estimation of antenna phase center offset for bds igso and ...€¦ · dc4 ds4 bc1 bs1 x-offset...
TRANSCRIPT
Chang’an University
Estimation of antenna phase center
offset for BDS IGSO and MEO
satellites
Guanwen Huang, Xingyuan Yan, Qin Zhang, et al.
Chang’an university, Xi’an, China
IGS Workshop· Wuhan
October 2018
Chang’an University
1. background
• Antenna phase center offsets (PCOs), which are not only the error
source but also translation-parameters of the center of mass, play
an important role in precise positioning. On the other hand,
inaccurate phase center corrections will also lead to the uncertainty
of the reference frame and degrade the performance of the orbit
product.
• For BDS, phase center corrections of IGSO and MEO satellites were
estimated by ESA/ESOC (Dilssner et al., 2014) and Wuhan
university (Guo et al., 2016). However, there were significant
differences between the two estimations of the z-offset for IGSO,
and satellites that prevented incorporation of these results into
igs08.atx. (Montenbruck et al. 2017).
• Thus, it is necessary to further study the accurate estimation of
PCOs for BDS satellites, especially for IGSO satellites.
Chang’an University
2. PCOs parameters model
The phase center offset correction for a given unit vector
Ԧ𝑒𝑟 = (𝑠𝑖𝑛𝛼𝑠𝑖𝑛𝜂, cos𝛼𝑠𝑖𝑛𝜂, 𝑐𝑜𝑠𝜂) of the observation can be
expressed as
∆𝜌 𝛼, 𝜂 = −𝑃 ∙ Ԧ𝑒𝑟= −𝑑𝑥 ∙ 𝑠𝑖𝑛𝛼𝑠𝑖𝑛𝜂 − 𝑑𝑦 ∙ 𝑐𝑜𝑠𝛼𝑠𝑖𝑛𝜂 − 𝑑𝑧 ∙ 𝑐𝑜𝑠𝜂
𝜂 is the nadir angle and 𝛼 is the azimuth in body-fixed sy
stem when viewing from the station; dx, dy, and dz are x-
offset, y-offset, and z-offset of PCOs, respectively.
Chang’an University
3. PCOs estimation for BeiDou-2
B1 & B2 combination
Chang’an University
Items Description
Number of stations 50 stations
Time interval From 1 January to 31 December 2015
observation Zero-difference GPS/BDS phase and code observations
Sampling rate 300 s
Elevation cut-off angle 3°
weighting Elevation-dependent; GPS:BDS = 4:1;
Stations coordinates Fixed to an IGS week-solution, others from GPS PPP;
orbits 72 hours arcs; 6 initial positions and velocities;
Solar radiation model ECOM1 VS ECOM2;
No a priori model;
Satellites PCOs initial values: (600,0, 1100) mm;
Constraints are (10, 1, 10) m;
Satellites PVs Set zero for all BDS satellites;
Receiver PCOs and PVs using L1 and L2 for B1 and B2
Strategies for data processing
3.1 PCOs estimation for BeiDou-2
Chang’an University
Distribution of the selected stations. Red cycle stations are from
MGEX and blue cycle stations are from iGMAS.
Distribution of the selected stations
3.1 PCOs estimation for BeiDou-2
Chang’an University
-0.4
-0.2
0.0
0.2
0.4
-0.4
-0.2
0.0
0.2
0.4
0 90 180 270 360
0 90 180 270 360-0.4
-0.2
0.0
0.2
0.4
0 90 180 270 360
C06
C07
C08
C09
Ho
rizo
nta
l o
ffse
ts c
orr
ectio
ns r
ela
tive
to
in
itia
l va
lue
s [
m]
C10
C11
x-offset
y-offset
C12
C14
-90
-45
0
45
90
beta
-90
-45
0
45
90
beta
-90
-45
0
45
90
beta
3.1 PCOs estimation for BeiDou-2
Impact of different SRP model – ECOM1
The rangers of x-offset variation related to beta angle are 0.2-0.4 m (IGSO)
and 0.1-0.2 m (MEO). The rangers of y-offset variation are 0.2m (IGSO)
and 0.1 (MEO).
