Transcript Ionospheric Considerations for Wide Area GPS Augmentation

Welcome to Stanford!
Civil Aviation Administration of China &
Federal Aviation Administration
Welcome to Stanford
•
•
•
•
•
•
•
Ms. Lu Xiao Ping, Deputy Director General, ATMB
Ms. Zhang Jing, Director, Inter. Cooperation Division, ATMB
Mr. Li Xin, Director, R&D Division, ATMB
Mr. Pan Yong Dong, Deputy Director, Planning Division, ATMB
Mr. Cao Hui, Manager, Aeronautical Data Communication Company
Mr. Cai KaiQuan, Engineer, ADCC
Mr. Shi Le, Engineer, ADCC
•
•
•
•
•
Mr. Chris Dufresne, Computer Engineer, FAA
Mr. CJ Jones, FAA / ATO
Mr. Dave Burkholder, FAA / ATO
Mr. Sam El-Zoobi, FAA / ATO
Mr. J.C. Johns, Navigation Director, FAA
2
Agenda
7:30
Parking and Logistics
8:00
Welcome from the FAA ATO International, Mr. David Burkholder
8:15
Introductory Remarks and GPS/RAIM Objectives, Madame Lu
8:30
Overview of FAA Satellite Navigation, Mr. J.C. Johns
8:45
History of Stanford Involvement with FAA SatNav, Prof Per Enge
9:00
Absolute RAIM, Dr. Todd Walter
9:30
Break
9:45
Civil Monitoring, Dr. Xingxin Gao
10:15 GPS Modernization, Prof. Brad Parkinson
10:45 Open discussion on potential for GPS RAIM cooperation
between FAA, Stanford and CAAC ATMB
11:30 Lunch at the Faculty Club
1:00
Closing Remarks and document agreed to Next Steps for
presentation at JATSG/6 on 4/21 (FAA, ATMB and Stanford)
3
History of Stanford Involvement with FAA Satellite
Navigation: From 1990 to the GEAS
for the Civil Aviation Administration of China
by Per Enge (with the help of many)
April 20, 2009
The Global Positioning System
User Segment
Control Segment
5
Stanford Scope:
Maximize Aviation Benefits from GNSS
• Worldwide approach capability with vertical
guidance, but no airport equipment.
• Worldwide landing capability (Cat. I/II/III) with high
availability.
• Robust against
–
–
–
–
–
Faults
Rare normal
Ionosphere
Scheduled RFI
Unscheduled RFI
• Safety Analyses
6
Very Brief History of Stanford Work for the FAA
1990
2000
2010
Brad Parkinson gathers a GPS team at Stanford
Early work on LADGPS, WADGPS & RAIM
LADGPS flight trials based on in-track pseudolites
WADGPS flight trials based on vector corrections
Co-chair RTCA WG-4
LADGPS flight trials based on airport pseudolites
Work on operational benefits
Tunnel in the sky displays, wake vortex, CSPA, etc.
Design for safety: faults & rare normal events
Interaction with prime contractors
WAAS integrity performance panel (WIPP co-chairs)
LAAS integrity performance panel (LIPP)
WAAS Operational
LAAS Operational
All of the above was funded by the FAA through Cooperative Agreements
7
93-G-004, 95-G-005, 97-G-012, 00-G-012 & 08-G-007.
First 10 Years Focused on Flight Trials
8
Second 10 Years:
Faults & “Rare Normal” Events
October 1993
modulation fault
Clock “runoffs”
7/28/01, 5/26/03
6/11/03 & more
40 notable
iono events
during the last
solar peak
RFI events:
• San Diego
• St Louis
• Santa Cruz
April 10, 2007 ephemeris
fault & 24 smaller faults
over the last 5 years
9
Truncation of the Error Tail
0.14
dual freq.
