Combination combination Ionospheric-Free the Ionospheric (LI=L1-L2) and the Melbourne Wbbena (W=Lw-Pw)
LI
Lw-Pw
Two independent combinations, LI and Lw, allow to detect Two independent combinations, LI and Lw, allow to detect two independent cycle-slips (in L1 and L2 phase measur.). two independent cycle-slips (in L1 and L2 phase measur.).
L1
L2
gAGE
Notice that, from L1, L2 is not possible to detect short cycle-slips Notice that, from L1, L2 is not possible to detect short cycle-slips
61
Lesson 4 Satellite coordinates
gAGE
JEAGAL, 2004-2005
62
gAGE
63
gAGE
The GPS navigation message provides pseudo-Keplerian elements to compute satellite coordinates
JEAGAL, 2004-2005
64
(X, Y, Z, Vx, Vy, Vz)

(a, e, i, , , V)
6 values are needed (x,y,z,vx,vy,vz) to provide the position and velocity of a body. They can be map into the six Keplerian elements (a, e, i, , , V), which provides the natural representation of the orbit!
gAGE
JEAGAL, 2004-2005
65
orbit shape
orbit orientation
position in the orbit
a: semi major axis e: eccentricity i: inclination : argument of ascending

node : argument of perigee V: true anomaly
gAGE
JEAGAL, 2004-2005
focus
ae
66
perigee
(a, e, i, , , V)
Position in the orbit: Satellite

True anomaly (V)

Fictitious body moving at velocity n=2/P=ctt. Mean anomaly (M)
E a ae
M
Perigee
satellites perigee
T0 : time of passage by
gAGE
T0 2 n= P , a, e
V(t)
2 M (t ) = n(t T0 ) ; n = = 3 P a E(t ) = M (t ) + e sin E(t ) 1+ e E(t ) tan V (t ) = 2arctan 2 1 e

JEAGAL, 2004-2005
67
Calculation of osculatrix orbital elements from position and velocity (rv2ele_orb.f)

gAGE
JEAGAL, 2004-2005
68
Calculation of position and velocity from orbital elements (ele_orb2rv.f, orb2xyz.f)

Where:
gAGE
JEAGAL, 2004-2005
69
Due to the non-spherical nature of gravitational potential, the attraction of the sun and moon, the solar radiation pressure, etc., the true satellite path deviates from the elliptic orbit.
At any time an elliptical orbit tangent to the true path can be defined. This is the osculatrix orbit, whose Keplerian elements vary with time t:
a(t),e(t),i(t),(t),(t),V(t)
True path
gAGE
Instantaneous elliptic tangent (osculatrix) orbit.
JEAGAL, 2004-2005
70
Exercise 3: Orbital elements variation:

File 1995-10-18.eci contains the precise position and velocities of GPS satellites every 5 minutes for October 18th, 1995. [from JPL/NASA server: ftp://sideshow.jpl.nasa.gov/pub/gipsy_products] a) Use program rv2ele_orb to compute the instantaneous orbital elements for each epoch in the (a, e, i, , , V ) file. That is: (X, Y, Z, Vx, Vy, Vz) b) Plot the orbital elements in function of time to show their variation: a(t),e(t),i(t), (t),(t),V(t) Solution: a) cat 1995-10-18.eci|rv2ele_orb> orb.dat b) See the following plots
gAGE
71
gAGE
72
gAGE
73
Ephemerids in navigation message:

gAGE
In order to calculate WGS84 satellite coordinates, you should apply de following algorithm [GPS/SPS-SS, table 2-15] (see in the book FORTRAN subroutine orbit.f, annex IV)
74
RINEX ephemeris file RINEX ephemeris file

gAGE
JEAGAL, 2004-2005
75
Computation of satellite coordinates from navigation message (orbit.f)

Computation of tk time since ephemerids reference epoch toe (t and toe are given in GPS seconds of week):
tk = t toe
Computation of mean anomaly Mk for tk,
Mk = M0 +
anomaly Ek :
+ n tk a3
Iterative resolution of Keplers equation in order to compute eccentric
M k = Ek e sin Ek
1 e 2 sin E k v k = arctan cos E k e
Calculation of true anomaly vk :
gAGE
Computation of latitude argument uk from perigee argument W, true anomaly vk and corrections cuc and cus:
uk = + vk + cuc cos 2 ( + vk ) + cus sin 2 ( + vk )

76
Computation of radial distance rk, taking into consideration corrections crc and crs:
rk = a (1 2 cos Ek ) + crc cos 2 ( + vk ) + crs sin 2 ( + vk )

Calculation of orbital plane inclination ik from inclination io at reference epoch toe and corrections cic and cis :
ik = i0 + itk + cic cos 2 ( + vk ) + cis sin 2 ( + vk )

Computation of ascending node longitude k (Greenwich), from of sidereal time at Greenwich between start of week and and reference time tk=t-toe, and also corrected from change of ascending node longitude since reference epoch toe. longitude 0 at start of GPS week, corrected from apparent variation
k = 0 + ( E ) tk E toe
Calculation of coordinates in CTS system, applying three rotations (around uk, ik, k) : X r
k k
gAGE
Y k Zk
= R ( )R (i )R (u ) 0 k k 3 1 3 k 0
JEAGAL, 2004-2005
77
t
Orbit.f
Nav. message (ephemeris)
(x,y,z)[CTS]
z
Conventional Terrestrial System (CTS): Earth-Fixed System the reference system rotates with Earth.
y X
Greenwich
gAGE
JEAGAL, 2004-2005
78
The program orbit.f provides the satellite coordinates in a Earth fixed system (CTS)
X Z
Greenwich North Pole
gAGE
JEAGAL, 2004-2005
79
Exercise 4: Orbits and clocks accuracy (S/A=on)

The file eph.on contains satellite coordinates (x,y,z) and clocks, computed from the navigation message of GPS satellites for March 23th, 1999. (with S/A=on)
[the coordinates has been computed using the program orbit.f]
The file sp3.on contains precise coordinates and clocks of GPS satellites for March 23th, 1999
[Provided by the IGS server ftp://igscb.jpl.nasa.gov/igscb/product]
Plot the error of broadcast orbits and clocks and discuss the results. Solution: See the following plots
gAGE
JEAGAL, 2004-2005
80
ERROR in coordinates and clock S/A=on
gAGE
JEAGAL, 2004-2005
81
Exercise 5: Orbits and clocks accuracy (S/A=off)

The file eph.on contains satellite coordinates (x,y,z) and clocks, computed from the navigation message of GPS satellites for May 15th, 2000 (with S/A=off)
[the coordinates has been computed using the program orbit.f]
The file sp3.on contains precise coordinates and clocks of GPS satellites for May 15th, 2000
[Provided by the IGS server ftp://igscb.jpl.nasa.gov/igscb/product]
Plot the error of broadcast orbits and clocks and discuss the results. Solution: See the following plots.
gAGE
JEAGAL, 2004-2005
82
ERROR in coordinates S/A=off
gAGE
JEAGAL, 2004-2005
83
Selective Availability (S/A): Intentional degradation of satellite clocks and broadcast ephemeris. (from 25 March, 1990)
GPS Before and After S/A was switched off

Colorado Springs, Colorado 160 140 120 100 80 60 40 20 0 -20 -40 -60 -80 -100 -120 -140 -160 -180 -200 0 1 2 May 2000
Horizontal Error (meters) Vertical Error (meters)
Instantaneous Error (meters)
ANALYSIS NOTES - Data taken from Overlook PAN Monitor Station, equipped with Trimble SVeeSix Receiver - Single Frequency Civil Receiver - Four Satellite Position Solution at Surveyed Benchmark - Data presented is raw, no smoothing or editing
gAGE
10
Time of Day (Hours UTC)
JEAGAL, 2004-2005
84
gAGE
JEAGAL, 2004-2005
Lesson 5 Model
85
Pseudorange modeling (code)

C1
The pseudorange modeling is based in the GPS Standard Positioning Service Signal Specification (GPS/SPS-SS).
sat sat sat sat sat C11rec = rec + c (dtrec dt sat ) + relrec +Troprec + Ion1rec + K1rec + K1sat +
gAGE
JEAGAL, 2004-2005
86
gAGE
JEAGAL, 2004-2005
87
Geometric range
sat rec
Euclidean distance between satellite coordinates at emission time and receiver coordinates at reception time.
sat rec
(x
sat
xrec ) + ( y
2
sat
yrec ) + ( z
2
sat
zrec ) =
2
gAGE
Of course, receiver coordinates are not known (is the target of this problem). But ....
sat sat sat sat sat C11rec = rec + c (dtrec dt sat ) + relrec +Troprec + Ion1rec + K1rec + K1sat + 88

sat rec
(x
sat
xrec ) + ( y
2
sat
yrec ) + ( z
2
sat
zrec )
Of course, receiver coordinates ( xrec , yrec , zrec ) are not known (they are the target of this problem). But, we can always assume that an approximate position ( x 0 , y 0 , z 0 ) is known (it can be computed using the Bancrofts method see next lesson--):
rec rec rec
gAGE
Thence, as it will be shown in next lesson, the navigation problem will Thence, as it will be shown in next lesson, the navigation problem will consist on: consist on: 1.- To start from an approximate value for receiver position 1.- To start from an approximate value for receiver position ( x 0 , y 0 , z 0 ) (it can be computed with Bancrofts method) (it can be computed with Bancrofts method) 2.- With the pseudorange measurements and the navigation 2.- With the pseudorange measurements and the navigation equations, compute the correction ( dxrec , dyrec , dzrec ) to have a equations, compute the correction to have a more precise position of the receiver. more precise position of the receiver.
rec rec rec
( xrec , yrec , zrec ) = ( x0
rec
, y0rec , z0rec + ( dxrec , dyrec , dzrec )
JEAGAL, 2004-2005
89
Satellite coordinates at emission time (rec2ems.f)

The GPS signal travels from satellite coordinates at emission time (tems) to receiver coordinates at reception time (trec). The satellite can move several hundreds of meters from tems to trec. The receiver time-tags are given at reception time and in the receiver clock time.
gAGE
An algorithm is needed to compute the satellite An algorithm is needed to compute the satellite coordinates at emission time in the GPS system time coordinates at emission time in the GPS system time from reception time in the receiver time M., Sanz J, Salazar D., Ramos P. from reception time in the receiver time tags. Hernndez-Pajares M., Juan tags. JEAGAL, 2004-2005 90
gAGE research group of Astronomy and Geomatics Emission time University of Catalonia, Barcelona, Spain T[ems] in the GPS system time Technical
The satellite offset The satellite offset clock dtSS can be clock dt can be C1 computed from the computed from the navigation message navigation message
C1= c t= c [trec(TR)-tems(TS)]
As it is known, the pseudorange measurements link the emission time (tems) in satellite clock (TS) with reception time (trec) in receiver clock (TR) (receiver time tags). Thence, the emission time in the satellite clock is:
tems(TS)= trec(TR)-C1/c
Finally, since dtS= tS T is the time offset between satellite clock (tS) and GPS system time (T), thence:
JEAGAL, 2004-2005
T[ems]= tems(TS)-dtS= trec(TR)-(C1/c+dtS) 91