Chang’an University
0
2
4
6
8
0
2
4
6
8
0 90 180 270 360
0 90 180 270 360
0
2
4
6
8
0 90 180 270 360
C06
C07
C08
C09
z-o
ffse
t co
rre
ctio
ns r
ela
tive
to
in
itia
l va
lue
s [
m]
C10
C11
z-offset
C12
C14
3.1 PCOs estimation for BeiDou-2
Impact of different SRP model – ECOM1
The rangers of z-offset variation are 2 m and 0.5 m for IGSO and MEO
respectively.
Chang’an University
0 60 120 180 240 300 360-1.0
-0.5
0.0
0.5
1.0
1.5
0 60 120 180 240 300 360 0 60 120 180 240 300
0 60 120 180 240 300-1.0
-0.5
0.0
0.5
1.0
1.5
0 60 120 180 240 300 0 60 120 180 240 300-0.5
0.0
0.5
0 60 120 180 240 300-0.5
0.0
0.5
0 60 120 180 240 300
Ho
rizo
nta
l P
CO
(m
) C06
C07
-90
-60
-30
0
30
60
90
be
ta (
de
g)
C08
Ho
rizo
nta
l P
CO
(m
) C09
C10
x-offset
y-offset
doys
-90
-60
-30
0
30
60
90
be
ta (
de
g)
C11
Ho
rizo
nta
l P
CO
(m
)
doys
C12
doys
-90
-60
-30
0
30
60
90
be
ta (
de
g)
C14
The x-offset variation are 1 and 0.5 m for IGSO and MEO respectively. For
the same types of satellites (IGSO or MEO), the larger beta angle, the
larger scatters, which is similar to Galileo PCOs (Steigenberge et al., 2016).
The y-offset variation is 4cm.
0 60 120 180 240 300 360-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
y-offset
YP
CO
(m
)
C09
-90
-60
-30
0
30
60
90
be
ta (
de
g)
3.1 PCOs estimation for BeiDou-2
Impact of different SRP models – ECOM2 (Arnold et al., 2015)
Chang’an University
Z-offset had a significant systematic variation of 5 m for IGSO (C06-C10), 1 m
for MEO.
Impact of different SRP model – ECOM2 (Arnold et al., 2015)
0 60 120 180 240 3000123456789
10
0 60 120 180 240 300 0 60 120 180 240 300
0 60 120 180 240 3000123456789
10
0 60 120 180 240 300 0 60 120 180 240 3000.0
0.5
1.0
1.5
2.0
0 60 120 180 240 3000.0
0.5
1.0
1.5
2.0
0 60 120 180 240 300
Ve
rtic
al P
CO
(m
)
C06
C07
-90
-60
-30
0
30
60
90
be
ta (
de
g)
C08
Ve
rtic
al P
CO
(m
)
C09
C10
z-offset
doys
-90
-60
-30
0
30
60
90
be
ta (
de
g)
C11
Ve
rtic
al P
CO
(m
)
doys
C12
doys
-90
-60
-30
0
30
60
90
be
ta (
de
g)
C14
3.1 PCOs estimation for BeiDou-2
Chang’an University
PX
PY
PZ
VX
VY
VZ
D0
Y0
B0
Dc2
Ds2
Dc4
Ds4
Bc1
Bs1
x-o
ffse
ty-o
ffse
tz-o
ffse
t
PXPYPZVXVYVZD0Y0B0
Dc2Ds2Dc4Ds4Bc1Bs1
x-offsety-offsetz-offset
-1.000
-0.6661
-0.3322
0.001739
0.3357
0.6696
1.000
C12 beta=41 deg
3.1 PCOs estimation for BeiDou-2
Correlation analysis between orbital and PCO parameters by using ECOM2P
XP
YP
ZV
XV
YV
ZD
0Y
0B
0D
c2
Ds2
Dc4
Ds4
Bc1
Bs1
x-o
ffset
y-o
ffset
z-o
ffset
PXPYPZVXVYVZD0Y0B0
Dc2Ds2Dc4Ds4Bc1Bs1
x-offsety-offsetz-offset
-1.000
-0.6661
-0.3322
0.001739
0.3357
0.6696
1.000
C06 beta=5.2 deg
PX
PY
PZ
VX
VY
VZ
D0
Y0
B0
Dc2
Ds2
Dc4
Ds4
Bc1
Bs1
x-o
ffse
ty-o
ffse
tz-o
ffse
t
PXPYPZVXVYVZD0Y0B0
Dc2Ds2Dc4Ds4Bc1Bs1
x-offsety-offsetz-offset
-1.000
-0.6661
-0.3322