GBAS
(Cat I/II/III)
air & ground
screening
(Cat I/II/III)
0.12
ground
screening
(Cat I)
0.1
PDF
0.08
0.06
0.04
0.02
0
0
5
10
15
20
25
30
35
40
45
User Vertical Position Error (meters)
10
Evolution of GNSS-Based Safety
2010
2020
2030
L1 Only
• RAIM
• SBAS
• GBAS
Dual freq. SBAS & GBAS
• 24 SVs Minimum
• 10-4 from GNSS
Dual freq. ARAIM
• Open service
• GPS: 30+ Slots
• Multi-constellation
• 10-4 from GNSS
GNSS Integrity Within
• GPS IIIC (1st 16) ++, or
• GNSS Safety of Life
• 24 SVs (GPS alone)
• 10-7 from GNSS
11
Our Current Emphasis: L5 & New Constellations
2010
2020
2030
Dual freq. SBAS & GBAS
• 24 SVs Minimum
• 10-4 from GNSS
Dual freq. ARAIM
• Open service
• GPS: 30+ Slots
• Multi-constellation
• 10-4 from GNSS
12
Absolute Receiver Autonomous Integrity
Monitoring (ARAIM) for 2020
GPS
Compass
VPLVPLVPL
VPL
Galileo
GLONASS
13
FAA Supported Graduates of the GPS Lab (1/2)
University Professors
• Penny Axelrad, UColorado
• Changdon Kee, SNU
• Boris Pervan, IIT
• Glenn Lightsey, UT Austin
• Demoz Gebre-Egziabher,
UMinn.
• Gabe Elkaim,f UCSC
• Shau-Shiun Jan, Taiwan
• David Bevly, Auburn U.
Novariant*
• Clark Cohen
• Stewart Cobb
• Dave Lawrence
• Paul Montgomery
• Mike O'Connor
• Tom Bell
• Frank Bauregger
Televigation*, Y.C. Chao
Traxis*
• Roger Hayward
• Jock Christie
• Rich Fuller
Nav3D*
• Andy Barrows
• Keith Alter
• Chad Jennings
Rossum*
• Matt Rabinowitz
• Guttorm Opshaug
• Ju-yong Do
M Shift*, Awele Ndili
Mapbar.com*, Donghai Dai
Meta-VR*, Andrew Hansen
NordNav*
• Per-Ludwig Normark
• Sasha Mitelman
OlinkStar*, Junlin Zhang
14
FAA Supported Graduates of the GPS Lab (2/2)
Medium Size Companies
• Jiyun Lee
• Ping-Ya Ko
• Yeou-Jyh Tsai
• Jaewoo Jung
• Gang Xie, SiRF
• Alexander Mitelman, NordNav
• Ung-Suok Kim
• Michael Koenig, SiRF
• Lee Boyce: Consultant
• Seebany Datta-Barua, ASTRA
• Euiho Kim, Wilcox
• Hiroyuki Konno, TopCon
• Harris Teague, Seagull
• Sharon Houck, Seagull
University or Govt. Researchers
• Sam Pullen
• Eric Phelts
• Sherman Lo
• Juan Blanch
• Konstantin Gromov, JPL
• Eric Olsen, Johns Hopkins APL
• Jenny Gautier, UC
• Ran Gazit, Rafael
• Hiro Uematsu, NASDA
• Andrew Hansen, FAA Volpe
Large Companies
• Andy Rekow, John Deere
• Eric Abbott: L3
15
Updated Stanford Work Plan for FY09
for JC Johns
by Per Enge (with the help of many)
April 20, 2009
Evolution of GNSS-Based Safety
from the GEAS
2010
2020
2030
L1 Only
• RAIM
• SBAS
• GBAS
Task 1: Optimize Single Frequency WAAS
Task 2: SDM from WAAS to LAAS
Task 3: GAST-D
Task 4: DCPS
17
Task 1: Optimize
Single Frequency WAAS
• Provide added robustness for upcoming
solar maximum
– Work with Raytheon to implement kriging
– Retune algorithms and storm detectors to
ensure maximum CONUS availability
• Work with Raytheon to implement improved
SQM for better continuity
• Provide training to Oklahoma City to ensure
they understand intent and design of
WAAS algorithms
18
Task 2: Use WAAS SDM to Validate Nominal
Model for New LAAS SVs
SV11 & 23 cause
LAAS SDM to trip
during SLS4000
development
1.5
WAAS SDM Trip Threshold
1
0.5
0
0
5
10
15
20
PRN Number
25
30
19
Task 2: Use WAAS SDM to Validate Nominal
Model for New LAAS SVs
[From WAAS PAN report]: Increasing trend used
to confirm SLS-4000 SDM was operating
correctly by flagging and excluding PRN 11.