Satellite coordinates computation at emission time

The algorithm provided by the GPS/SPS-SS (orbit.f) supplies satellite coordinates in an Earth-Fixed reference frame. To compute the satellite coordinates At the emission time, the following algorithm can be applied: 1. From receiver time-tags, compute emission time in GPS system time: S
See rec2ems.f See rec2ems.f
T[ems]= trec(TR)-(C1/c+dt )
2. Compute satellite coordinates at emission time T[ems]
T[ems]
[orbit]
(Xsat,Ysat,Zsat)CTS[emission]
gAGE
3. Account for Earth rotation during traveling time from emission to reception t (CTS reference system at reception time is used to
build the navigation equations).
(Xsat,Ysat,Zsat)CTS[reception] =R3(E t).(Xsat,Ysat,Zsat)CTS[emission]

92
Exercise 6: Using the GCAT program, compute satellite

coordinates at emission time and at reception time. Plot the module of the vector difference between both positions (use October 13th, 1998 data files).
gAGE
JEAGAL, 2004-2005
93
gAGE
94
Satellite and receiver clock offsets

system time (provided by the ground control segment):
They are time-offsets between satellite/receiver time and GPS

-The receiver clock offset (dtrec) is estimated together with receiver coordinates. - Satellite clock offset (dtsat) may be computed from navigation message:
dtsat=a0 + a1(t-t0) + a2(t-t0)2
gAGE
dtsat=a0 + a1(t-t0) + a2(t-t0)2

YY MM DD H M S
t0
a0
a1
a2
PRN
22 NAVIGATION DATA GPS RINEX VERSION // TYPE NAVIGATION DATA GPS RINEX VERSION TYPE srx/v1.8.1.4 BAI 95/10/19 03:18:35 PGM // RUN BY // DATE srx/v1.8.1.4 BAI 95/10/19 03:18:35 PGM RUN BY DATE CASA COMMENT CASA COMMENT -2444431.2031 -4428688.6270 3875750.1442 COMMENT -2444431.2031 -4428688.6270 3875750.1442 COMMENT END OF HEADER END OF HEADER 14 95 10 18 00 51 44.0 1.129414886236D-05 1.136868377216D-13 0.000000000000D+00 14 95 10 18 00 51 44.0 1.129414886236D-05 1.136868377216D-13 0.000000000000D+00 1.730000000000D+02-5.175000000000D+01 4.375182243902D-09-5.836427291652D-01 1.730000000000D+02-5.175000000000D+01 4.375182243902D-09-5.836427291652D-01 -2.712011337280D-06 2.427505562082D-03 8.568167686462D-06 5.153718931198D+03 -2.712011337280D-06 2.427505562082D-03 8.568167686462D-06 5.153718931198D+03 2.623040000000D+05 4.470348358154D-08 1.698435481558D+00 1.676380634308D-08 2.623040000000D+05 4.470348358154D-08 1.698435481558D+00 1.676380634308D-08 9.636381916043D-01 2.153437500000D+02 3.056960010495D+00-8.030691653399D-09 9.636381916043D-01 2.153437500000D+02 3.056960010495D+00-8.030691653399D-09 -5.178787145843D-11 1.000000000000D+00 8.230000000000D+02 0.000000000000D+00 -5.178787145843D-11 1.000000000000D+00 8.230000000000D+02 0.000000000000D+00 3.200000000000D+01 0.000000000000D+00 1.396983861923D-09 1.730000000000D+02 3.200000000000D+01 0.000000000000D+00 1.396983861923D-09 1.730000000000D+02 2.592180000000D+05 0.000000000000D+00 0.000000000000D+00 0.000000000000D+00 2.592180000000D+05 0.000000000000D+00 0.000000000000D+00 0.000000000000D+00
gAGE
JEAGAL, 2004-2005
96
gAGE
97
Relativistic correction (relij)

A constant component depending only on nominal value of satellites orbit major semi-axis, being corrected modifying satellites clock oscillator frequency*:
f 0' f 0 1 v U = + 2 f0 2c c
2
4.464 1010
A periodic component due to orbit eccentricity (to be corrected by user receiver):
rel = 2
a
c
e sin( E ) = 2
rv (meters ) c
Being =3.986005 1014 (m3/s2) universal gravity constant, c =299792458 (m/s) light speed in vacuum, a is orbits major semi-axis, e is its eccentricity, E is satellites eccentric anomaly, and r and v are satellites geocentric position and speed in an inertial system. *being f0 = 10.23 MHz, we have f=4.464 10-10 f0= 4.57 10-3 Hz so satellite should use fo=10.22999999543 MHz.
gAGE
98
gAGE
99
Ionospheric Delay
I j i
ion =
40.3 I 2 f
As a first approach, ionospheric delay depends on frequency as given by:
Where I is number of electrons per area unit in the direction of observation, or STEC (Slant Total Electron Content) I= N
ds
For two-frequency receivers, it may be cancelled (99.9%) using ionosphere-free combination f12 L1 f 22 L 2
LC = f12 f 22
For one-frequency receivers, it may be corrected (about 60%) using Klobubhar model (defined in GPS/SPS-SS), whose parameters are sent in navigation message. (See program klob.f)
gAGE
Klobuchar model (klob.f)

Vertical delay
It was designed to minimize user computational complexity. Minimum user computer storage Minimum number of coefficients transmitted on satellite-user link At least 50% overall RMS ionospheric error reduction worldwide. It is assumed that the electron content is concentrated in a thin layer at 350 Km in height. The slant delay is computed from the vertical delay at the ionospheric Pierce Point (IPP), multiplying by the obliquity factor.
IPP Slant delay
Obliquity factor
gAGE
JEAGAL, 2004-2005
101
IONOSPHERIC PIERCE POINTS (IPP)
IPPs trajectories for a receiver in Barcelona, Spain
IPP Slant Delay Vertical Delay

Ionospheric Layer (350 Km in height)
gAGE
JEAGAL, 2004-2005
102
Klobuchar model
30 Time Delay (ns at 1.6 GHz) 25 20 Amplitude 15 10 5 Dc=5ns 4 8
Klobuchar coefficients
IonVERT
2 (t ) DC + A cos (day) P = DC ; if 2 (t ) > (night ) 2 P

Being :
*Period 12 16 18 24
A=
n=0
P=
n=0
= Geomagnetic Latitude
Local Time (hours)
IonSLANT = IonVERT m(elev) m(elev) = 1+16( 0.53 elev / )

JEAGAL, 2004-2005
Where: DC= 5ns = 14 (ctt. phase offset) t = Local Time

103
gAGE
sat,0,1,2,3,0,1,2,3) (time, rrsta, rrsat,0,1,2,3,0,1,2,3) (time, sta,
[Klob] [Klob]
Iono Iono
elev,
2 NAVIGATION DATA RINEX VERSION / TYPE CCRINEXN V1.5.2 UX CDDIS 24-MAR- 0 00:23 PGM / RUN BY / DATE IGS BROADCAST EPHEMERIS FILE COMMENT 0.3167D-07 0.4051D-07 -0.2347D-06 0.1732D-06 ION ALPHA -0.2842D+05 -0.2150D+05 -0.1096D+06 0.4301D+06 ION BETA -0.121071934700D-07-0.488498130835D-13 319488 1002 DELTA-UTC: A0,A1,T,W 13 LEAP SECONDS END OF HEADER 1 99 3 23 0 0 0.0 0.783577561379D-04 0.113686837722D-11 0.000000000000D+00 0.191000000000D+03-0.106250000000D+01 0.487163149444D-08-0.123716752769D+01 -0.540167093277D-07 0.476544268895D-02 0.713579356670D-05 0.515433833885D+04 0.172800000000D+06-0.260770320892D-07-0.850753478531D+00 0.763684511185D-07 0.957259887797D+00 0.241437500000D+03-0.167990552187D+01-0.823998608564D-08 0.174650132022D-09 0.100000000000D+01 0.100200000000D+04 0.000000000000D+00 0.320000000000D+02 0.000000000000D+00 0.465661287308D-09 0.191000000000D+03 0.172800000000D+06 0.000000000000D+00 0.000000000000D+00 0.000000000000D+00
gAGE
JEAGAL, 2004-2005
104
gAGE
105
Tropospheric Delay
It does not depend on frequency and affects both the code and carrier phases in the same way It can be modeled (about 90%) by:
sat Troprec = (ddry + dwet ) m(elev)
m(elev) =
1.001
0.002001+ sin2 (elev)
ddry corresponds to the vertical delay of the dry atmosphere (basically oxygen and nitrogen in hydrostatical equilibrium) It can be modeled as an ideal gas. dwet corresponds to the vertical delay of the wet component (water vapor) difficult to model. A simple model is:
d dry = 2.3 exp( 0.116 10 3 H ) d wet = 0.1m meters ; [ H : heigh ] over the sea level
gAGE
A more accurate model for ddry and dwet is provided for SBAS receivers in RTCA-Do229C. This model depends on the latitude and the day-of-year, being interpolated over a table of several meteorological parameters. More sophisticated models uses two different mappings (for wet and dry)
gAGE
107
Instrumental Delays
Some sources for these delays are antennas, cables, as well as several filters used in both satellites and receivers. They are composed by a delay corresponding to satellite and other to receiver, depending on frequency:
sat K1rec = R1rec TGD sat sat K 2rec = R 2rec
f12 2 TGD sat f2
R1rec may be assumed as zero (including it in receiver clock offset). TGDsat is transmitted in satellites navigation message (Total Group Delay).
According to ICD GPS-2000, control segment monitors satellite timing, so TGD cancels out when using free-ionosphere combination. That is why we have that particular equation for L2
gAGE
TGD
108
gAGE
109
Measurement noise (thermal noise)

Antispoofing (A/S): Antispoofing (A/S): The code P is encrypted to Y. The code P is encrypted to Y. signal Only theGPSC at Only thecode C at code
frequency L1 is available. frequency L1 is available. C1 P1 (Y1): encrypted P2 (Y2): encrypted L1 L2
Wavelength (chip-rate)
noise (1% of ) [*]
Main characteristics
Code measurements
300 m 30 m 30 m 19.05 cm 24.45 cm 3m 30 cm 30 cm 2 mm 2 mm Unambiguous but noisier
Phase measurements Precise but ambiguous
gAGE
[*] codes may be smoothed with the phases in order to reduce noise (i.e., C1 smoothed with L1 50 cm noise)
sat sat sat sat sat C11rec = rec + c (dtrec dt sat ) + relrec +Troprec + Ion1rec + K1rec + K1sat 110 +
Multipath
One or more reflected signals reach the antenna in addition to the direct signal. Reflective objects can be earth surface (ground and water), buildings, trees, hills, etc. It affects both code and carrier phase measurements, and it is more important at low elevation angles.
Butterfly shape
gAGE
Code: up to 1.5 chip-length Phase: up to /4
up to 450m for C1 [theoretically] Typically: less than 2-3 cm.
up to 5 cm for L1 and L2 [theoretically] Typically: less than 1 cm

