Mitigating Ionospheric Threat Using a Dense Monitoring Network

1
ION GNSS 2007 Fort Worth, TX Sept. 25-28, 2007
Mitigating Ionospheric Threat Using a Dense
Monitoring Network
T. Sakai, K. Matsunaga, K. Hoshinoo, K. Ito,
ENRI T. Walter, Stanford University
2
Introduction

• The ionospheric effect is a major error source
  for SBAS
• The ionospheric term is dominant factor of
  protection levels
• Necessary to reduce GIVE values not only in the
  storm condition but also in the nominal condition
  to improve availability of vertical guidance.
• The problem is caused by less density of IPP
  samples
• The current planar fit algorithm needs inflation
  factor (Rirreg) and undersampled threat model to
  ensure overbounding residual error
• Solution integrating the external network such
  as GEONET and CORS
• Developed a GIVE algorithm suitable to such a
  situation.
• Evaluated a new GIVE algorithm with GEONET
• 100 availability of APV-II (VAL20m) at most of
  Japanese Airports
• Still protects users No HMI condition found.

3
MSAS Status

• All facilities installed
• 2 GEOs MTSAT-1R (PRN 129) and MTSAT-2 (PRN 137)
  on orbit
• 4 GMSs and 2 RMSs connected with 2 MCSs
• IOC WAAS software with localization.
• Successfully certified for aviation use
• Broadcast test signal since summer 2005 with
  Message Type 0
• Certification activities Fall 2006 to Spring
  2007.
• Began IOC service on Sept. 27 JST (1500 Sept. 26
  UTC).

Launch of MTSAT-1R (Photo RSC)
4
Position Accuracy
_at_Takayama (940058) 05/11/14 to 16 PRN129
_at_Takayama (940058) 05/11/14 to 16 PRN129
GPS
GPS
MSAS
MSAS
Horizontal RMS 0.50m MAX 4.87m
Vertical RMS 0.73m MAX 3.70m
5
Concerns for MSAS

• The current MSAS is built on the IOC WAAS
• As the first satellite navigation system
  developed by Japan, the design tends to be
  conservative
• The primary purpose is providing horizontal
  navigation means to aviation users Ionopsheric
  corrections may not be used
• Achieves 100 availability of Enroute to NPA
  flight modes.

• The major concern for vertical guidance is
  ionosphere
• The ionospheric term is dominant factor of
  protection levels
• Necessary to reduce GIVE to provide vertical
  guidance with reasonable availability.

6
APV-I Availability of IOC MSAS
MSAS Broadcast 06/10/17 0000-2400 PRN129
(MTSAT-1R) Test Signal Contour plot for APV-I
Availability HAL 40m VAL 50m Note
100 availability of Enroute through NPA
flight modes.
7
Components of VPL
VPL
Ionosphere (5.33 sUIRE)
Clock Orbit (5.33 sflt)
MSAS Broadcast 06/10/17 0000-1200 3011
Tokyo PRN129 (MTSAT-1R) Test Signal

• The ionospheric term is dominant component of
  Vertical Protection Level.

8
Problem Less Density of IPP

• Ionospheric component GIVE
• Uncertainty of estimated vertical ionospheric
  delay
• Broadcast as 4-bit GIVEI index.
• Current algorithm Planar Fit
• Vertical delay is estimated as parameters of
  planar ionosphere model
• GIVE is computed based on the formal variance of
  the estimation.
• The formal variance is inflated by
• Rirreg Inflation factor based on chi-square
  statistics handling the worst case that the
  distribution of true residual errors is not
  well-sampled a function of the number of IPPs
  Rirreg 2.38 for 30 IPPs
• Undersampled threat model Margin for threat that
  the significant structure of ionosphere is not
  captured by IPP samples a function of spatial
  distribution (weighted centroid) of available
  IPPs.

9
Using External Network

• Integrating the external network to the SBAS
• Increase the number of monitor stations and IPP
  observations dramatically at very low cost
• Just for ionospheric correction Clock and orbit
  corrections are still generated by internal
  monitor stations because the current
  configuration is enough for these corrections
• Input raw observations OR computed ionospheric
  delay and GIVE from the external network
  loosely-coupled systems.
• Necessary modifications
• A new algorithm to compute vertical ionospheric
  delay and/or GIVE is necessary because of a great
  number of observations
• Safety switch to the current planar fit with
  internal monitor stations when the external
  network is not available.

10
Available Network GEONET

• GEONET (GPS Earth Observation Network)
• Operated by Geographical Survey Institute of
  Japan
• Near 1200 stations all over Japan
• 20-30 km separation on average.
• Open to public
• 30-second sampled archive is available as RINEX
  files.
• Realtime connection
• All stations have realtime datalink to GSI
• Realtime raw data stream is available via some
  data providers.

GEONET station
MSAS station
11
Sample IPP Distribution

• A snap shot of all IPPs observed at all GEONET
  stations at an epoch
• GEONET offers a great density of IPP
  observations
• There are some Japan-shape IPP clusters each
  cluster is corresponding to the associated
  satellite.

12
New Algorithms

• (1) Residual Bounding
• An algorithm to compute GIVE for given vertical
  delays at IGPs
• Vertical delays are given For example, generated
  by planar fit
• Determine GIVE based on observed residuals at
  IPPs located within 5 degrees from the IGP Not
  on the formal variance of estimation
• Improves availability of the system.
• (2) Residual Optimization
• An algorithm to optimize vertical delays at IGPs
• Here Optimum means the condition that sum
  square of residuals is minimized
• GIVE values are generated by residual bounding
• Improves accuracy of the system.

