multi-receiver vector tracking - stanford university
TRANSCRIPT
Multi-Receiver Vector TrackingYuting Ng and Grace Xingxin Gao
please feel free to view the .pptx version for the speaker notes
Cutting-Edge Applications
Sampling Device
NetworkSync.
time
voltage/ current
synchronized phasor measurements
PMU
interference mitigation
SAR-GNSSProcessing
remotesensing
UAV formation
flight
1
State-of-the-Art Techniques
Multiple antennas
Multiple receivers
Signal tracking
beamstearing
increased availability
PPP
DGPS
EKF IMU
clock
VDLL
VFLL
External aiding
increased robustness
RTK
2
State-of-the-Art Techniques
Multiple antennas
Multiple receivers
Signal tracking
External aiding
beamstearing
increased availability
PPP
DGPS
EKF IMU
clock
VDLL
VFLL
increased robustness
RTK
2
State-of-the-Art Techniques
Multiple receivers
Multiple antennas
Signal tracking
External aiding
beamstearing
increased availability
PPP
DGPS
EKF IMU
clock
VDLL
VFLL
Cooperative NavigationJoint Signal Tracking
increased robustness
RTK
2
Joint tracking of multiple GNSS receivers for increased accuracy, robustness and reliability
Multi-Receiver Vector Tracking
3
•Motivation | Coop Nav, Joint Signal Tracking
•Our Approach | Multi-Receiver Vector
Tracking
•Implementation | Hardware, Python SDR
•Experimental Validation
•Summary
Outline
Receiver Network
Channel 1
Channel N
...
...
...
Receiver
...
Receiver
N ChannelsM Receivers4 Unknowns
Scheme Scalar SRVT MRVT
Entities N x M Channels M Receivers 1 Network
Unknowns N x M x 4 M x 8 1 x 14
M Receivers8 Unknowns
1 Network14 Unknowns
incoming raw signal
MRVT Receiver Architecture
4
Comparison with other GNSS Processing Implementations
Single-Receiver Vector Tracking (SRVT) Multi-Receiver Vector Tracking (MRVT)
Antenna array Differential GPS (DGPS)
GNSS Receiver
GNSS Receiver
GNSS Receiver
Nav Filter
measurements/ coordinates
5
MRVT Block Diagram
6
MRVT Block Diagram
6
Single Receiver Vector Tracking Loop
EKF-based Filter
Vector Delay Lock Loop - James J. Spilker, Jr. Global Positioning System-Theory and Applications, Volumes I, II. AIAA, Washington, DC., 1996.7
Single Receiver PIA Vector Tracking
State Representation Process Noise Cov.
8
Single Receiver PIA Vector Tracking
State Representation Process Noise Covariance Matrix
σ2σ2
9
MRVT Block Diagram
10
MRVT Loop
11
MRVT Variations
MRPIAVT• Enforce timing constraints
→clock bias differ by constant→clock drift same
MRVTEnforce baseline, attitude
and velocity constraintsEnforce timing constraints
12
Correlator
Code and Carrier Discriminators
Measurement Update
Code and CarrierNCO
Channel
incoming GPS signal
a prioriinformation
TimeUpdate
LOS Projections
EKF
navigationsolution
Multi-Receiver Vector Tracking
MRVT Filter
Receiver
ReceiverNetwork
13
•Motivation | Coop Nav, Joint Signal Tracking
•Our Approach | Multi-Receiver Vector
Tracking
•Implementation | Hardware, Python SDR
•Experimental Validation
•Summary
Outline
Overall Implementation
Software (Python SDR)
GNSS antennas ADCs
Receiver
Network
4.60m
Hardware
14
Python SDR : Research Platform
• Share information effectively• SDR (Software Defined Radio)
• OOP (Object-Oriented Programming)
15
Python SDR : Research Platform
16
Python SDR : Teaching Platform
17
Python SDR : Teaching Platform
17
•Motivation | Coop Nav, Joint Signal Tracking
•Our Approach | Multi-Receiver Vector
Tracking
•Implementation | Hardware, Python SDR
•Experimental Validation
•Summary
Outline
MRVT Positioning Results
SRVT MRVT
18
zoomed inSRVT
zoomed inMRVT
MRVT Positioning Results
SRVT MRVT
18
zoomed inSRVT
zoomed inMRVT
MRVT Results: Reliable Altitudes
19
MR
VT
Res
ults
: Acc
urat
e B
asel
ines 19
MR
VT
Res
ults
: Les
s N
oisy
Spe
eds 19
Res
ults
with
Bui
ldin
gs+A
ccel
erat
ions
SRVT
MRV
T
20
Res
ults
with
Bui
ldin
gs+A
ccel
erat
ions
zoom
ed in
zoom
ed in
20
SRVT
MRV
T
MR
PIA
VT
Res
ults
:Im
prov
ed T
imin
g A
ccur
acy
SRVT
MR
PIAV
T
21
MR
PIA
VT
Res
ults
:Im
prov
ed T
imin
g A
ccur
acy
SRVT
MR
PIAV
T
21
MR
PIA
VT:
Res
ista
nt to
Jam
min
gA
dded
noi
se o
f diff
eren
t lev
els Clo
ck d
rift c
ompa
rison
ac
ross
diff
eren
t noi
se p
ower
Clo
ck b
ias
com
paris
on
acro
ss d
iffer
ent n
oise
pow
er
22
MR
PIA
VT:
Rob
ust t
o M
eaco
ning
cloc
k bi
ascl
ock
drift
||sta
te e
rror c
ov||
cloc
k bi
ascl
ock
drift
||sta
te e
rror c
ov||
wea
k m
eaco
ning
(1:1
pow
er ra
tio)
stro
ng m
eaco
ning
(100
:1 p
ower
ratio
)
robu
st to
at
tack
atta
ckde
tect
ed
23
MR
PIA
VT:
Det
ects
Dat
a S
poof
ing
cloc
k bi
ascl
ock
drift
||sta
te e
rror c
ov||
sate
llite
clo
ck b
iase
s sh
ifted
by
4s
atta
ckde
tect
ed
Atta
cked
af 0
in
Sub
fram
e 1
of
Eph
emer
ides
24
•Mot
ivatio
n | C
oop
Nav
, Joi
nt S
igna
l Tra
ckin
g
•Our
App
roac
h | M
ulti-
Rec
eive
r V
ecto
r
Trac
king
•Impl
emen
tatio
n | H
ardw
are,
Pyt
hon
SD
R
•Exp
erim
enta
l Val
idat
ion
•Sum
mar
y
Out
line
•Pr
opos
ed M
ulti-
Rec
eive
r Ve
ctor
Tra
ckin
g (M
RVT
)–
Coo
pera
tive
navi
gatio
n–
Join
t sig
nal t
rack
ing
•D
esig
ned
MR
VT A
rchi
tect
ure
Impl
emen
ted
in P
ytho
n SD
R
•C
ondu
cted
MR
VT E
xper
imen
ts
•Va
lidat
ed P
erfo
rman
ce Im
prov
emen
ts
–M
RVT
: D
emon
stra
ted
incr
ease
d ac
cura
cy in
hor
izon
tal
head
ing,
alti
tude
s, b
asel
ines
, spe
eds,
tim
ing
–M
RPI
AVT
: Dem
onst
rate
d im
prov
ed ti
min
g ac
cura
cy
and
prec
isio
n, ro
bust
ness
aga
inst
tim
ing
atta
cks
Sum
mar
y
25
Than
k yo
u!
From
all
of u
s @
UIU
C