JEAGAL, 2004-2005
111
gAGE
Exercise 7: Plot code and phase ionospheric combination for satellite PRN 15 of file 97jan09coco___r0.rnx and discuss the results.
Butterfly shape: High multipath for low elevation rays (when satellite rises and sets)
JEAGAL, 2004-2005
Note: A/S=on
112
Exercise 8: Files gage2710.98.a, 2720.98.b and gage2730.98.a contain 1-second measurements collected by a static receiver in three consecutive days. Plot the combination P1-L1 and identify the multipath (note: shift the plots 3m56s= 236 sec each day)
The trend is due to the ionosphere
For a static receiver: For a static receiver: Each sidereal day is Each sidereal day is m s (24hh-3m56s),the (24 -3 56 ), the geometry repeats geometry repeats the multipath repeats the multipath repeats
JEAGAL, 2004-2005
gAGE
L1sat P sat = 2Ion1sat + ctt + ambig + [Multipath + noise] 113 sta sta 1sta
Exercise 9: Computation of modeled pseudorange

Using data of files 13oct98.rnx and 13oct98.eph, compute by hand the modeled pseudorange for satellite PRN 14 at t=38230 sec (10h37m10s).
sat sat sat sat sat PR[mod]1rec = 0,rec cdt sat + relrec +Troprec + Ion1rec + K1sat
Follow these steps:
gAGE
JEAGAL, 2004-2005
114
1. Select orbital elements closer to 38230 2. Compute satellite clock offset 3. Compute satellite-receiver aprox. geometric range
3.1 Compute emission time from receiver (reception) time-tags and code pseudorange. 3.2 Compute satellite coordinates at emission time 3.3 Compute approximate geometric range.
gAGE
4. 5. 6. 7. 8.
Compute Compute Compute Compute Compute
satellite Instrumental delay (TGD): relativistic correction tropospheric delay ionospheric delay modeled pseudorange from previous values:
115
1. Selection of orbital elements: From file 13oct98.eph,

select the last transmitted navigation message block before instant t=38230 s (10h37m10s).
Transmission time: 979 208818 10h 0m 18s
PRN
14 98 10 13 12 0 0 +1.28000000000E+02 -3.31364572048E-06 +2.16000000000E+05 +9.73658001335E-01 -1.45720352451E-10 GPS sec of week +3.20000000000E+01 +2.08818000000E+05 +5.65452501178E-06 -6.10000000000E+01 +1.09227513894E-03 -6.33299350738E-08 +2.74031250000E+02 +1.00000000000E+00 +0.00000000000E+00 +0.00000000000E+00 +9.09494701773E-13 +4.38125402624E-09 +5.67547976971E-06 +1.00409621952E+00 GPS week +2.66122811383E+00 +9.79000000000E+02 -2.32830643654E-09 +0.00000000000E+00 +0.000000000000E+00 +8.198042513605E-01 +5.153795101166E+03 -3.725290298462E-09 -8.081050495434E-09 +0.000000000000E+00 +1.280000000000E+02 +0.000000000000E+00
gAGE
JEAGAL, 2004-2005
116
2. Satellite clock offset computation: From file PRN

13oct98.eph, compute satellite clock offset at time t=3830 s for PRN14:
t0
a0
+5.65452501178E-06 -6.10000000000E+01 +1.09227513894E-03 -6.33299350738E-08 +2.74031250000E+02 +1.00000000000E+00 +0.00000000000E+00 +0.00000000000E+00
a1
+9.09494701773E-13 +4.38125402624E-09 +5.67547976971E-06 +1.00409621952E+00 +2.66122811383E+00 +9.79000000000E+02 -2.32830643654E-09 +0.00000000000E+00
a2
+0.000000000000E+00 +8.198042513605E-01 +5.153795101166E+03 -3.725290298462E-09 -8.081050495434E-09 +0.000000000000E+00 +1.280000000000E+02 +0.000000000000E+00
14 98 10 13 12 0 0 +1.28000000000E+02 -3.31364572048E-06 +2.16000000000E+05 +9.73658001335E-01 -1.45720352451E-10 +3.20000000000E+01 +2.08818000000E+05
t = 38230 sec t0= 12h 0m 0s= 43200 s
gAGE
dtsat= a0 + a1(t-t0) + a2(t-t0)2 = 5.65 10-6 s

JEAGAL, 2004-2005
117
3. Satellite-receiver geometric range computation:

Use the following values (4789031, 176612, 4195008) as approximate coordinates.
3.1: Emission time computation from receiver time-tag and code pseudorange:
T[ems]= trec(TR)-(C1/c + dtsat)
Measurement file 13oct98.a Ephemeris file 13oct98.eph
Pseudorange C1 at receiver time-tag t=38230: C1=23585247.703 m Satellite clock offset at t=38230 sec dtsat= 5.65 10-6 sec (see previous
results)
gAGE
Thence, the emission time in GPS satellite clock is:
T[ems]= 38230 (23585247.703/c + 5.65 10-6) =

= 38229.9213224
JEAGAL, 2004-2005
(where c=299792458)
118
3.2: Satellite coordinates at temission time pseudorange:

T[emission]=
38229.921 sec
Orbit.f
X= 11453479.346 Y= 122468524.004 Z= 8245076.145
CTS [emission]
Use the selected ephemeris for PRN14 (from file 13oct98.eph)
gAGE
The previous coordinates are given in an Earth-fixed reference frame (CTS) at t=T[emission]= 38229.921 s. This reference frame rotates by un amount E t during traveling time t=T[reception]-T[emission].

119

11453350.377 cos( ) sin( ) 0 11453479.346 122468589.797 = sin( ) cos( ) 0 122468524.004 8245076.145 0 0 1 8245076.145
CTS [ reception ] CTS [ emission ]
= E t = 5.74 106 rad . ( where E = 7.2921151467 105 rad / sec)

t =
sat rec = sat rec
= 0.079sec
2 2 2
( x sat xrec ) + ( y sat yrec ) + ( z sat zrec )
23616673.3m
( x, y, z ) satellite (11453479, 22468524, 8245076) ( x, y, z ) receiver (4789031, 176612, 4195008)
gAGE
An approximate value is enough to compute t.
Note: Both satellite and receiver coordinates must be given in the Note: Both satellite and receiver coordinates must be given in the same reference system! same reference system! the CTS[reception] will be usedto build navigation equations. P. JEAGAL,theCTS[reception] will be used Hernndez-Pajares M., Juan M., Sanz J, equations. 2004-2005 to build navigation Salazar D., Ramos
120
3.2: Geometric range computation

The geometric range between satellite coordinates at emission time and the approximate position of the receiver at reception time (both coordinates given in
the same reference system [for instance the CTS system at reception time]) is computed by:
satellite 0, receiver
(x
sat
x0,rec ) + ( y
2
sat
y0,rec ) + ( z
2
sat
z0,rec ) = 23616699.124m
2
( x, y, z ) satellite = (11453350.2771, 22468589.7975, 8245076.1448)CTS [ reception ] ( x0 , y0 , z0 ) receiver = (4789031, 176612, 4195008)CTS [ reception ]
Approximate receiver coordinates at reception time.
gAGE
JEAGAL, 2004-2005
121
4. Satellite Instrumental delay (TGD): From file

13oct98.eph, compute the Total Group Delay for PRN14:
PRN
14 98 10 13 12 0 0 +1.28000000000E+02 -3.31364572048E-06 +2.16000000000E+05 +9.73658001335E-01 -1.45720352451E-10 +3.20000000000E+01 +2.08818000000E+05 +5.65452501178E-06 -6.10000000000E+01 +1.09227513894E-03 -6.33299350738E-08 +2.74031250000E+02 +1.00000000000E+00 +0.00000000000E+00 +0.00000000000E+00
TGD (in sec)

+9.09494701773E-13 +4.38125402624E-09 +5.67547976971E-06 +1.00409621952E+00 +2.66122811383E+00 +9.79000000000E+02 -2.32830643654E-09 +0.00000000000E+00 +0.000000000000E+00 +8.198042513605E-01 +5.153795101166E+03 -3.725290298462E-09 -8.081050495434E-09 +0.000000000000E+00 +1.280000000000E+02 +0.000000000000E+00
gAGE
TGD= -2.32830643654E-09 * c= -0.69801 m
JEAGAL, 2004-2005
122
5. Relativistic correction: PRN

research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain 14 98 10 13 12 0 0 +1.28000000000E+02 -3.31364572048E-06 +2.16000000000E+05 +9.73658001335E-01 -1.45720352451E-10 +3.20000000000E+01 +2.08818000000E+05 +5.65452501178E-06 -6.10000000000E+01 +1.09227513894E-03 -6.33299350738E-08 +2.74031250000E+02 +1.00000000000E+00 +0.00000000000E+00 +0.00000000000E+00
sqrt(a)
+0.000000000000E+00 +8.198042513605E-01 +5.153795101166E+03 -3.725290298462E-09 -8.081050495434E-09 +0.000000000000E+00 +1.280000000000E+02 +0.000000000000E+00
+9.09494701773E-13 +4.38125402624E-09 +5.67547976971E-06 +1.00409621952E+00 +2.66122811383E+00 +9.79000000000E+02 -2.32830643654E-09 +0.00000000000E+00
T[emission] =
38229.921 s
Orbit.f
(eccentric anomaly)
E =0.095 rad.
gAGE
rel
satellite receiver
=2
a
c
e sin(E) = 0.07m
= 3.9860051014 m3s2
c = 299792458 ms1
JEAGAL, 2004-2005
123
6. Tropospheric correction
Trop = ( ddry + dwet ) m(elev) = 6.76m

sat rec
ddry = 2.3e m(elev) =
0.116103 H
= 2.3m
dwet = 0.1m 1.001 0.002001+ sin (elev)

2
See klob.f
elev = 20.57
180 H =160m (heigh over the ellipsoid)
= 0.359rad
gAGE
(x,y,z)sta (x,y,z)sta
[car2geo] [car2geo]
(Lon, Lat, H)sta (Lon, Lat, H)sta
JEAGAL, 2004-2005
124
7. Ionospheric correction
sat,0,1,2,3,0,1,2,3) (time, rr ,,rrsat,0,1,2,3,0,1,2,3) (time, sta sta
[Klob] [Klob]
Iono=10.26m Iono=10.26m
2 NAVIGATION DATA GPS RINEX VERSION/ TYPE XPRINT v1.1 gAGE 00/06/04 17:36:23 PGM / RUN BY / DATE gAGE BROADCAST EPHEMERIS FILE COMMENT +1.9558E-08 +0.0000E+00 -1.1921E-07 +0.0000E+00 ION ALPHA +1.2288E+05 -1.6384E+04 -2.6214E+05 +1.9661E+05 ION BETA -8.381903171539E-09-1.421085471520E-14 405504 979 DELTA_UTC: A0,A1,T,W 12 LEAP SECONDS END OF HEADER
t = 38230sec ( x, y, z ) satellite = (11453350.2771, 22468589.7975, 8245076.1448)CTS [ reception ] ( x0 , y0 , z0 ) receiver = (4789031, 176612, 4195008)CTS [ reception ]
Approximate values for receiver or satellite coordinates are enough
gAGE
125
7. Compute the modeled pseudorange.