0.001739
0.3357
0.6696
1.000
C10 beta=-41 deg
Higher beta:
x-offset and Dc2, Ds2, Bc1,
Bs1 (higher correlations);
y-offset has a weaker
correlation;
(a) (b)
(c)
Beta=-41 degBeta range:±42°
Beta=41 degBeta range:±42°
Beta=5.2 degBeta range:±77°
C10 C12
C06
Chang’an University
Higher beta:
y-offset and positions and velocities
(higher correlation);
x-offset has a weaker correlation
than that of ECOM2.P
XP
YP
ZV
XV
YV
ZD
0Y
0B
0B
cB
sx-o
ffse
ty-o
ffse
tz-o
ffse
t
PXPYPZVXVYVZD0Y0B0BcBs
x-offsety-offsetz-offset
-1.000
-0.6522
-0.3043
0.04348
0.3913
0.7391
1.000
C12 beta=41 deg
Beta=41 degBeta range:±42°
3.1 PCOs estimation for BeiDou-2
PX
PY
PZ
VX
VY
VZ
D0
Y0
B0
Bc
Bs
x-o
ffse
ty-o
ffse
tz-o
ffse
t
PXPYPZVXVYVZD0Y0B0BcBs
x-offsety-offsetz-offset
-1.000
-0.6522
-0.3043
0.04348
0.3913
0.7391
1.000
C10 beta=-41 deg
Beta=-41 degBeta range:±42°
PX
PY
PZ
VX
VY
VZ
D0
Y0
B0
Bc
Bs
x-o
ffse
ty-o
ffse
tz-o
ffse
t
PXPYPZVXVYVZD0Y0B0BcBs
x-offsety-offsetz-offset
-1.000
-0.6522
-0.3043
0.04348
0.3913
0.7391
1.000
C06 beta=7 deg
Beta=7 degBeta range:±77°
Correlation analysis between orbital and PCO parameters by using ECOM1
ECOM1 had smaller variations of x-offset and z-offset series than that of
ECOM2, and ECOM1 was chose to estimate PCOs in the next.
(a) (b)
(c)
C10 C12
C06
Chang’an University
CHA ESOC WHU
PRN x-offset y-offset z-offset x-offset y-offset z-offset x-offset y-offset z-offset
C06 554.17 0.00 3894.01 549.00 0.00 3049.00 586.40 0.00 2513.70
C07 554.17 0.00 4469.97 549.00 0.00 3236.70 586.40 0.00 2721.90
C08 554.17 0.00 4283.91 549.00 0.00 3842.60 586.40 0.00 3440.00
C09 554.17 0.00 4469.40 549.00 0.00 3973.60 586.40 0.00 3551.90
C10 554.17 0.00 4571.83 549.00 0.00 3882.10 586.40 0.00 4087.00
C11 572.00 0.00 2143.34 549.00 0.00 2069.50 575.00 0.00 1990.70
C12 572.00 0.00 2348.76 549.00 0.00 2313.50 575.00 0.00 2249.10
C14 572.00 0.00 2256.46 549.00 0.00 2311.70 575.00 0.00 2144.30
Estimations of PCOs from CHA, ESOC and WHU (unit: mm)
3.1 PCOs estimation for BeiDou-2
Chang’an University
orbit and clock offset overlapping precision
C06 C07 C08 C09 C10 C11 C12 C140
10
20
30
40
50
60
70
80
90
alo
ng-t
rack d
iffe
rence/m
m MGEX PCOs
Estimated PCOs
C06 C07 C08 C09 C10 C11 C12 C140
10
20
30
40
50
60
70
80
90
Cro
ss-t
rack d
iffe
rence/m
m
C06 C07 C08 C09 C10 C11 C12 C140
10
20
30
40
50
60
70
80
90
radia
l diffe
rence/m
m
C06 C07 C08 C09 C10 C11 C12 C140.00
0.05
0.10
0.15
0.20
0.25
ST
D o
f clo
ck o
ffset o
verlap/n
s
IGSO:13 mm (17%);MEO: 19 mm (31%) IGSO:12 mm (21%);
MEO: 5 mm (14%)
IGSO:5 mm (19%);MEO: 2 mm (15%)
STD:0.03ns (20%);
Time period: 2015,301-331
The influence of the reference clock data was subtracted by differencing with the
reference clock offset series, and C14 was chose as the reference satellite.
