An ionospheric scintillation mitigation method for RTK positioning at

Transcrição

MundoGEO #Connect, Latin America 2014, São Paulo, 9th May 2014
An ionospheric scintillation mitigation
method for RTK positioning at low latitudes
Lei Yang, Jihye Park, Marcio Aquino and Alan Dodson
Presented by: Vinícius Stuani
Presentation Outline
• Work package overview – objective and approach
• Data Selection
• Mitigation strategies for ionospheric scintillation
+ preliminary results
• Interface to receiver
2
• Data Selection
3
Project Structure
UNOTT
Objective of this work package:
Develop strategic improvements in positioning and navigation algorithms
against the effects of ionospheric disturbances
4
NRTK Data Flow
Conventional
Ref Stn(s)
GNSS Obs
Ref Stn(s)
GNSS Obs
Iono model
5
Iono
Monitoring
Network Data
NRTK Central
Processing Facility
Rover
The CALIBRA Approach
NRTK Central
Processing Facility
Mitigation
Algorithm
Rover
NRTK Data Flow
Conventional
Ref Stn(s)
GNSS Obs
Ref Stn(s)
GNSS Obs
Iono model
6
Iono
Monitoring
Network Data
NRTK Central
Processing Facility
Rover
The CALIBRA Approach
NRTK Central
Processing Facility
Mitigation
Algorithm
Rover
Tasks
• Algorithm review and identification of weaknesses (completed)
• Data and software preparation (completed)
• Algorithm development at the observable level
• Screening (completed)
• Weighting (ongoing)
• Algorithm development at the positioning level
• Adaptive ambiguity resolution threshold (about to start)
• Network processing (ongoing)
• Testing and refinement
7
Topics to be
covered in this
presentation
• Data Selection
8
Data Selection [1]
• Single baseline RTK
• 10km base line
• Rover uses its locally measured ionospheric indices
• Network solution is being studied separately, which aims to give less noisy VRS
observations, and also provide the network computed ionospheric indices for the
rover location
• Scintillation indices
• Amplitude:
S4
• Phase:
σΦ  T (spectral strength @1Hz) and p (slope of the PSD)
• Data processing
• RTK engine
• Performance evaluated by the percentage of RTK ambiguity fixing, 3D error
9
mean/std/max of the fixed/float solutions
Data Selection [1]
10
• Testing Days
- 12 days in 2012: different seasons / levels of scintillation activity
• 24hr data with 15 sec interval
047
094
190
282
018
085
191
270
043
093
200
283
1.5
1
0.7
0.3
1.5
Max S4
Data Selection [2]
1
0.7
0.3
1.5
1
0.7
0.3
0
4
8 12 16 20 24
0
4
8 12 16 20 24
0
4
8 12 16 20 24
0
4
8 12 16 20 24
Time (Hr)
11
Low
Moderate
Strong
Spring
DOY 018, DOY 047
DOY 043
Summer
DOY 094
DOY 085
DOY 093
Autumn
DOY 190, DOY 200
DOY 191
-
Winter
DOY 282
DOY 270, DOY 283
• Data Selection
12
•
•
•
Based on the assumption that only a few satellites are affected by scintillation
Remove the most affected satellites from the positioning solution
Currently based on the S4 index only; an integrated index based on S4 and σΦ
is under investigation
Analyses carried out using two S4 thresholds, 0.7 and 0.5
On selected days (quiet, moderate, strong)
•
•
SV - 18
SV - 21
50
30
0.3
10
0
0.7
50
30
0.3
10
5
10
15
Time (Hr)
13
70
S4
0.7
Elevation (Degree)
S4
70
Elevation (Degree)
Screening [1]
Moderate to strong
scintillation case
20
0
5
10
15
Time (Hr)
Low scintillation
case
20
Observations
to be screened
Screening [2]
•
The impact of screening is small, but clearly noticeable
– The number of epochs/observation involved in screening is low
Day
018
043
047
085
093
094
190
191
200
270
282
283
Category
M
S
M
M
S
Q
Q
M
Q
S
M
S
% of screened
observations
0.57
1.30
0.08
0.19
1.17
0
0
0.08
0.002
0.87
0.02
1.46
•
Positive and negative impact trade-off
– If only use GPS, the negative impact (worse DOP) of introducing screening
is larger than the positive impact (‘cleaner observables’)
– If GPS+GLONASS is used, the positive impact of screening out 1-2 SVs is
small but noticeable; no test data is investigated for a more severe case
•
Float solutions are impacted more than the fixed solutions
•
If S4 threshold is lowered from 0.7 to 0.5, the influence on positioning accuracy
varies from day to day, can be either positive or negative
14
Weighting [1]
•
Based on weighting individual observations on the positioning solution
depending on their scintillation contamination level; assigning a very low
weight to a satellite in the positioning engine will be equivalent to
screening out that satellite