sat sat sat sat sat PR[mod]1rec = 0,rec cDt sat + relrec +Troprec + Ion1rec + K1sat
sat 0,rec = 23616699.124m
cdt sat = 5.6510-6 c =1693.828m rel = 0.071m

sat rec sat Troprec = 6.760m sat Ion1rec =10.260m
sat PR[mod]1rec = 23615021.689m
gAGE
K1sat = 0.698m
JEAGAL, 2004-2005
126
Prefit residual:
Is the difference between measured and modeled pseudorange

sat sat sat sat sat C11rec = rec + c (Dtrec Dt sat ) + relrec + Troprec + Ion1rec + K1rec + K1sat +
sat sat sat sat sat PR[mod]1rec = 0,rec cDt sat + relrec + Troprec + Ion1rec + K1sat
Pref = C1 PR1[mod] =
sat rec sat rec sat rec sat rec
sat 0,rec
+ c Dtrec + K1rec +
In the previous example (PRN14 at tt = 38230 s): In the previous example (PRN14 at = 38230 s): Pref= 23585247.703 -- 23615021.689 = -29773.986 m Pref= 23585247.703 23615021.689 = -29773.986 m
gAGE
127
Lesson 6
Navigation Equations
gAGE
JEAGAL, 2004-2005
128
Solving navigation equations

Input: - Pseudoranges (receiver-satellite j): Pj - Navigation message. In particular: satellite position when transmitting signal: rj = (xj, yj, zj) offsets of satellite clocks: dtj (j = 1,2,n) (n>=4) Unknowns: - receiver position: r = (x, y, z) - receiver clock offset: DT
gAGE
129
For each satellite in view

rel.+ Iono+Tropo+TGD
C 1ij = i j + c ( D t i D t j ) +
j = io +
j io
+
dzi + c ( Dti Dt j ) + k
xio x j
j io
dxi +
yio y j
j io
dyi +
zio z j
where:
xi = xio + dxi
Prefit-residuals (Prefit)
j C1ij io + cDt j k =
; yi = yio + dyi
; z i = z io + dz i
xio x j
gAGE
j io
dxi +
yio y j
j io
dyi +
zio z j
j io
dzi + cDti
measurement
JEAGAL, 2004-2005
computed
unknown
130
For all sat. in view

Prefit Prefit Prefit 22 Prefit ........ ........ Prefit nn Prefit

11
xio x1 1 io x x2 io 2 = io xio x n n io
yio y1
1 io
zio z1
1 io
yio y
2 io
zio z
2 io
.......... yio y n
n io
zio z n
n io
1 1 1
dx xii dy y ii dz zii cDT cdTii
Observations
(measured/computed)
Geometry of rays
gAGE
rn 0,i
in0
rn 0 ,i
in0
Line of sight vector from receiver to satellite

131
JEAGAL, 2004-2005
COMMENTS:
P r e fit 1 2 P r e fit ........ P r e fit n
x io x 1 i1o x x2 io 2 = io x io x n ino
y io y 1
z io z 1
1 io
1 io
y io y 2
i2o
z io z 2
i2o
.......... y io y n
ino
z io z n
ino
1 1 1
dxx i i dy yi i dzz i i c d Ti cDTi
Of course, receiver coordinates ( xrec , yrec , zrec ) are not known (they are the target of this problem). But we can always assume that an approximate position ( x 0 , y 0 , z 0 ) is known:
rec rec rec
gAGE
That is: That is: The navigation problem will consist on: The navigation problem will consist on: 1.- To start from an approximate value for receiver position 1.- To start from an approximate value for receiver position ( x 0 , y 0 , z 0 ) (it can be computed with Bancrofts method) (it can be computed with Bancrofts method) 2.- With the pseudorange measurements and the navigation 2.- With the pseudorange measurements and the navigation equations, to compute the correction ( dxrec , dyrec , dzrec ) to have a equations, to compute the correction to have a more precise position of receiver. more precise position of receiver.
rec rec rec
( xrec , yrec , zrec ) = ( x0
rec
, y0rec , z0rec + ( dxrec , dyrec , dzrec )
JEAGAL, 2004-2005
132
Note: An (approximate) initial receiver position (x0,y0,z0) is assumed to be known in former equations!! The method of Bancroft (see next lesson) allows to compute approximate receiver coordinates from pseudorange measurements and satellite coordinates.
T= tt0: Compute an approximate initial position using T= 0: Compute an approximate initial position using Bancroft algorithm Bancroft algorithm rr0(to)= (x00,y0,z0) 0(to)= (x ,y0,z0)
the coordinates computed in the previous epoch. the coordinates computed in the previous epoch.
T= ttn: take as an approximate position rr0(tn)= r(tn-1)), T= n: take as an approximate position 0(tn)= r(tn-1 ,
gAGE
Bancroft Bancroft to initialize to initialize

JEAGAL, 2004-2005
Linearized model:
from tn-1 to tn
rr0(to) 0(to)
r0(tn)=r(tn-1)
r(tn)
133
Solving the equations
gAGE
JEAGAL, 2004-2005
134
Let be the basic linearized GPS measurement equation:

Y = AX
Least Squares solution:
= ( A T A ) 1 A Y X
m in Y Y Y = A X
= m in i
( yi
2 yi )
The same error is assumed in all measurements
Weighted Least Squares solution

If the measurements have different errors, the equations can be weighted by matrix W : And the weighted least squares solution is:
w y1 W = 0
2 W
0 w yn
gAGE
Uncorrelated errors are assumed
X = (A W A ) A W Y
T 1
JEAGAL, 2004-2005
min Y Y Y = AX
2 = min wi ( y i y i ) i
135
Assuming that measurements Y have random errors with zero mean and variance 2, and assuming that error sources for each satellite are uncorrelated with error sources for any other satellite, the following weighted matrix may be used:
2 1/ y1 0 O W = 2 0 1/ yn
wi =
2 yi
2 y
i
wi
less weight
greater error
Minimum Variance Solution: the error covariance matrix for measurements Y. 1 W=Py If the weighting matrix is taken as W = PY-1 , thence the
Minimum Variance Solution is found: Let be PY
gAGE
= ( A T P 1 A ) 1 A P 1 Y X Y Y
And the error covariance matrix for the estimation X is:
PX = ( A P
T
1 Y
A)
JEAGAL, 2004-2005
136
Exercise 10:
With program GCAT and data files 13oct98.a, 13oct98.eph, and taking (x0,y0,z0)=[4789031 176612 4195008 ]: a) Generate data file 13oct98.a.dmx with the prefit-residuals (Y) and the design matrix (A) for t = 38230 s. b) Use these values to compute the navigation solution at that instant X = inv(A*W*A)*A*Y [note X=(dx,dy,dz)] c) Compute receiver position:
(x,y,z)= (x0,y0,z0)+ (dx,dy,dz)
gAGE
JEAGAL, 2004-2005
137
Resolution: a)
research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain ------------------------------- 13oct98.a.dmx ---------------------------x_sat-x0 y_sat-y0 z_sat-z0 rec sec PRN Prefit-res ---------------------- sig2_y rho0 rho0 rho0 ------------------------------------------------------------------------gAGE 38230.000 27 C -29790.69346 -0.825386165 0.083009922 0.558432656 1 gAGE 38230.000 14 C -29773.88675 -0.282186738 -0.943907431 -0.171491711 1 gAGE 38230.000 16 C -29692.55363 -0.524933459 -0.498000630 -0.690246504 1 gAGE 38230.000 19 C -29730.62312 -0.958194420 -0.270768810 0.092453794 1
t=38230 t=38230
Y=[-29790.69346] Y=[-29790.69346] [-29773.88675] [-29773.88675] [-29692.55363] [-29692.55363] [-29730.62312] [-29730.62312] A=[-0.825386165 0.083009922 0.558432656 1] A=[-0.825386165 0.083009922 0.558432656 1] [-0.282186738 -0.943907431 -0.171491711 1] [-0.282186738 -0.943907431 -0.171491711 1] [-0.524933459 -0.498000630 -0.690246504 1] [-0.524933459 -0.498000630 -0.690246504 1] [-0.958194420 -0.270768810 0.092453794 1] [-0.958194420 -0.270768810 0.092453794 1] Py=[1 0 0 0] Py=[1 0 0 0] [0 1 0 0] [0 1 0 0] [0 0 1 0] [0 0 1 0] [0 0 0 1] [0 0 0 1]
gAGE
138
Resolution: b)
Y=[-29790.69346] Y=[-29790.69346] [-29773.88675] [-29773.88675] [-29692.55363] [-29692.55363] [-29730.62312] [-29730.62312]
A=[-0.825386165 0.083009922 0.558432656 1] A=[-0.825386165 0.083009922 0.558432656 1] [-0.282186738 -0.943907431 -0.171491711 1] [-0.282186738 -0.943907431 -0.171491711 1] [-0.524933459 -0.498000630 -0.690246504 1] [-0.524933459 -0.498000630 -0.690246504 1] [-0.958194420 -0.270768810 0.092453794 1] [-0.958194420 -0.270768810 0.092453794 1] Py=[1 0 0 0] Py=[1 0 0 0] [0 1 0 0] [0 1 0 0] [0 0 1 0] [0 0 1 0] [0 0 0 1] [0 0 0 1]
Note data set collected under S/A=on dx dx dy dy dz dz clock clock
[ -99.09] [ -99.09] x=inv(A*W*A)*A*Y= [ 6.02] x=inv(A*W*A)*A*Y= [ 6.02] where W=inv(Py) [ -105.24] where W=inv(Py) [ -105.24] [-29814.21] [-29814.21]
gAGE
Resolution: c)
x= xo + dx = [[4789031] + [[ -99.09] = [[ 4788931.91] x= xo + dx = 4789031] + -99.09] = 4788931.91] y = y0 + dy = [[ 176612 ]] + [[ 6.02 ]] = [[ 176618.02] y = y0 + dy = 176612 + 6.02 = 176618.02] zz= z0 + dz = [[4195008 ]] + [[-105.24] J,=[[4194902.76 139 Hernndez-Pajares -105.24] = z0 + dz = 4195008 + M., Juan M., Sanz =Salazar D., Ramos P. 4194902.76]]
JEAGAL, 2004-2005
The program Gnav.f