300 310 320 330-90
-60
-30
0
30
60
90
beta angle (deg)
doys
C06 C07 C08 C09
C10 C11 C12 C14
3.2 Validation of PCOs estimation for BeiDou-2
Chang’an University
CUT0
DJIG
DYN
G
FTNA
GM
SD
HAR
B
JFNG
KIR
UKRGG
KZN
2
MAL2
METG
NNOR
NRM
G
SIN1
ZIM3
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.14
0.15
RM
S o
f E
-com
ponent/
m
CUT0
DJIG
DYN
G
FTNA
GM
SD
HAR
B
JFNG
KIR
UKRGG
KZN
2
MAL2
METG
NNOR
NRM
G
SIN1
ZIM3
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.14
0.15
RM
S o
f N
-com
ponent/
m
MGEX PCOs
Estimated PCOs
CUT0
DJIG
DYN
G
FTNA
GM
SD
HAR
B
JFNG
KIR
UKRGG
KZN
2
MAL2
METG
NNOR
NRM
G
SIN1
ZIM3
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.14
0.15
RM
S
of
U-c
om
ponent/
m
coordinate estimation
E N U
MGEX PCOs 0.017 0.008 0.080
Estimated PCOs 0.014 0.006 0.056
Average RMSs of coordinates in the
ENU components (unit: m)
3 mm (17%) 2 mm (16%)
24 mm (30%)
Static solution, Only C06-C12,C14 were used; 2015,301-331;
16 stations selected randomly.
3.2 Validation of PCOs estimation for BeiDou-2
Chang’an University
Comparisons with the ESOC and WHU results – orbit and clock offset
Improved percent of the precision of orbit and clock offset overlap when compared with
the MGEX model
C06 C07 C08 C09 C10 C11 C12 C14
-10
0
10
20
30
40
imp
rove
d p
erc
en
t o
f o
rbit 3
D R
MS
(%
)
ESOC WHU CHA
C06 C07 C08 C09 C10 C11 C12
-10
0
10
20
30
imp
roved
pe
rce
nt
of
clo
ck o
ffset
ST
D (
%)
ESOC WHU CHA
ESOC WHU CHA
3D RMS of orbit 15.09% 16.17% 21.34%
STD of clock offset 12.18% 16.46% 18.93%
ESOC: PCOs and PVs; (Dilssner et al. 2014)
WHU: PCOs and PVs; (Guo, 2014)
CHA: PCOs and set PVs=0;
3.2 Validation of PCOs estimation for BeiDou-2
Chang’an University
RMS_E RMS_N RMS_U
MGEX 0.017 0.008 0.080
CHA 0.015 0.006 0.056
ESOC 0.017 0.013 0.057
WHU 0.016 0.013 0.056
The average RMSs in ENU components, the unit is m
Although there are differences in the orbit and clock precision by using the
different phase center correction models, but the impact on the coordinate
estimation is small for ESOC, WHU, and CHA models.
Comparisons with the ESOC and WHU results – coordinate estimation
3.2 Validation of PCOs estimation for BeiDou-2
Chang’an University
4. Coarse PCOs estimation for BeiDou-3
B1 & B3 combination
Chang’an University
Distribution of the stations used for POD. Blue are BeiDou-3 and
BeiDou-2 stations, and green are BeiDou-2 stations, red are GPS-only
stations.