20
Task 3: GAST-D is Seeking Iono Freedom
Ionosphere Anomaly Threat Model
425
Flat 375
mm/km
Flat 425
mm/km
Slant iono. gradient bound (mm/km)
375
300
200
100
5
15
30
45
65
90
SV elevation angle (deg)
21
Task 3: GAST-D
Truncates the Error Tail
0.14
Most errors are exactly zero due to,
e.g., CCD detection and exclusion
before anomaly affects users, but all
zero errors have been removed from
the histogram.
0.12
0.1
PDF
0.08
Safety limit
derived from
OCS  28 m
0.06
0.04
Worst-case
error  41 m
0.02
0
0
5
10
15
20
25
30
35
40
45
User Vertical Position Error (meters)
RTCA-24 Constellation; All-in-view, all 1-SV-out, and all 2-SV-out subsets
included; 2 SVs impacted simultaneously by ionosphere anomaly
22
Task 3: GAST-D
Truncates the Error Tail
0.14
dual freq.
GBAS
(Cat I/II/III)
air & ground
screening
(Cat I/II/III)
0.12
ground
screening
(Cat I)
0.1
PDF
0.08
0.06
0.04
0.02
0
0
5
10
15
20
25
30
35
40
45
User Vertical Position Error (meters)
23
Task 3: GAST-D Availability Tool
LAAS ground and
user PR error
models
GPS SV
almanac
Loop through
all outage
cases (0, 1, 2,
3 SV out)
Loop
through all
time epochs
over 1 day
Loop
through 12
U.S. airport
locations
Increment availability
counters & terminate
outage counters
Availability requires:
• VPL ≤ 10-meter VAL
• max. Svert constraints met
• max |Svert|  3.0
• max |Svert| + 2nd max |Svert|  5.0
Increment outage
counters
Compute User
Protection Levels for
current SV geometry
Yes
Are GAST-D
req’ts. met?
No
24
Task 3: GAST-D Availability vs Constellation
0
10
IRT 21-SV
-1
10
IRT 24-SV
IRT 27-SV
-2
Wk465 26-SV 10
IRT 30SV
Wk465 31-SV
-3
IRT 33SV
Un-availability
10
IRT 36SV
-4
10
-5
10
-6
10
0 SVs Out
1 SV Out
2 SVs Out
Number of SV's Unhealthy
3 SVs Out
25
Task 3: GAST-D (L1-only CAT II/III)
• Fall 2008 review led by Jason Burns &
John Warburton
– SU, IIT, & other KTAs reviewed GAST-D
technical status
– Identified issues that need further study (e.g.,
multiple faults, time-to-alert)
• SU supporting FAA review of ionosphere
anomaly mitigation alternatives
• ICAO/RTCA technical concept validation
desired by end of 2009
26
Task 4: DCPS
• Latest LAAS MOPS (DO-253C) limits DCPS to
horizontal navigation (i.e., no VPL for DCPS)
• Further changes needed - seek best combination of
several options:
– Implement a “screening HAL” of 50 – 200 meters  below
this value, HPL is not guaranteed to bound worst-case HPE
– Add airborne geometry screening via max. Shoriz (similar to
GAST-D), min. Nsat, or max. (NLGF_corr – Nsat)
– Add airborne RAIM integrity monitoring (for DCPS only)
• These include changes to MOPS and to iono. threat
model (relative to PA  1 SV impact only)
27
Evolution of GNSS-Based Safety
2010
2020
2030
Dual freq. SBAS & GBAS
• 24 SVs Minimum
• 10-4 from GNSS
Task 5: Dual freq WAAS (L5 Roadmap)
Task 6: Dual freq GBAS & JPALS
28
Aviation Benefits from
New Constellations & Signals
• Worldwide approach capability with vertical
guidance, but no airport equipment.
• Worldwide landing capability (Cat. II/III) with high
availability.
