Practical Implementation Considerations in the Detection of GPS Satellite Signal Failure A.

Download Report

Transcript Practical Implementation Considerations in the Detection of GPS Satellite Signal Failure A.

Practical Implementation Considerations in the
Detection of GPS Satellite Signal Failure
A. J. Van Dierendonck, AJ Systems
Dennis Akos, Sam Pullen, Eric Phelts and Per Enge
Stanford University
27 June 2000
IAIN/ION Meeting
1
Topics










Paper Objective
Satellite Failure Threat Models
Airborne Receiver Design Constraints
Detection Metric Definitions/Implementation
Detection Metric Performance Requirements
Detection Metric Performance
Some Simulation Results
Noise/Interference Considerations
Scan Peak Requirements
Summary & Conclusions
27 June 2000
IAIN/ION Meeting
2
Paper Objectives


Provide summary and update of SQM
requirements
Present practical implementation of SQM
detection metrics



Reason why detection is not as complex as it may
seem
Provide an update on metric performance
Justification for changes to airborne receiver
requirements
27 June 2000
IAIN/ION Meeting
3
Satellite Failure Thread Models
27 June 2000
IAIN/ION Meeting
4
Undistorted Correlation Function
27 June 2000
IAIN/ION Meeting
5
Distorted Correlation Function
27 June 2000
IAIN/ION Meeting
6
Satellite Failure Threat Models
Threat A
Lead/Lag
-0.12 chips    0.12 chips
Threat B
Ringing
4 MHz  fd  17 MHz
0.8    8.8
Threat C
Combination
-0.12 chips    0.12 chips
4 MHz  fd  17 MHz
0.8    8.8
27 June 2000
IAIN/ION Meeting
7
Threat Model A
27 June 2000
IAIN/ION Meeting
8
Threat Model B
27 June 2000
IAIN/ION Meeting
9
Threat Model C
27 June 2000
IAIN/ION Meeting
10
Airborne Receiver Design Constraints
27 June 2000
IAIN/ION Meeting
11
Airborne E-L Discriminator Constraint
Space
Region
3 dB
Precorrelation
bandwidth (MHz)
Average
Correlator
Spacing
(Chips)
Instantaneous
Correlator
Spacing
(Chips)
Differential
Group Delay
(nsec)
1
2 < BW  7
0.045 - 1.1
0.04 - 1.2
 600
2
7 < BW  16
0.045 - 0.21
0.04 - 0.235
 150
3
16 < BW  20
0.045 - 0.12
0.04 - 0.15
 150
4
20 < BW  24
0.08 – 0.12
0.07 – 0.13
 150
27 June 2000
IAIN/ION Meeting
12
Airborne  Discriminator Constraint
Space
Region
3 dB
Precorrelation
bandwidth
(MHz)
Average
Correlator
Spacing
(Chips)
Instantaneous
Correlator
Spacing
(Chips)
Differential
Group Delay
(nsec)
1
2 < BW  7
0.045 – 0.6
0.04 – 0.65
 600
2
7 < BW  14
0.045 - 0.24
0.04 - 0.26
 150
3
14 < BW  16
0.07 - 0.24
0.06 - 0.26
 150
27 June 2000
IAIN/ION Meeting
13
SQM Detection Metric
Definitions/Implementation


Don’t require pseudoranges
Don’t need a “picket fence”
27 June 2000
IAIN/ION Meeting
14
Close-in Correlation Function Values
P0.1
E0.025
E0.075
L0.025
(E - L)0.1
E0.1
27 June 2000
IAIN/ION Meeting
L0.075
L0.1
15
SQM2b Test Metrics
ESFD1,k , j 
ESFD 2,k , j 
2 I P0.2,k , j

I E  L0.1,k , j
2 I P0.1,k , j
I E0.075,k , j  I L0.075,k , j
2 I P0.2,k , j
I E0.075,k , j
S P1,k , j 
S P 2,k 
I E  L0.2,k , j
I P0.2,k , j
I L0.075,k
I P0.2,k

  E0.2,0.1, j  0.2,k , j  0.1,k , j
I E  L0.1,k , j
2 I P0.1,k , j
  E0.15,0.1, j  0.15,k , j  0.1,k , j
  SP1, j
  SP 2, j
• Can use more ratios
27 June 2000
IAIN/ION Meeting
16
Detection Thresholds, Mean Values,
MDEs and MDRs


Real World Thresholds, Mean Values
D, test   D ,test  TD
TD  K ffd D, test
R, test   R ,test  TR
TR  K ffd  R , test
Used for Simulations (Include missed
detection probability)
MDE  ( K ffd  Kmd )D,test
MDR  ( K ffd  Kmd )R,test
27 June 2000
IAIN/ION Meeting
17
Fault-Free Probability Densities
0.5
Fault-Free
 test
0.45
Negative Failure
Positive Failure
0.4
Probability Density
0.35
0.3
0.25
Kmd
0.2
Kffd
0.15
0.1
0.05
0
-16.7
-8.35
0
8.35
16.7
Hypothetical Test Value Normalized with  test
27 June 2000
IAIN/ION Meeting
18
Metric Statistics

Based upon Stanford and NovAtel Roof Tests


Two environmental extremes
Multipath dominates noise and interference
27 June 2000
IAIN/ION Meeting
19
Difference Metric Performance
Comparisons
0.45
0.4
0.15 - 0.1, NovAtel 502
0.2 - 0.1, NovAtel 502
Delta Sigma Value - meters
0.35
0.2 - 0.1 Stanford
0.15 - 0.1 Stanford
0.3
0.25
0.2
0.15
0.1
0.05
0
0
10
20
30
40
50
60
70
80
90
Elevation Angle
27 June 2000
IAIN/ION Meeting
20
Ratio Metric Performance
Comparisons
0.0035
Ratio +0.1, NovAtel 502
Ratio -0.1, NovAtel 502
Stanford Ratio -0.1
Stanford Ratio +0.1
0.003
Ratio Sigma Value
0.0025
0.002
0.0015
0.001
0.0005
0
0
10
20
30
40
50
60
70
80
90
Elevation Angle
27 June 2000
IAIN/ION Meeting
21
Some Simulation Results


Justify simple metric definition
Justify change in airborne receiver design
constraints

Additional constraints on  discriminator
receivers


Smaller constraint space
Reduction of constraints on E – L discriminator
receivers


27 June 2000
Larger constraint space
No “Scan Peak” requirement
IAIN/ION Meeting
22
 Discriminator Error Performance –
SQM3
27 June 2000
IAIN/ION Meeting
23
 Discriminator Error Performance –
SQM2b
27 June 2000
IAIN/ION Meeting
24
Threat Model C E-L Discriminator
Protection Results
8.0
0.09
Max
Spacing
Maximum Test Value
0.08
6.0
0.07
5.0
0.06
4.0
0.05
3.0
0.04
2.0
0.03
1.0
0.02
0.0
0
500
1000
1500
2000
Chip Spacing - chips, Bandwidth - GHz
BW GHz
7.0
0.01
2500
Test Number (With Multiple Zero Crossings)
27 June 2000
IAIN/ION Meeting
25
Summary & Conclusions

SQM doesn’t have to be exceeding complex


Performance is dominated by multipath


Use correlation function amplitude measurements
instead of PRs
Performance varies widely with environment
Performance simulations are useful to define
airborne receiver design constraints



 constraint space
E – L scan peak requirement
DoD receiver accommodation
27 June 2000
IAIN/ION Meeting
26