1. Loran Integrity & Performance Panel (LORIPP)
Per Enge, Stanford University, November 2003
Based on the work of:
Federal Aviation Administration, U.S. Coast Guard,
Peterson Integrated Geopositioning, Booz Allen Hamilton, Ohio University,
JJMA, ITT, University of Wales, Reelektronika & Stanford University
But the opinions may be mine alone & and the mistakes certainly are!

2. Performance Requirement
RNP 0.3 Requirements
Performance Requirement
Value
Accuracy (target)
307 meters
Monitor Limit (target) (HAL)
556 meters
Integrity
for all users in the coverage area (cannot average Boulder against Colorado Springs)
at all times (cannot average solar peak against quiet times)
under all conditions (in the presence of hazards)
10-7/hour
Time-to-alert
10 seconds
Availability at primary or alternate airport (minimum/target)
99.9/ 99.99%
Continuity (minimum/target)

3. Integrity Hazards (from Sherman Lo)
LORIPP work is organized around these hazards
with a system engineering group predicting coverage.

4. 10-7?
10-7 means:
Use the best available engineering to think through the corner cases.
Find the data that describes the hazard.
If the right data does not exist, collect some.
Design monitors to address any real integrity issues.
Remember, over design hurts continuity, availability and coverage.

5. Error Bounds, Not Accuracy
Prob(HPE > HPL) < 10-7 per hour
One or more cycle errors:
Envelope TOA at short ranges
Residuals test at long ranges
residual of
temporal ASF
temporal ASF
transmitter
receiver noise & RFI
spatial ASF

6. Integrity Analysis is best taught by example.
My favorite example (hazards) are:
evil waveforms for GPS
early skywave for Loran
remember these are only two examples from two long hazard lists.

7. DGPS Position Error Measured by Trimble at the 1993 Oshkosh Air Show
Differential vertical error
up to 8.5 meters
Altitude (meters)
SV19 Visibility Period
Local time of day

8. C/A and P(Y) Measurement from Camp Parks

9. C/A and P(Y) Measurement from Camp Parks

10. Modeling Evil Waveforms (from Eric Phelts)
C/A PRN Codes
Chips
Volts  
1/fd
Correlation Peaks
Code Offset (chips)
Normalized Amplitude

11. Signal Quality Monitoring (from Eric Phelts)
48 Correlator Receiver Spacings
SQM2b 1
0.8
0.6
Normalized Magnitude
SQM2b E-L Spacings:
0.1 chips*
0.15 chips
0.2 chips
0.4
0.2 -1
-0.8
-0.6
-0.4
-0.2
0.2
0.4
0.6
0.8 1
Spacing (chips)

12. WAAS Safety Processor Safety Processor DO 178 level B WREs, level D
CNMP
Corrections Processor
DO 178, level D
Iono.
correct.
& GIVE
L1/L2
Biases
Range
Domain
Position
Domain
UDRE
GIVE + +
SV orbit
determination
& corrections
UDRE + + +

13. Error Bounds, Not Accuracy (from Sherman Lo)
LORAN WAAS with Latency Removed

14. Back to Loran – Early Skywave

15. ECD Perturbations at Fairbanks (from Bob Wenzel)
time in days UT (n.0 is early afternoon on n-1 in Western Alaska)
Large solar proton event on Jan. 10

16. TD Perturbations at Fairbanks (from Bob Wenzel)
300 nsec
time in days UT (n.0 is early afternoon on n-1 in Western Alaska)

17. Previous plots blown up
Caribou (9960W) to Sandy Hook 463NM or 857 km
from Bob Wenzel

18. Monitor Using 228 Paths < 900 NM (from Ben Peterson)

19. Early Skywave Cures Monitor at LorStas and SAMs (not at airports!)
Range limits
Sample earlier (at 20 or 25 microseconds) & maybe speed the rise time of the pulse.
Channel sounding pulse
Receiver autonomous detection
See talks by Peter Morris, Bob Wenzel, Frenand Le Roux & Ben Peterson for much more.

20. Summary
10-7 means:
Use the best available engineering to think through the corner cases.
Find the data that describes the hazard.
If the right data does not exist, collect some.
Design monitors to address any real integrity issues.
Remember, over design hurts continuity, availability and coverage.
For Loran
We are well underway.
We have the right people, working the right issues.
But it is a big job

21. Backup Viewgraphs

22. Major Hazards
Temporal Variations of Groundwave including ASF, ECD and SS
Spatial Variations of ASF, ECD & SS
Weather related noise (p-static & atmospheric)
Early skywave
Aircraft dynamics
Man-made RFI
Transmitter Hazards
LORIPP work is organized around these hazards with a system engineering group predicting coverage.

23. Typical Distributions of TOA Measurement (from Ben Peterson)
Pcycle error = fn(Envelope uncertainty)
Probability Density of TOA
Accuracy = fn(Phase uncertainty)
Blue - Low SNR, Red - High SNR

24. Threat Flow from GPS Work
Ground
control
segment
upload
GPS
satellite
nav. message
signal dist.
Airborne
radio
environ.
RFI
multipath
Airborne
fault
detection
Airborne
rcvr.
ionosphere
& troposphere
Data
faults
Ground
radio
environ.
RFI
multipath
Ref. rcvr.
Level D code
cycle slips
Fault
detection
Data
broadcast

25. Monitor Performance nominal prob. density function Pr(false alarm)
faulted
prob. density
function
Pr(miss detect)

26. Simulation Data for Locus LRS IIID (from Bob Wenzel)