(source code provided with the book)
gAGE
The software code of Gnav.f emulates The software code of Gnav.f emulates a GPS receiver. a GPS receiver. It computes the navigation solution from It computes the navigation solution from RINEX measurement and ephemeris files. RINEX measurement and ephemeris files. It applies the Bancroft method for cold It applies the Bancroft method for cold start and the linearized equations for start and the linearized equations for navigating navigating
JEAGAL, 2004-2005
140
Kalman filtering:
It is based on computing the weighted average between: It is based on computing the weighted average between: Y (n) the measurement Y(n) (i.e., at tt= tt )) the measurement Y(n) (i.e., at = nn X ( n from the estimation X ( n 1) the prediction of the state X(n), ) from the estimation X(n-1) the prediction of the state X(n), X(n-1)
1. Weighted average:
Y (n) = A(n) X (n) X (n) = X (n)

Y ( n ) A ( n ) = X (n) X (n) I
Lets assume, that we have the prediction X ( n ) , with P X ( n ) thence, it can be used to add an additional set of equations to the measurement equation Y= A X
gAGE
JEAGAL, 2004-2005
0 PY ( n ) W = 0 PX ( n ) Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.
141
Y ( n ) A ( n ) = X (n) X (n) I
PY ( n ) W = 0
PX ( n ) 0
And the following solution of the previous equation system can be found with some elemental algebraic manipulations:
1 X ( n ) = PX ( n ) A t ( n ) PY( 1 ) Y ( n ) + PX ( n ) X ( n ) n
gAGE
PX ( n ) = A ( n ) PY ( n ) A ( n ) + P
t 1
1 X (n)
JEAGAL, 2004-2005
142
2.- Prediction
Scalar case: Scalar case:

Lets Lets
x n 1
be the state at tt= n with variance be the state at = n with variance
2 x n 1
The simplest prediction model is to assume that the prediction The simplest prediction model is to assume that the prediction at tt= n is proportional to the state at tt= n-1..That is: at = n is proportional to the state at = n-1 That is:
x n = x n 1
Thence, existing a linear relation between x n 1 Thence, existing a linear relation between
variance of the prediction will be: variance of the prediction will be:
n
x2 = 2 x2
the x n ,,the and and

+ q2
n 1
gAGE
An additional term is added to account for modeling error!

143
Generalization to the vector case: Generalization to the vector case:

x n = x n 1
xn
2 xn
=
X (n) (n) PX ( n ) Q (n)
2 x n 1
+ q2
x2
q
(n) : Q (n) :
transition m atrix process noise m atrix
gAGE
X ( n ) = ( n 1) X ( n 1) PX ( n ) = ( n 1) PX ( n 1) T ( n 1) + Q ( n 1)
144
Kalman filter
(see kalman.f)
Measurements
X (n) PX ( n )
Prediction Estimation
Y (n) PY ( n )
X (n) = X (n 1) PX (n) = PX (n1) t + Q

X (0) PX ( 0 )
Initialization
1 X (n) = PX ( n ) At PY( 1) Y (n) + PX ( n ) X (n) n
PX ( n ) = A PY ( n ) A + P
t 1
1 X (n)
gAGE
X (n) PX ( n )
JEAGAL, 2004-2005
145
Kalman filter (classical version)

Measurements research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain
X (n) PX ( n )
Estimation Prediction
Y (n) PY ( n )
X (n) = X (n 1) PX (n) = PX (n1) + Q

t
X (n) = X (n) + K(n) Y(n) A(n) X (n) K(n) = PX (n) A (n) A(n) PX (n) A (n) P (n) Y PX (n) = [ I K(n) A(n)] PX (n)
t t 1
X (0) PX ( 0 )
Initialization
gAGE
X (n) PX ( n )
JEAGAL, 2004-2005
146
Some simple examples to define matrices and Q

X ( n ) = ( n 1) X ( n 1) PX ( n ) = ( n 1) PX ( n 1) T ( n 1) + Q ( n 1)
a) Static positioning:
State vector to be determined is X = (xrec , yrec , zrec , dtrec) where coordinates (xrec , yrec , zrec) are considered constant (because receiver is fixed) and clock offset dTrec is treated as white noise with zero mean. In these conditions, matrices have the form:
1 (n) = 1 1 0
0 (n) = 0 0
2 DT
gAGE
2 Being D T process noise associated to clock offset (in some way, the uncertainty in clock value).
JEAGAL, 2004-2005
147
Static positioning: constant coordinates and white noise clock
gAGE
S/A=on
148
b) Kinematic positioning
1) In case of a fast moving vehicle, coordinates will be modeled as white noise with zero mean, and the same rationale applies for clock offset:
0 (n) = 0 0 0
d2x (n) =
d2y d2z
2 DT
2) In case of a slow moving vehicle, coordinates2 may be modeled as random walk, process spectral density q = d : &
dt
1 (n) =
gAGE
1 1
& q dx t (n) =
& q dy t & q dz t
2 DT
JEAGAL, 2004-2005
149
Pure Kinematic positioning: white noise coordinates and clock
gAGE
S/A=on
150
Kinematic positioning: random walk noise coordinates and white noise clock
gAGE
S/A=on
151
Positioning: Random walk coordinates with Q=0 and white noise clock
gAGE
S/A=on
152
Annex
Bancroft method for direct estimation of receiver position and clock offset
Bancroft method allows to get a direct solution of receiver coordinates and clock offset (x,y,z,DT), without any a priori knowledge. So, this method may be used to obtain the initial value (x0, y0, z0) for navigation equations (cold start)
gAGE
JEAGAL, 2004-2005
153
1 P = Pn =
(x x ) +( y y ) +(z z )
1 2 1 2
1 2
+ cDT
(x x ) +( y y ) +(z z )
n 2 n 2
n 2
+ cDT
x y = M B1 ( 1 + a) z cDT
Where:
B 1 1, B 1 1 2 + 2 B 1 1, B 1 a 1 + B 1 a , B 1a = 0
x = M xn
1
y M
z M
yn
zn
P M Pn
1
1 a1 1 a 1 1 = a = 2 aj = a3 2 1 a4 1
1 0 a , b = [ a1 , , a 2 , a 3 , a 4 ] 0 0
xj xj j j y , y zj zj P j P j
0 1 0 0 0 0 1 0 0 b1 0 b2 0 b3 1 b4
154
gAGE
Satellite coordinates
Measured pseudorange
JEAGAL, 2004-2005
Comments:
To improve the estimation, the measured pseudorange can be substituted by:
P j + dt j rel j TGD j
Nevertheless, to compute the ionospheric and tropospheric delay, the receiver coordinates are needed. Thence, after computing the approximate receiver coordinates, a second iteration is needed to compute the solution using all the corrections (which is done using the linearized equations).
gAGE
155
Lesson 7
Code and phase differential positioning. Floating versus fixing ambiguities
gAGE
JEAGAL, 2004-2005
156
Differential Positioning
A receiver (rover) is positioned regarding to another receiver (reference) whose coordinates are known. It allows to cancel out most of the common errors between both receivers, improving dramatically the positioning accuracy.
gAGE
JEAGAL, 2004-2005
157
ERORS on the Signal

Space Segment Errors: Space Segment Errors: Clock errors Clock errors Ephemeris errors Ephemeris errors Propagation Errors Propagation Errors Ionospheric delay Ionospheric delay Tropospheric delay Tropospheric delay Local Errors Local Errors Multipath Multipath Receiver noise Receiver noise
JEAGAL, 2004-2005
Common Strong spatial correlation Weak spatial correlation No spatial correlation
gAGE
158
DIFFERENTIAL GPS (DGPS) with code

Computed position
Computed position
True position (known)
Error
More precise position
gAGE
In short baselines, similar errors are expected

159
gAGE
Positioning error in station bell (S/A=on)

160
gAGE
Basically, the same errors are found in station ebre, 100Km far from bell
(the same satellites are used in the navigation solution)
161
gAGE
Most of the errors cancels out when computing the difference between errors in ebre and bell.
(the same satellites are used in the navigation solution)
162
gAGE
Zoom of previous plot
JEAGAL, 2004-2005
163
In the previous example, the differential error has been cancelled in the position domain. But: It requires to use the same satellites in both stations. It does not takes benefit from the internal correlations between measurements and geometry Thence, is much better to solve the problem in the pseudorange domain than in the position domain. Concept: Concept: A reference station (its exact position is known) computes A reference station (its exact position is known) computes a differential correction for each satellite in view. a differential correction for each satellite in view. The user applies this correction to remove most part The user applies this correction to remove most part of the common errors in pseudorange. of the common errors in pseudorange.
gAGE
164
Range Differential Correction Calculation

Broadcast SV Position
Actual SV Position
Calculated Range
ref
PRref
Measured Pseudoranges
PRuser
REF
Corrections Calculation
Differential Message Broadcast
PRC
USER
Known Reference Location
The first receiver, in a reference station, can calculate these The first receiver, in a reference station, can calculate these errors knowing its exact location (corrections PRC errors knowing its exact location (corrections PRC calculated by the ground station): PRC= PRref --ref calculated by the ground station): PRC= PRref ref The second receiver (the user) will use these corrections to The second receiver (the user) will use these corrections to adjust its own measurements and increase the accuracy of adjust its own measurements and increase the accuracy of these measurements: PRuser--PRC these measurements: PRuser PRC
gAGE
165
1. Differential positioning (with simple differences)

Lets P jrov and P jref be code measurements of rover (rov) and reference station (ref), respectively, for satellite j. Thence:
j Prefitrov =
x0,rov x j
j 0,rov
dxrov +
y0,rov y j
j 0,rov
dyrec +
z0,rov z j
j 0,rov
j dzrov + cDTrov + rov
j j Prefitref = cDTref + ref
Introducing the following notation to symbolize the simple difference j j j of measurements rov ref , the difference of previous equation can be written as:
gAGE
Prefit j =
x0,rov x j
j 0,rov
dxrov +
y0,rov y j
j 0,rov
dyrec +
z0,rov z j
j 0,rov
dzrov +(cDT) + j
JEAGAL, 2004-2005
166
And, considering all satellites in view from both receivers, the following equation system can be written. This system is similar to that of absolute positioning, but the relative clock between both receivers is instead estimated: (cDT)=cDTrov-cDTref
xo , rov x1 1 0, rov 2 x o , rov x 2 = 0, rov n x o , rov x n 0, rov yo , rov y1
1 0, rov
zo , rov z1
1 0, rov
Prefit 1 Prefit 2 ........ n Prefit
yo , rov y
2 0, rov
zo , rov z
2 0, rov
.......... yo , rov y n
n 0, rov
zo , rov z n
n 0, rov
1 1 1
dxrov dy rov dz rov (cDT )
gAGE
This system can be solved applying the same mathematic tools than in absolute positioning (LMS, WMS, Kalman filtering,...)
JEAGAL, 2004-2005
167
Residual GPS Pseudorange Errors