• Robust against
–
–
–
–
Ionosphere
Scheduled RFI
Unscheduled RFI
Malevolent RFI
29
Task 5: Dual Frequency WAAS
Convert Orange to Green
30
Task 5: Dual Frequency WAAS
• Finalize & support the L5 transition plan based on
–
–
–
–
L5 roadmap meetings
L2 semi-codeless sunset
Support L1/L5 avionics
Meetings with service providers, RTCA & Eurocae
• Publish scintillation white paper based on:
– International working group
– Jiwon Seo: outage statistics & correlation
– Tsung-Yu Chiou: receiver design
• L5 SBAS MOPS development at RTCA &
EUROCAE
– Determination of message contents and formats
– Determination of user algorithms
– Coordination with receiver manufacturers
31
Task 6: Dual-Freq. LAAS Combines DivergenceFree & Ionosphere-Free Smoothing
32
Task 6: Land-Based JPALS Ground Facility
Changed to divergence-free (DF) smoothing in LDGPS –iono-free
(IF) smoothing and LAAS-like SF smoothing are backups
In JPALS, adjust thresholds based on
observed local RF interference
GPS
SIS
Database
A
SISRAD
B
Removed in
JPALS– air
and ground
receivers
have similar
designs
JPALS
includes
multiple
corrections
for L1 vs. L2
and different
smoothing
types
MQM
SQR
Smooth
LAAS
SIS
SQM
DQM
Executive Monitor (EXM)
VDB
Message
Formatter
&
Scheduler
VDB
TX
LAAS
SIS
Correction
Average
MRCC
sm-Monitor
LAAS Ground System Maintenance
VDB
Monitor
VDB
RX
Multiple sets of B-values are
computed and monitored
33
Evolution of GNSS-Based Safety
2010
2020
Dual freq. ARAIM
• Open service
• GPS: 30+ Slots
• Multi-constellation
• 10-4 from GNSS
2030
GNSS Integrity Within
• GPS IIIC (1st 16) ++, or
• GNSS Safety of Life
• 24 SVs (GPS alone)
• 10-7 from GNSS
Task 7: System Def., Requirements & PL equations
Task 8: Experimental Validation
Task 9: International outreach
Task 10: F/A-18 Hornet
34
Task 7: System Definition
ARAIM for 2020
GPS
Compass
VPLVPLVPL
VPL
Galileo
GLONASS
35
Task 7: System Definition
ARAIM for 2020
Compass
Compass
Galileo
Compass
Galileo
Compass
Galileo
Compass
Galileo
Galileo
dual frequency open service
GLONASS
GLONASS
GLONASS
GLONASS
GLONASS
GPS
GPS
GPS
GPS
GPS
Ground
Control
Ground
Monitors
Air Traffic
Status
1. How do we close the loop? through Air Traffic
2. What do we ground monitor? URA & biases
3. How often do we close the loop? once per hour
4. What monitor network do we use? civil monitoring net
5. What do we ask of other service providers? same as us
36
Task 7: System Definition
2030: GPSIII C
GLONASS
GPS
GPS
GPS
GPS
GPS
GLONASS
GLONASS
GLONASS
GLONASS
Compass
Compass
Compass
Compass
Compass
Galileo
Galileo
Galileo
Galileo
Galileo
dual frequency PPS
Ground
Control
1.
2.
3.
4.
5.
Ground
Monitors
How do we close the loop? through the GPS SVs
What do we ground monitor? ephemeris + URA to 10-7
How often do we close the loop? once per minute (RRAIM)
What monitor network do we use? GPS
What do we ask of other GNSS service providers? Not so much (ARAIM?)
37
Task 7: Proposed VPL Equations
• GPS IIIC
n
s
VPLGPSIII  5.33

  URA  s
2
3,i
i1
2
1
2
i
2
trop,i
s
2
user,i
n
  s  
3,i
2
 URAi   3 
i1
• ARAIM
VPL0  K 0
n
s
2
3,i
i1
VPL j  K j
n


   URA  s
2
1
2
i
2
trop,i
i1,i j
n
  s  
3,i
2
 URAi   3 
i1

s    URA  s
2
j 3,i
s
2
user,i
2
1
2
i
2
trop,i
s
2
user,i
n
 
s   2  URAi   3    j
j 3,i
i1,i j
VPLARAIM  max VPL j
j0,n
38
Task 7: Proposed Translation from
Probabilities to Ground Monitoring
39
Task 7: Assertions Under Consideration
1. Ergodicity
2. Smoothness
3. Short Temporal Correlations
4. Symmetry
5. Independence of Errors
6. Full Threat Analysis
7. Corrective Action
40
Task 7: System Definition, Requirements &
Protection Level Equations
• Continue work briefed by Todd Walter at
February 2009 GEAS
– Finalize URE definition
– Finalize URA monitoring
– PL equations for multi-constellation ARAIM
– PL equations for GPS IIIC
• Seek wider review
• Coordinate with PSICA group to ensure
FAA assurance requirements are being
monitored and met.