Common Name SVN COSPAR ID PRN Launch date
BeiDou-3 M1 C201 2017-069A C19 2017.11.05
BeiDou-3 M2 C202 2017-069B C20 2017.11.05
BeiDou-3 M7 C203 2018-003A C27 2018.01.11
BeiDou-3 M8 C204 2018-003B C28 2018.01.11
BeiDou-3 M3 C205 2018-018A C22 2018.02.12
BeiDou-3 M4 C206 2018-018B C21 2018.02.12
BeiDou-3 M9 C207 2018-029A C29 2018.03.29
BeiDou-3 M10 C208 2018-029B C30 2018.03.29
Status of BeiDou-3 satellites.
http://mgex.igs.org/IGS_MGEX_Status_BDS.php
4.1 BeiDou-3 satellites
BeiDou-2: 24 (14 MGEX + 10 iGMAS)
BeiDou-3: 13 ( 3 MGEX + 10 iGMAS)
BeiDou-3e: 13 ( 3 MGEX + 10 iGMAS)
Chang’an University
0 60 120 180 240 300-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
0 60 120 180 240 300 0 60 120 180 240 300 0 60 120 180 240 300
0 60 120 180 240 300 360-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
0 60 120 180 240 300 0 60 120 180 240 300 0 60 120 180 240 300
0 60 120 180 240 300-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
0 60 120 180 240 300 0 60 120 180 240 300 0 60 120 180 240 300
Horizonta
l P
CO
(m
) C19
C20
C21
-90
-60
-30
0
30
60
90
beta
(deg)
Horizonta
l P
CO
(m
) C27
C28
C29
-90
-60
-30
0
30
60
90
beta
(deg)
C30
Horizonta
l P
CO
(m
)
doys
C31
doys
C32
doys
C33
x-offset
y-offset
doys
-90
-60
-30
0
30
60
90
beta
(deg)
C34
ECOM1 was used to estimate PCOs, a loose constraint of 10 m for all PCOs (x-offset,
y-offset, and z-offset); time period is from 044 to 292, 2018.
B1I & B3I combination
4.2 PCOs estimation for BeiDou-3
Chang’an University
0 60 120 180 240 3000123456789
10
0 60 120 180 240 300 0 60 120 180 240 300 0 60 120 180 240 300
0 60 120 180 240 3000123456789
10
0 60 120 180 240 300 0 60 120 180 240 300 0 60 120 180 240 300
0 60 120 180 240 3000123456789
10
0 60 120 180 240 300 0 60 120 180 240 300 0 60 120 180 240 300
Vert
ica
l P
CO
(m
) C19
C20
C21
-90
-60
-30
0
30
60
90
beta
(d
eg
)
C22
Vert
ica
l P
CO
(m
) C27
C28
C29
-90
-60
-30
0
30
60
90
beta
(d
eg
)
C30
Vert
ica
l P
CO
(m
)
doys
C31
doys
C32
doys
C33
z-offset
doys
-90
-60
-30
0
30
60
90
beta
(d
eg
)
C34
4.2 PCOs estimation for BeiDou-3
B1I & B3I combination
Chang’an University
PRN x-offset y-offset z-offset
C06 602.6 -27.1 5386.2
C07 557.6 1.1 4982.2
C08 581.3 -46.6 6232.6
C09 574.4 -70.4 5662.8
C10 588.5 -24.6 5336.4
C11 584.7 14.1 2539.7
C12 601.1 16.3 2764.0
C13 590.5 -18.9 4247.8
C14 596.1 -3.2 2600.3
C19 -244.2 -21.4 2399.2
C20 -247.5 -16.3 2545.4
C21 -242.8 -28.6 2520.0
C22 -262.3 -4.4 2474.6
C27 29.4 -18.0 1699.1
C28 47.7 -2.8 1648.4
C29 -35.9 -19.8 1880.1
C30 8.0 -12.0 1655.8
C31 -37.1 -124.4 1657.9
C32 45.1 -341.2 2831.6
C33 -244.8 -9.3 2486.6
C34 -260.4 -0.7 2610.2
PRN x-offset y-offset z-offset
C06 545.0 0.0 3509.5
C07 545.0 0.0 4121.2
C08 545.0 0.0 4710.2
C09 545.0 0.0 5029.8
C10 545.0 0.0 4935.1
C11 545.0 0.0 2214.2
C12 545.0 0.0 2401.9
C13 —— —— ——
C14 545.0 0.0 2450.2
ESOC(Dilssner et al. 2014)
X-offset: 3~5 cm;
Y-offset: 1~7 cm;
Z-offset: 0.6~1.8 m (IGSO), 15~40 cm (MEO)
B1I & B3I combination
4.2 PCOs estimation for BeiDou-3
Short time of data leads to the
poorer accuracy of PCOs for
BeiDou-3 satellites, thus we
named the estimations as the
coarse PCOs.