Without DGPS Correction ERRORS Receiver Noise Multipath Satellite Clock (S/A=off) Satellite Clock (S/A=on) Ephemeris (S/A=off) Ephemeris (S/A=on) Ionosphere (raw iono) Troposphere Bias (meters) 0.5 0.3 to 3.0 2.0 21.0 10.0 (extreme) 100.0 (extreme) 2 to 10 (times obliquity) 2 (times obliquity) Random (meters) 0.2 0.2 to 1.0 0.0 0.1 0.0 0.0 <0.1 (times obliquity) <0.1 (times obliquity) Zero Basline Zero Latency DGPS Bias (meters) 0.5 0.4 to 3.0 0.0 0.0 0.0 0.0 0.0 Random (meters) 0.3 0.2 to 1.0 0.10 0.14 0.0 0.0 <0.14 Decorrelation with Latency Velocity (m/s) 0.0 0.0 0.02 0.21 negl. <0.01 0.092 Acceleration (m/s2) 0.0 0.0 0.004 0.004 negl. <0.001 negli. Geographic Decorrelation (m/100 Km)
0.0 0.0 0.0 0.0 <0.05 <0.5 <0.2
gAGE
0.0
<0.14
negl.
negli.
<0.05
JEAGAL, 2004-2005
168
2. Differential positioning with double differences

Taking one station and one satellite as a reference, the double differences between satellites and stations can be computed from the single ones:
rov j
ref
=
Thence:
=
ref
rov
gAGE
j =
JEAGAL, 2004-2005
j rov
j ref
R rov
R ref
j rov
R rov
j ref
R ref

169
And applying the double differences to the measurement equations, it follows that:
x0,rov x j y0,rov y j z0,rov z j j Prefit j = j dxrov + dyrec + j dzrov + j 0,rov 0,rov 0,rov
where:
(cDT) =( cDT) =0
Considering all satellites in view from both receivers, the following equation system may be written, where the clock term is cancelled.
xo , rov x1 1 0, rov 1 Prefit 2 xo , rov x Prefit 2 2 = 0, rov ........ n Prefit n xo , rov x n 0, rov
JEAGAL, 2004-2005
gAGE
yo , rov y1 1 0, rov yo , rov y 2 2 0, rov .......... yo , rov y n n 0, rov
zo , rov z1 1 0, rov zo , rov z 2 2 0, rov zo , rov z n n 0, rov
dxrov dy rov dz rov
170
gAGE
Positioning error in station bell (S/A=on)
JEAGAL, 2004-2005
171
gAGE
172
gAGE
173
As it has been seen in the previous plots, double differences and single differences performs similar when positioning with code pseudoranges. Nevertheless, doubles differences have an important application when positioning with carrier phases, in particular for the fixing ambiguities techniques.
gAGE
JEAGAL, 2004-2005
174
3. Carrier phase positioning

3.1 Differential positioning with double differences using code and phase measurements (floating the ambiguities) Phase measurements are modeled in similar way than code ones, but taking into account the phase ambiguities B, which must be estimated together with rover coordinates (dx,dy,dz).
xo , rov x1 1 0 , rov 1 Prefit ( P )1 xo , rov x 2 1 Prefit ( L ) 0, rov = ........ M Prefit ( P ) n xo , rov x n n n Prefit ( L ) 0, rov n xo , rov x 0n, rov yo , rov y1 1 0, rov yo , rov y1 2 0 , rov M yo , rov y n n 0, rov yo , rov y n n 0 , rov zo , rov z1 1 0, rov zo , rov z1 2 0 , rov M zo , rov z n n 0 , rov zo , rov z n n 0, rov 0 L 0 L M L 0 L 0 L 0 L L L L L dxrov 0 L 0 dyrov d z rov 0 L 0 B1 M M L M Bk 0 L 0 M Bl M 1 L 0 { B (l ) s
(k )
1 { M
gAGE
This system can be solved with a Kalman filter, assuming ambiguities B constant along continuous carrier phase arcs, and white noise at cycle-slips.
JEAGAL, 2004-2005
The tropospheric delay can be also estimatedM., Sanz J,random walk Hernndez-Pajares M., Juan as a Salazar D., Ramos P.
175
Comments: More accurate modeling for phase positioning is needed. It should involve: Phase Wind-up (the GPS signal is polarized wave) Antenna Phase center To estimate tropospheric delay (random walk) To adjust broadcast orbits (long baselines) Tidal Effects (in particular solid Earth Tides)
gAGE
JEAGAL, 2004-2005
176
Solid Earth tides

Wind-Up (Lc)
Tropospheric delay
Antenna Phase center (sat.)
gAGE
JEAGAL, 2004-2005
177
Differential positioning at 420 km from reference station

Broadcast orbits
Broadcast orbits (adjusted)
gAGE
JEAGAL, 2004-2005
178
Differential positioning with double differences using code and phase measurements (floating the ambiguities)
Phase measurements are modeled in similar way than code ones, but taking into account the phase ambiguities B, which must be estimated together with rover coordinates (dx,dy,dz).
xo , rov x1 1 0 , rov 1 Prefit ( P )1 xo , rov x 2 1 Prefit ( L ) 0, rov = ........ M Prefit ( P ) n xo , rov x n n n 0, rov Prefit ( L ) n xo , rov x 0n, rov yo , rov y1 1 0, rov yo , rov y1 2 0 , rov M yo , rov y n n 0, rov yo , rov y n n 0 , rov zo , rov z1 1 0, rov zo , rov z1 2 0 , rov M zo , rov z n n 0 , rov zo , rov z n n 0, rov 0 L 0 L M L 0 L 0 L L L L L L dxrov 0 L 0 dyrov d z rov 0 L 0 B1 M M L M Bk 0 L 0 M Bl M 1 L 0 { B (l ) s
(k )
1 { M
gAGE
This system can be solved with a Kalman filter, assuming ambiguities B constant along continuous carrier phase arcs, and white noise at cycle-slips. The tropospheric delay can be also estimated as a random walk
179
Resolving the Ambiguities:

Ambiguities must be estimated with a Kalman filter, together with coordinates (and troposphere,).
Ambiguity
The easier way to resolve the ambiguities is to treat them as REAL numbers, which are constant along continuous phase arcs and white noise when cycle-slips (this technique is called: floating ambiguities) The filter needs some time span to converge, and also there is estimation noise (due to the correlations with the other states of filter), which degrades the navigation solution at the level of few Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P. JEAGAL, 2004-2005 180 decimeters.
gAGE
Notice: If the ambiguities were fixed, we would be Notice: If the ambiguities were fixed, we would be positioning with measurements of few millimeters of noise. positioning with measurements of few millimeters of noise. Differential Absolute be removed using Differential On the other hand, the ionosphere can be removed using with 2-freq and On the other hand, the ionosphere can Error budget positioning positioning dual frequency measurements (with Lc, Pc combinations)! dual frequency measurements (with Lc, Pc combinations)! carrier phases with C1 code with C1 code Note: (Lc,Pc) Note: Klobuchar model only accounts for ~60% Satellite clock 1.0 for 0m Klobuchar model only accounts m ~60% For long baselines the ionosphere does not cancel! Orbits 2.0 m 0m For long baselines the ionosphere does not cancel!
Ionosphere Troposphere Receiver Noise Multipath SA on (SA=off) 5-10 m 0.5-1.0 m 0.5 m 0.6 m 30 (0)m 0.4m 0.2 m 0.5m 0.6m 0m Differential ~ mm Estimate 2 mm + Ambig. 5 mm
Carrier phase ambiguities must be estimated/fixed Typical GPS Differential with receiver coordinates: Precision standalone The accuracy will depend (5)m Horizontal 60 on how they are solved. 1.5m (also the tropospheric 70 (6)m orbits adjustments delay or Vertical 2.0m should be considered for high precision). 3-D 90 (8)m 2.5m
JEAGAL, 2004-2005
gAGE
181
Solution converges after ~20 minutes
New satellites are incorporated to the navigation solution.
gAGE
Differential Kinematic Positioning (floating ambiguities) using broadcast orbits and LC, PC measurements. Baseline 100Km
182
gAGE
Zoom of previous figure

183
Kinematic Posit. with Lc, Pc, FLOATING ambiguities (base line 100Km)
New satellites are incorporated to the navigation solution.
Solution converges after ~20 minutes
gAGE
(see the exercise in the book)

JEAGAL, 2004-2005
184
Why FIXING instead of FLOATING ambiguities ?

Lc
Carrier phase Lc repaired using floated ambiguities. The estimations need some time to converge, and do not converge to the true value The ambiguities are fixed to their true values. There is no estimation noise. The carrier phase is fully repaired and provides a very accurate pseudorange (few millimeters of error)
Lc
gAGE
JEAGAL, 2004-2005
185
FIXING versus FLOATING ambiguities
gAGE
(see the exercise in the book) All the ambiguities have been fixed in post-process.. All the ambiguities have been fixed in post-process
After fixing the ambiguities the carrier phases are use for positioning After fixing the ambiguities the carrier phases are use for positioning Hernndez-Pajares JEAGAL, 2004-2005 very accurate pseudoranges (with fewM., Juan M., Sanz J, Salazar D., Ramos P. as very accurate pseudoranges (with fewmillimeters of noise) 186 as millimeters of noise)
Fixing the Ambiguities:

To obtain centimeter accuracy it is necessary to apply FIXING To obtain centimeter accuracy it is necessary to apply FIXING ambiguities techniques: ambiguities techniques: These techniques exploit the fact that the ambiguities in f1 and These techniques exploit the fact that the ambiguities in f1 and f2 are INTEGER NUMBERS of wavelengths, and if it were f2 are INTEGER NUMBERS of wavelengths, and if it were possible to build combinations of measurements with a level of possible to build combinations of measurements with a level of noise under 1 wavelength, it would be possible to obtain the noise under 1 wavelength, it would be possible to obtain the exact value of the ambiguity by ROUNDING. exact value of the ambiguity by ROUNDING. Lets B1 and B2 be carrier phase bias in frequencies f1 and f2. Thence:
sat sat B1rec = 1 N1rec + k1rec + k1sat
N1,N1 are the ambiguities (integers)
sat sat B2rec = 2 N 2rec + k2 rec + k2 sat k1, k2 are the instrumental delays (real)
The instrumental delays cancel out when making the double differences. B1 = 1 N1 On the other hand B1, B2, Bw are integer B 2 = 2 N 2 multiples of their wavelengths. But, the last is not the true for Bc N1 W N 2 Bc = C W BW = W NW
gAGE
JEAGAL, 2004-2005
187
RTK: it isassumed that ionospheric refraction cancels out assumed that ionospheric refraction cancels out RTK: it is gAGE research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain (base lines < 15-20 Km) (base lines < 15-20 Km)
Example of algorithm for fixing L1, L2 ambiguities