Observations with
different weights
15
Weighting [2]
•
The weighting scheme can be achieved by estimating the variances of the
phase/code tracking jitter of the specific observations
•
Three possible approaches
1. Standard constant variance per observable type, e.g.
(1mm)2
16
(30cm)2
(L1)
2.
Variance from tracking error depending on measured c/n0:
3.
Variance from L1 phase and code tracking error depending on c/n0 and
scintillation level (S4, p and T): [Conker et al., 2003]
Approach 1 x Approach 3
DOY 191 – short period of strong scintillation
Approach 1
Approach 3
Positioning error in North (dN), East (dE), and Up (dU)
with Ambiguity resolution index (1: fixed, 0: float)
S4
Approach 1
81.6%
Approach 23
75.3%
All
Fixed
Float
All
Fixed
Float
0.009/0.068
0.006/0.023
0.023/0.150
0.008/0.055
0.006/0.018
0.015/0.106
0.009/0.070
0.0120/0.021
-0.003/0.157
0.009/0.051
0.013/0.018
0.000/0.097
0.019/0.224
0.019/0.044
0.016/0.514
0.009/0.223
0.019/0.042
-0.022/0.441
0.7
0 4 8 12 16 20 24
[UTC hr]
DOY 191 Positioning result of scintillation hours (1- 4 UTC, Post sunset LT)
DOY 191
3D error:
Approach 1
RMS: 0.475 m
Approach 3
RMS: 0.441 m
Height error:
Approach 1
RMS: 0.253 m
Approach 3
RMS: 0.244 m
S4
DOY 283 – longer period of strong scintillation
Approach 1
Approach 3
Positioning error in North (dN), East (dE), and Up (dU)
S4
with Ambiguity resolution index (1: fixed, 0: float)
Approach 1
66.7%
Approach 3
62.7%
All
Fixed
Float
All
Fixed
Float
-0.033/0.429
0.006/0.031
-0.111/0.737
0.008/0.191
0.010/0.026
0.005/0.310
0.032/0.146
0.026/0.031
0.045/0.250
0.039/0.197
0.022/0.030
0.066/0.318
0.102/1.313
0.035/0.063
0.235/2.271
0.013/0.359
0.025/0.079
-0.007/0.578
0.7
[UTC hr]
DOY 283 Positioning result of scintillation hours (0-6 UTC, Post sunset LT)
3D error:
Approach 1
RMS: 0.773 m
Approach 3
RMS: 0.544 m
Height error:
Approach 1
RMS: 0.390 m
Approach 3
RMS: 0.244 m
S4
Variance comparison
SV - 18
0.7
50
S4
Elevation
70
30
Blue – Elevation
Green – S4
0.3
10
0
Code Variance ( cm2 )
Carrier Variance ( mm2 )
5
10
15
20
2500
Blue – approach 2
Red –approach 3
2000
1500
1000
500
21
0
0
5
10
15
20
15
Blue – approach 2
Green – approach 3
10
5
0
0
5
10
15
Time ( hr )
20
• Work package overview –objective and approach
• Methodology
22
Interface to receiver
•
Bespoke interfaces between the mitigation engine and the receiver have been
defined and are being implemented, with close collaboration between the
academic and the industrial partners
•
Aiming to provide key parameters for screening and weighting while trying to
minimize the implementation effort
•
In terms of screening, the satellites to be screened out are flagged
•
In terms of weighting, the estimated values of the code/phase tracking error
variances are provided for each observation
23
RTCM-like Message
SBF file
RTCM 3
Message
RTCM 3
Type 4025
Sub-block A
Header
Header
Sub-block A
Type 1001
Sub-block 1
Type 1002
Sub-block 2
DiffCorrtn
Sub-block
Type 1003
Sub-block 3
Header
•
•
•
•
•
•
•
•
•
Content
(for RTCM3)
24
•
•
•
Type 4025
(new)
•
•
•
For screening
For weighting
For future uses
Implementation Plan [1]
Rover SBF
Rover ISMR
UNOTT Mitigation Algorithm
Rover SBF
PPSDK
Rover SBF
Iono model
UNOTT Mitigation Algorithm
Rover SBF
PPSDK
Iono Monitoring
Network Data
Offline tests
25
Implementation Plan [2]
SSN Rover Receiver
UNOTT Mitigation
Algorithm
Rover SBF
Rover ISMR
Rover SBF
Rover ISMR
UNOTT Data
Centre
RTCM 4025 for
the SBF updating
SSN Rover Receiver
UNOTT Mitigation
Algorithm
Iono model
UNOTT Data
Centre
RTCM 4025 for
the SBF updating
SSN Rover Receiver
UNOTT Mitigation
Algorithm
Iono Monitoring
Network Data
• Real time tests using SSN receiver
– Using locally measured scintillation indices
• Real time tests with indices generated from the monitoring network
26
(INGV model need to be integrated)
Summary
•
The externally monitored ionospheric scitillation parameters could be
used to mitigate the ionospheric disturbance, and improve the
positioning performance
•
Two scintillation mitigation strategies, screening and weighting, have
been investigated
•
Both could provide clear positive improvement to the positioning
performance
27
THANK YOU

An ionospheric scintillation mitigation method for RTK positioning at

Transcrição

Documentos relacionados

A filtering method developed to improve GNSS receiver data quality

REGISTRO DE DELEGADOS INTERNACIONALES Quito