41
Task 8: Validate Civil Monitoring
(Collaboration with FAATC & AMTI)
Civil monitoring is a trade between:
• Constellation size
• Robustness to SV failures
• Network size (URA bounding)
ARAIM 99.5%
coverage
24-1
24
27-1
27
30-1
30
No Real Time
Monitoring
3.7%
27.5%
9.56%
87.9%
79.8%
99.6%
8 stations
50.8%
88.3%
71.5%
96.7%
98.7%
100%
38 stations
71.2%
98.9%
90.0%
100%
99.9%
100%
42
Task 8: Validate Civil Monitoring of Signals
(Collaboration with FAATC & AMTI)
L-band Feed
Cavity
Filter
Low Noise
Amplifier
45dB
Azimuth/Elevation
Control
Digital filter bank with
control circuits
Automatic data recording
based on triggers
Nova for Windows
Satellite Tracking Software
15 m cable
Agilent Vector
Signal Analyzer (VSA)
43
Task 8: Validate Civil Monitoring of Measurements
(Collaboration with FAATC & AMTI)
• Objectives
– Nominal range errors & URE
– Detect & characterize range error & URE blunders
• Challenges
– Compute range error without post-processed truth
– Volume of data and computational load
• Milestones
– MATLAB Interface is ready
– Integrate the algorithm within the current monitoring
– Migration from nationwide network to worldwide
network
– Migration from L1/L2 to L1/L5
44
Task 8: Validate Flight Performance
Detection Process
Undetected Fault
Real-Time VPL and Error from True Position
meters
1500
75
500
50
35
0
0
0
2000
4000
Healthy subset VPE
Healthy subset VPL
100
All-in-view VPE
All-in-view VPL
1000
FDE VPL Interval and Position Error
6000
8000
10000
0
VPE : VPL ratio
Error/ VPL Ratio before Elimination
1
0.5
0.5
0
0
2000
4000
6000
seconds
8000
4000
6000
10000
8000
Error/ VPL Ratio after Elimination
1
0
2000
10000
0
2000
4000
6000
10000
8000
seconds
45
Task 9: International Outreach to
Enable Multi-Constellation ARAIM
Key: common understanding of providers responsibility
• All parties need to participate in requirement development
• Each constellation can choose level of performance
• Performance must be monitored to enable ARAIM
• New document for constellation requirements
• New document for avionics requirements
– MOPS & SARPS
– PL equations
– Message contents
46
Task 9: International Outreach
(Schedule for next 4 months)
• March 2-3 ICG Workshop on GNSS Interoperability: Benefits of multiconstellation ARAIM and requirements from GNSS service providers
• March 3-5 Munich Satellite Summit: GEAS vision and benefits of
multi-constellation ARAIM
• March 25-26: First meeting of Working Group C on interoperability of
integrity service provision.
• April 2-4: Outreach to African nations on GNSS - aviation applications
and the importance of understanding ionospheric
behavior. Emphasis on getting measurements during the upcoming
solar maximum.
• April 2-4: SBAS iono working group meeting.
• April 20: Civil Aviation Administration of China ATMB will visit Stanford
(HoD: Deputy Director Madam Liu)
• April 21-23: Eurocae WG-62 is having substantial discussion on the
use of multi-constellation ARAIM for vertical guidance.
• May or June: IWG to be hosted at Stanford. Will discuss L5 plans
• June 23: RTCA WG-2 need to start serious work on L5 MOPS
47
Task 10: Naval Aviation Enterprise
has use for LPV worldwide.
F/A-18 Hornet
• What research or study is
required to certify the system?
• Joint testing?
• ARAIM with one constellation
• SBAS in coverage
• SBAS/ARAIM in mixed
• WAGE based URE
• Availability/integrity
requirements in theatre
• TIM at Stanford in early April
48
Optional Task 11: Security
Against DoS & Spoofing Attacks:
Tri-lateration Based on Mode-S
Bring the safety
perspective early on.
What VALs & HALs
can be supported?
Preliminary FMEA
49