1. Using Lc, Pc, the Kalman filter estimates the ambiguity together with the receiver coordinates (floating it such as in previous examples). 2. From the rough estimation of
Bc
Bc , the ambiguity N w is fixed:
Bw Lw Lc + Bc N w = int = int w w
3. Being fixed the ambiguity can be also fixed:
N w , the ambiguities N1
N 2
B1 L1 L2 2 Nw N1 = int = int 1 2 1
N 2 = N1 N w
4. From the exact values of the ambiguities N1 N 2 , the exact value of the NON-INTEGER ambiguity in Lc can be fixed: Being fixed the ambiguity in Bc, , Being fixed the ambiguity in Bc N1 N 2 then the measurement is accurate at then the measurement is accurate at Bc = c w the level of fewmillimeters!!! the level Juan M., millimeters!!! JEAGAL, 2004-2005 2 1 Hernndez-Pajares M.,offewSanz J, Salazar D., Ramos P. 188
(100 times more accurate than code). (100 times more accurate than code)
Error budget
GPS Standalone
1.0 m 2.0 m 5-10 m 0.5-1.0 m 0.5 m 0.6 m 30 (0)m
Differential CA code
0m 0m 0.4m 0.2 m 0.5m 0.6m 0m
Differential with 2-freq

(Phases, L1,L2 [Lc])
Satellite clock Orbits Ionosphere Troposphere Receiver Noise Multipath SA on (SA=off)
~mm Estimate 2 mm 5 mm
Typical GPS Differential Precision standalone From the conceptual point of view, when the Horizontal 60 (5)m 1.5m ambiguity is fixed, the phase measurements are repaired and thence, the positioning is like if 2.0m Vertical 70 (6)m very accurate codes (few millimeters of noise) 3-D 90 (8)m 2.5m were used!!!
JEAGAL, 2004-2005
Differential 3 cm 4 cm 5 cm
gAGE
189
Limitations: Limitations:
The previous algorithm is based on assuming The previous algorithm is based on assuming that ionospheric refraction is the same in the that ionospheric refraction is the same in the reference station and the rover, and cancels out reference station and the rover, and cancels out when forming the double differences. when forming the double differences. This hypothesis can only be assumed for short This hypothesis can only be assumed for short baselines (<15-20Km). baselines (<15-20Km).
gAGE
JEAGAL, 2004-2005
190
Wide Area Real Time Kinematics (WARTK)

For long baselines, the ionospheric refraction For long baselines, the ionospheric refraction STEC must be taken into account in previous STEC must be taken into account in previous equations. Thence: equations. Thence:
Lw Lc 1.98STEC + Bc N w = int w
L1 L2 STEC 2 N w N1 = int 1 2
gAGE
Note: being 2-1 = 5.4 cm, accuracy in STEC Note: being 2-1 = 5.4 cm, accuracy in STEC
must be better than 2.7 cm of L1-L2 delay. must be better than 2.7 cm of L1-L2 delay.
JEAGAL, 2004-2005
191
Vertical Ionospheric Refraction Experiment Day (077 2000, 15UT)

Central Europe: > 2.6cm L1-L2 at 15 km or more
At middle latitudes like in At middle latitudes like in Europe we also find large Europe we also find large ionospheric gradients which ionospheric gradients which do not allow to extend RTK up do not allow to extend RTK up to more than 15-20 Km. to more than 15-20 Km. delay Units: cm of L1-L2
gAGE
192
GPS alone standalone
ACCURACY
RASANT
SBAS (WAAS, EGNOS, MSAS)
1m
20cm
Global DGPS (IGDG)
10cm
TCAR RTK RTK
WARTK
gAGE
20km
100km
400km
BASELINE
JEAGAL, 2004-2005
Carrier phase
193
Code pseudoranges
But, how to achieve error ( STEC) < 2.7 cm L1-L2 at the user location?
Tomographic model STEC <10 cm in L1-L2
gAGE
JEAGAL, 2004-2005
194
But, how to achieve error ( STEC) < 2.7 cm L1-L2 at the user location?
Tomographic model STEC <10 cm in L1-L2
gAGE
JEAGAL, 2004-2005
195
WARTK: Can be based on Wide Area GPS Networks (i.e. EGNOS)

gAGE
An STEC with error less than 2.7 cm of L1-L2 delay must An STEC with error less than 2.7 cm of L1-L2 delay must achieved at the rover location!! achieved at the rover location!!
JEAGAL, 2004-2005
[gAGE/UPC patent,2000]
196
WARTK layout
WARTK Central Processing Facility

Permanent Receivers Network
....
CPF
1. Kalman
WARTK User
Real-time Iono. model
Iono. model
Sat. corr. Tropo. corr.
gAGE
2. Ambiguity fixing 3. Ionospheric Integrity 4. WARTK message JEAGAL, 2004-2005

LI = STEC+ BI =
SAT REC
Nedl + BI = ( Ne )i, j ,k si, j,k +BI

i j k
Real-time geodetic model

197
(Kp>4)
1998 May 3rd
JEAGAL, 2004-2005
gAGE
Ref.: Colombo O., Hernndez-Pajares M., J.M. Juan,Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P. J. Sanz. Wide-Area, carrier-phase ambiguity resolution using a tomographic model of the Ionosphere. Navigation 49(1), pp. 61- 69, 2002.
198
gAGE
JEAGAL, 2004-2005
Kp>4,
199
gAGE
200
Floating ambiguities: Broadcast orbits adjusted
gAGE
Kinematic versus true position of ALBH, with HOLB as base station (a 429 km baseline). Adjusted broadcast orbits used. Ambiguities "floated" (Lc biases estimated). Tropospheric refraction errors estimated. Triangles: dUP; black circles: dN, squares: dE; all meters.Sanz J, Salazar D., Ramos P. 201 Hernndez-Pajares M., Juan M., JEAGAL, 2004-2005
Fixing ambiguities: Broadcast orbits adjusted
gAGE
Kinematic versus true position of ALBH, with HOLB as base station (a 429 km baseline). Adjusted broadcast orbits used. Ambiguities fixed (Lc biases fixed). Tropospheric refraction errors estimated. Triangles: dUP; black circles: dN, squares: Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P. 202 dE; all meters. JEAGAL, 2004-2005
Fixing ambiguities: Precise SP3 orbits
gAGE
Kinematic versus true position of ALBH, with HOLB as base station (a 429 km baseline). Precise SP3 orbits used. Ambiguities fixed (Lc biases fixed). Tropospheric refraction errors estimated. Triangles: dUP; black circles: dN, squares: Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P. 203 dE; all meters. JEAGAL, 2004-2005
Broadcast orbits
Broadcast orbits Adjusted
Differential positioning Geomatics gAGE research group of Astronomy and at
420Technical University of Catalonia, Barcelona, Spain km from reference station
Floating
20 cm
20h
T(GPS hours)
23h
Precise SP3 orbits

8min: three ambig. fixed
Broadcast orbits Adjusted

10 min: three ambig. fixed
Fixing
10 cm
20h
JEAGAL, 2004-2005
10 cm
23h 20h
T(GPS hours)
T(GPS hours)
23h
204
Fixing ambig. with three frequencies: Galileo

Techniques proposed to improve the instantaneous positioning by using three-frequency systems share a similar basic approach: the double differenced integer ambiguities are successively solved from the longest to the shortest beat-wavelength, including "extrawidelane" and "widelane" combinations of carrier phases. In particular TCAR is a straightforward approach that tries to instantaneously solve (single-epoch) the full set of ambiguities. But TCAR (and ITCAR) is strongly affected by the ionos. refraction decorrelation with the distance. Ionosphere is a problem (for both 2 & 3 freq.) when :
gAGE
I > 0.26TECU
205
Classical TCAR Technique (1)

TCAR
1. To solve the extra-wide lane ambiguity by adding a pseudo-range combination 2. The wide lane combination ambiguity is estimated from the unambiguous extra-wide lane carrier phase, obtained in step 1 3. The L1 phase ambiguity is derived from the difference between L1 and the unambiguous wide lane obtained previously
Multipath Mitigation
P= 1 (P1 + P2 + P3 ) 3
gAGE
1 1 (M ew + ew ) + ... N ew = (Lew Pew ) = N ew +
ew
ew
JEAGAL, 2004-2005
206
Classical TCAR Technique (2)

TCAR
1. To solve the extra-wide lane ambiguity by adding a pseudo-range combination 2. The wide lane combination ambiguity is estimated from the unambiguous extra-wide lane carrier phase, obtained in step 1 3. The L1 phase ambiguity is derived from the difference between L1 and the unambiguous wide lane obtained previously
1
gAGE
0.058 cycles Nw/TECU

1
(Lw Lew + ewN ew ) = N w
( ew + mew mw ) +
( w ew )I + ... = N w
JEAGAL, 2004-2005
207
Classical TCAR (3)

TCAR
1. To solve the extra-wide lane ambiguity by adding a pseudo-range combination 2. The wide lane combination ambiguity is estimated from the unambiguous extra-wide lane carrier phase, obtained in step 1 3. The L1 phase ambiguity is derived from the difference error (I ) between L1 and the unambiguous wide lane obtained previously
< 0.26TECU
gAGE
-1.95 cycles N1/TECU

1
1 1 1 N1 = (L1 Lw + w N w ) = N1 ( w + mw m1 ) + (1 w )I
JEAGAL, 2004-2005
208
Comments about TCAR

TCAR is a straightforward approach that tries to TCAR is a straightforward approach that tries to instantaneously solve (single-epoch) the full set of instantaneously solve (single-epoch) the full set of ambiguities in three-frequency systems. But TCAR (and ambiguities in three-frequency systems. But TCAR (and ITCAR) is strongly affected by the ionos. refraction ITCAR) is strongly affected by the ionos. refraction decorrelation with the distance. decorrelation with the distance. The Ionosphere is a problem (for both 2 & 3 freq.) The Ionosphere is a problem (for both 2 & 3 freq.) when :: when
I > 0.26TECU
gAGE
JEAGAL, 2004-2005
209
Classical TCAR (3) + RT iono.