13
Residual Bounding (1)

• An algorithm to compute GIVE for given vertical
  delays at IGPs
• The MCS knows ionospheric correction function
  (bilinear interpolation) used in user receivers,
  Iv,broadcast(l,f), for given vertical delays at
  IGPs broadcast by the MCS itself
• Residual error between the function and each
  observed delay at IPP, Iv,IPPi, can be computed
• Determine GIVE based on the maximum of residuals
  at IPPs located within 5 degrees from the IGP.

Vertical delay for user
Observed delay at IPP
14
Residual Bounding (2)

• Determine GIVE based on the maximum of residuals
  at IPPs located within 5 degrees from the IGP.

15
Residual Optimization

• An algorithm to optimize vertical delays at IGPs
• Vertical delays at IGPs can also be computed
  based on IPP observations as well as GIVE values
• Again, define residual error between the user
  interpolation function and each observed delay at
  IPP, Iv,IPPi
• The optimum set of vertical delays minimizes the
  sum square of residuals GIVE values are
  minimized simultaneously
• The optimization can be achieved by minimizing
  the energy function (often called as cost
  function) following over IGP delays (See paper)

Function of IGP delays
16
Number of Available IPPs

• The histogram of the number of IPPs available at
  each IGP (located within 5 deg from the IGP)
• For 68 cases, 100 or more IPPs are available
• Exceeds 1000 for 27 cases.

17
GIVE by Residual Bounding (1)
Planar Fit
Residual Bounding (All GEONET sites)

• Histogram of computed GIVE values in typical
  ionospheric condition for two algorithms
• Residual bounding with GEONET offers
  significantly reduced GIVE values
• Blue lines indicate quantization steps for GIVEI.

18
GIVE by Residual Bounding (2)
Planar Fit
Residual Bounding (All GEONET sites)

• Histogram of computed GIVE values in severe storm
  condition for two algorithms
• The result is not so different from case of
  typical condition.

19
Reduction of GIVEI
Planar Fit
Residual Bounding (All GEONET sites)

• Histogram of 4-bit GIVEI index broadcast to
  users
• Lower limit of GIVEI is 10 for planar fit
• Residual bounding can reduce GIVEI as well as
  GIVE values.

20
Comparison with FOC WAAS
Planar Fit (FOC WAAS)
Residual Bounding (All GEONET sites)

• FOC WAAS Dynamic Rirreg, RCM, multi-state storm
  detector, and CNMP
• GIVE values derived by residual bounding are
  still smaller than FOC WAAS algorithms.

21
Residual Optimization

• Histogram of difference of IGP delays with and
  without residual optimization
• Adjustment of IGP delay stays 0.052m
• In comparison with quantization step of 0.125m,
  the effect is little.

22
User Position Accuracy
Planar Fit (RMS 1.47m)
Residual Bounding (RMS 1.10m)
Residual Optimization (RMS 1.10m)

• User vertical position error at Tokyo in typical
  ionospheric condition
• Residual bounding improves user position
  accuracy, while residual optimization is not
  effective so much.

23
Evaluation by Prototype SBAS

• Prototype SBAS software developed by ENRI (NTM
  2006)
• Computer software running on PC or UNIX
• Generates the complete 250-bit SBAS messages
  every seconds
• Simulates MSAS performance with user receiver
  simulator
• Available as an MSAS testbed Measures benefit of
  additional monitor stations and evaluates new
  candidate algorithms.
• Integration with the proposed algorithms
• Scenario of vertical ionospheric delay and GIVE
  is generated based on GEONET archive data with
  application of the proposed algorithms
• The prototype generated augmentation messages
  with ionospheric corrections induced as the
  scenario
• Tested for typical ionospheric condition (July
  2004) and severe storm condition (October 2003).

24
User Protection

• PPWAD Simulation
• 03/10/29-31
• 3011 Tokyo
• Condition
• Severe Storm
• Algorithm
• Residual Bounding
• (All GEONET sites)
• Users are still protected by this algorithm
  during the severe storm.

25
System Availability
PPWAD Simulation 04/7/22-24 Condition Typical
Ionosphere Algorithm Residual Bounding (All
GEONET sites) Contour plot for APV-II
Availability HAL 40m VAL 20m
26
Conclusion

• Introduced new algorithms and usage of the
  external network to mitigate ionospheric threats
• Algorithms for bounding ionospheric corrections
  based on optimization of residual error measured
  by dense monitoring network
• Integration of GEONET as an external network.
• Evaluation by prototype SBAS software
• Reduced GIVEI enables 100 availability of APV-II
  flight mode (VAL20m) at most of Japanese
  airports
• No integrity failure (HMI condition).
• Further investigations
• Consideration of threats against the proposed
  algorithms
• Reduction of the number of stations required for
  residual bounding
• Temporal variation and scintillation effects.

27
Announcement

• Ionospheric delay database will be available
  shortly
• The datasets used in this study and
• Recent datasets generated daily from August 2007
• Each dataset is a file which consists of slant
  delays observed at all available GEONET stations
  with 300-second interval Hardware biases of
  satellites and receivers are removed
• Access to URL
• http//www.enri.go.jp/sat/pro_eng.htm