Chang’an University
C06 C07 C08 C09 C10 C11 C12 C13 C14
0
10
20
30
40
50
60
3D
RM
S c
om
pare
d w
ith g
bm
(cm
)
manufacturer PCOs
estimated PCOs
Compared with GBMCompared with overlap
4.3 Validation of PCOs estimation for BeiDou-3C
06
C07
C08
C09
C10
C11
C12
C13
C14
C19
C20
C21
C22
C27
C28
C29
C30
C31
C32
C33
C34
0
20
40
60
3D
RM
S o
f o
rbit o
ve
rlap
(cm
)
manufacturer PCOs
estimated PCOs
type Average median
BeiDou-2 3.2 2.4
BeiDou-3 2.6 2.7
BeiDou-3e 2.6 2.0
Improvement of orbit overlap (cm)
BeiDou-2: 3.2 cm
BeiDou-3: 2.6 cm
BeiDou-3e: 2.6 cm
type Average median
BeiDou-2 1.2 2.3
Improvement of accuracy compared
with GBM final product(cm)
Orbit precision
Chang’an University
4.3 Validation of PCOs estimation for BeiDou-3
Clock offset precisionC
06
C07
C08
C09
C10
C11
C12
C13
C14
C19
C20
C21
C22
C27
C28
C29
C30
C31
C32
C33
C34
0.0
0.1
0.2
0.3
0.4
0.5
ST
D o
f clo
ck o
ffse
t (n
s)
manufacturer PCOs
estimated PCOs
For IGSO satellites, the STD
values of clock offset overlap are
increased;
For most of MEO satellites, the
STD values have an average
improvement of 0.03 ns, except for
C19 and C20.
Short time of data processing, and lack of stations limited the accuracy of
PCOs.
Chang’an University
5. Summary
• Estimated PCOs can improve the precision of orbit and clock offset;
• Estimated PCOs can improve the accuracy of BDS-only PPP, especially in
U-component;
• The three estimated PCOs have small impact on the coordinate results;
• Coarse PCOs of BeiDou-3 can slightly improve the precision of orbit and
clock offset, However, the improvement is limited by the accuracy of PCOs,
and higher accurate PCOs will be achieved by using long time data
processing in the future.
Chang’an University
0 60 120 180 240 300
-1.0
-0.5
0.0
0.5
1.0
PCOX
PCOY
Horizonta
l P
CO
(m
)
doys
-90
-60
-30
0
30
60
90
beta
(deg)
C14
0 60 120 180 240 300
-0.2
0.0
0.2
0.4
0.6
0.8
PCOX
PCOY
Horizonta
l P
CO
(m
)
doys
-90
-60
-30
0
30
60
90
beta
(deg)
C12
ECOM1 is suffer from stretch shape of satellite body, like Galileo and BeiDou;
higher beta angle caused larger scatters of x-offset and y-offset. Thus, a priori
box-wing model is more necessary for higher beta satellites to get a more stable
series of x-offset and y-offset.
6. Outlook
(1) SRP model
(2) IGSO z-offset
BeiDou-2 MEO and IGSO have the same shape of body, the main difference is
the geometry of the observations from the earth. The maximum nadir angles are
9°and 13°for the IGSO and MEO, the smaller the maximum nadir angle means
the higher correlation between z-offset and satellite clock offset, also some extra
correlations such as PZ,VY, VZ.
Adding the LEO data to enlarge the nadir angle is a capable way to improve the
accuracy of IGSO PCOs, especially for z-offset.
Chang’an University
Thanks for your attention!