TCAR
1. To solve the extra-wide lane ambiguity by adding a pseudo-range combination 2. The wide lane combination ambiguity is estimated from the SAT unambiguous extra-wide B = STEC + BI = N e dl + lane I carrier phase,REC obtained in step 1
WARTK-3
(ref.sites)
LI =
( N )
i j k
e i , j ,k
s i , j , k +B I
3. The L1 phase ambiguity is derived from the difference error (I ) between L1 and the unambiguous wide lane obtained previously
< 0.26TECU
gAGE
-1.95 cycles N1/TECU

1
1 1 1 N1 = (L1 Lw + w N w ) = N1 ( w + mw m1 ) + (1 w )I
JEAGAL, 2004-2005
210
Centimetric navigation at hundreds of km from ref. stations Centimetric navigation at hundreds of km from ref. stations
GALILEO: WARTK3 [gAGE-ESA patent,2002] GALILEO: WARTK3 [gAGE-ESA patent,2002]

ADVANTAGES DISADVANTAGES
Seriously limited by iono. refraction The iono. delay still limits the 3rd ambiguity fixing. In spite of speeding up the navigation Kalman filter, a significant convergence time can be still needed (5-15 min).
TCAR ITCAR WARTK
Low computational load. Improved results by integrating TCAR in a navigation filter Accurate RT ionos. modelling, allows precise navigation at hundreds of km far from the nearest site Uses the extra-widelane, and an accurate iono. model to provide singleepoch navigation capabilities at hundreds of km far, and greatly speeding up the convergence of the Nav. Filter to just a few epochs.
WARTK-3
JEAGAL, 2004-2005
INSTANTANEOUS INSTANTANEOUS (i.e., single epoch) (i.e., single epoch) ambiguity fixing at ambiguity fixing at hundreds of Km far from hundreds of Km far from theJuan M., Sanz J, Salazar D., Ramos P. nearest ref. station Hernndez-Pajaresthe nearest ref. station M.,
WARTK-3 allows WARTK-3 allows
gAGE
211
Experiment : MADR at 404km from the nearest station (EBRE) at Solar Maximum Conditions
gAGE
FIXED STATION TREATED AS ROVER (MADR)

JEAGAL, 2004-2005
FIXED STATIONS (NETWORK)
212
ROVE: WARTK3 vs WARTK2 (starting up everything each 100/300 sec., including tropo.)
X-coordinate
Vertical coordinate/millimeters
Y-coordinate
gAGE
Z-coordinate
Starting-up everything: WARTK-2 (i.e. with GPS data) provides equivalent results to WARTK-3 (RMS of 2 cm and 100% amb. fixed), but after a convergence time of ~100 sec. (instead of instantaneously).
JEAGAL, 2004-2005
213
WARTK3 Lab.Test camp.: Frequencies

S2 1256.244 MHz
Mid-wide lane (S3-S2) =0.9834 m
S3 (E2) S1 (E1) 1256.244 MHz 1589.742 MHz
Short-wide lane (S1-S2) =0.8989 m
Extra-wide lane (S1-S3) =10.4662 m
Extra-wide lane (S1-S3): =10.4662 m Wide lane (S1-S2): = 0.8989 m S1: = 0.1886 m
gAGE
214
WARTK: subdecimeter-error navigation hundreds of kilometers away, and in single-epoch with 3-frequency systems
After general cycle-slips: Instantaneous recovery with 3-freq. systems (Galileo, modernized GPS), and about 1 minute with GPS.
ACCURACY
ACCURACY
GPS alone SBAS (WAAS, EGNOS, MSAS)
1m
GPS alone SBAS
20cm
Global DGPS (IGDG)
10cm
TCAR RTK
1m
WARTK
20km
20cm
100km
BASELINE
400km
gAGE
Global DGPS (IGDG)
10cm
WARTK-3 TCAR WARTK-2 RTK
Potential use of WARTK on SBAS service areas
Single epoch
JEAGAL, 2004-2005
1 minute
POSITIONING CONVERGENCE TIME
1 hour
215
WARTK*: experiments & results

Experiment Shortest baseline/km Rover/Fixed 116 / 286 400/900 Ionospheric Activity Mid.Solar Cycle & Quite Mid Solar Cycle & Active high lat. (Kp=6) Mid-Low Solar Cycle & Irreg. Solar Maximum Solar Max. & Supestorm Travelling Iono. Disturb. (TIDs) Solar Max. & Equator & Very Active (Kp to 9) Solar Max. Fixed Rec. Ambiguity success % 97 90-100 Roving Rec. Ambiguity success % 80-100 80 Kind of rover 4x4 Car IGS Site IGS Site IGS Site IGS Site IGS Site IGS Sites Sim. car Car 1:Car 2:Air. Fixed Site Region Dates Reported in
BellKin99 NWPacific1
Catalonia, NE Spain NWCanadaUSA NWCanadaUSA Central Europe Central Europe North Europe Central Asia to Oceania Central Europe Barcelona, NE Spain Iberian Peninsula Iberian Peninsula
23-03-99 03-05-98
Colombo et al. 99
(ION)
Hernndez et al. 00a, Colombo et al. 00 (GRL, PLANS) Hernndez et al. 00b (ION) Hernndez et al. 01 (GRL) Hernndez et al. 00b (ION) Hernndez et al. 01b (ION) Hernndez et al. 02 (JGR) Hernndez et al. 03a-b (IEEE ION-GPS2004 ION-GPS 2004 ION-GPS 2004
NWPacific2 SolarMax1 SolarMax2 Baltic99 Equator01
162/900 130/500 130/500 144/285 1000-3000/.
95-100 85-95 50-95 97 90
80-90 80 80 83 ---
28-04 to 01-05-98 19 to 22-04-00 12 to 15-07-00 25-08-99 06-03 to 02-04-01 17-03-00 (noon) 11-06-01 31-03-90 (ionos.) 31-03-90 (ionos.)
TCARdata (simulated) UNBAR01 WARTK3 Lab.Test 1-2 WARTK3 Lab. Test 3
130/300
100
92 (singleepoch) ~80 (with integrity) 100 99
TGARS, Navigation)
gAGE
70-115/100 178-238 /250-600 416/250600
Solar Max. & Strong TIDs Solar Max. Solar Max.
100 100 100
JEAGAL, 2004-2005
216
References:
Books:
Pratap Misra, Per Enge. Global Positioning System. Signals, Measurements, and Performance. Ganga Jamuna Press, 2004. Bierman, G.J., Factorization Methods for Discrete Sequential Estimation, Vol. 128 in Mathematics in Science and Engineering, Academic Press, New York, 1977. P.J.G. Teunissen, A. Kleusberg (Eds.) GPS for Geodesy, Springer, 1998. BW. Parkinson. Global Positioning System: Theory and Applications. Vol. I y Vol. II. Progress in Astronautics and Aeronautics. Vol 164. Published by the Institute of Aeronautics and Astronautics, Inc., 1996. ED. Kaplan. Understanding GPS: principles and applications. Artech House, cop. 1996. A. Leick. GPS Satellite Surveying. Ed. Wiley-Interscience Publication, 1994. B. Hofmann-Wellenhof et al. GPS, Theory and Practice. Springer-Verlag. Wien, New York, 1994. G. Seeber. Satellite Geodesy. Walter de Gruyter. New York, 1993. D. Wells. Guide to GPS Positioning. Canadian GPS Associates, 1990.
gAGE
217
Some papers about WARTK and TCAR

Colombo, O.L., M. Hernndez-Pajares, J.M. Juan, J. Sanz and J. Talaya, Resolving carrier-phase ambiguities on-the fly, at more than 100 km from nearest site, with the help of ionospheric tomography, ION GPS99, Nashville, USA, September 1999.
Colombo O.L., Hernndez-Pajares M., Juan J.M. and Sanz J., Ionospheric Tomography Helps Resolve GPS Ambiguities Onthe-Fly At Distances of Hundreds of Kilometers During High Geomagnetic Activity, Position Location and Navigation Symposium (PLANS 2000 IEEE conference), San Diego (USA), March 2000. Hernndez-Pajares M., J.M. Juan and J. Sanz, New approaches in global ionospheric determination using ground GPS data, Journal of Atmospheric and Solar Terrestrial Physics. Vol 61, 1237-1247, 1999a. Hernndez-Pajares M., J.M. Juan, J. Sanz and O.L. Colombo, Precise ionospheric determination and its application to real-time GPS ambiguity resolution, Institute of Navigation ION GPS99, Nashville, Tennessee, USA, September 1999b. Hernndez-Pajares, M., J.M. Juan, J. Sanz and O.L. Colombo, Application of ionospheric tomography to real-time GPS carrierphase ambiguities resolution, at scales of 400-1000 km, and with high geomagnetic activity, Geophysical Research Letters, 27, 2009-2012, 2000a. Hernndez-Pajares, M., J.M. Juan, J. Sanz, O. Colombo, H. Van der Marel, Real-time integrated water vapor determination using OTF carrier-phase ambiguity resolution in WADGPS networks, ION GPS2000, Salt Lake City, September 2000b. Hernndez-Pajares, M., J.M. Juan, J. Sanz, O.L. Colombo, and H. van der Marel, A new strategy for real-time Integrated Water Vapour determination in WADGPS networks, Geophysical Research Letters, 28, 3267-3270, 2001. Hernndez-Pajares, M., J.M. Juan, J. Sanz, O.L. Colombo, Tomographic modeling of GNSS ionospheric corrections: Assessment and real-time applications, ION GPS2001, Salt Lake, USA, September 2001b. Hernndez-Pajares, M., J.M. Juan, J. Sanz, O.L. Colombo, Improving the real-time ionospheric determination from GPS sites at Very Long Distances over the Equator, Journal of Geophysical Research, in press, 2002. Jung, J., P.Enge, B.Pervan, Optimization of Cascade Integer Resolution with Three Carrier GPS Frequencies, Proceedings of the ION-GPS 2000. Vollath, U., E. Roy, Ambiguity Resolution using Three Carriers -Performance Analysis using "Real" Data, GNSS Symposium,
JEAGAL, 2004-2005 Seville, May 2001. Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.
gAGE
218
Thats all, Thank you for your attention!
gAGE
JEAGAL, 2004-2005
219

GPSDP Gage

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

GPSDP Gage

Transféré par

Droits d'auteur :

Formats disponibles

research group of Astronomy and Geomatics Technical University of Catalonia, Barcelona, Spain

GPS Data processing: Code and Phase

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

11 due to the error clocks

How GPS Works

How GPS Works

One of the solutions is not on the Earth surface.

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

How GPS Works

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

How GPS Works Lesson 3:

GPS measurements and its combinations

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

How GPS Works

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

How GPS Works

Measurements: Ranges Atmospheric propagation: IONO, TROPO

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

How GPS Works

x iixi i y ii iiy z iiz ii cdtii cdti i

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

How GPS Works

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

GPS SIGNAL STRUCTURE

GPS Pseudorange Measurements

L1 ( t ) = a1 P ( t ) D ( t ) sin( f1t ) + a1C ( t ) D ( t ) cos( f1t )

GPS Phase Measurements f

L1(t), L2(t) (integrated Doppler)

C1 pseudorange computed from C binary code (on frequency 1)

Phase and Code pseudorange measurements

P1 + clock offset 20.000 Km

Phase and Code pseudorange measurements

Relative measurement (shifted by the unknown ambiguity N)

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

Code and phase measurements

Code (unambiguous but noisier)

Carrier Phase (ambiguous but precise)

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

GPS measurements: Code and Phase pseudoranges

L1 (t ) = a1 P (t ) D(t ) sin( f1t ) + a1C (t ) D(t ) cos( f1t ) L2 (t ) = a2 P (t ) D(t ) sin( f 2t )

noise (1% of ) [*]

Phase measurements Precise but ambiguous

RINEX measurement file RINEX measurement file

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

RINEX measurement file RINEX measurement file

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

RINEX measurement file RINEX measurement file

Measurement time (receive time tags) Number of tracked satellites

One satellite per row

RINEX measurement file RINEX measurement file

sat sat rec rec

sat sa sat = rel rec + T ro p rect + Io n rec + K rec + K sat +

Resolution: a) cat 95oct18casa___r0.rnx| rnx2txt > 95oct18casa.a b) See next plots:

Hernndez-Pajares M., Juan M., Sanz J, Salazar D., Ramos P.

An example of program to read the RINEX: rnx2txt

+ c (dtsta dt ) + rel + Trop + Ion