1 / 50

Wide Area Augmentation System (WAAS) Operations Team AJW-1921

Offline Monitoring. Wide Area Augmentation System (WAAS) Operations Team AJW-1921. WIPP. B. J. Potter Brad Dworak Chad Sherrell March 7, 2011. Introduction. This presentation covers the 4th Quarter of 2010 (2010-10-01 – 2010-12-31) Next Steps Analyze data for entire quarter

nairi
Télécharger la présentation

Wide Area Augmentation System (WAAS) Operations Team AJW-1921

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Offline Monitoring Wide Area Augmentation System (WAAS) Operations TeamAJW-1921 WIPP B. J. Potter Brad Dworak Chad Sherrell March 7, 2011

  2. Introduction • This presentation covers the 4th Quarter of 2010 • (2010-10-01 – 2010-12-31) • Next Steps • Analyze data for entire quarter • Transition all OLM analysis to SOS

  3. Clock Runoff Assertion The a priori probability of a GPS satellite failure resulting in a rapid change in the GPS clock correction is less than 1.0x10-4 per satellite. Monitoring Approach Events typically result in a fast correction that exceeds 256 meters When this occurs, the satellite is set Do Not Use until the correction reaches a reasonable size Events where the satellite is set Do Not Use from excessively large fast corrections while the satellite is healthy are recorded

  4. Clock Runoff No Clock Runoff Events between 2010-10-01 – 2010-12-31

  5. Ephemeris • Assertion The CDF of GPS ephemeris errors in a Height, Cross-track, and Along-track (HCL) coordinate system is bounded by the CDF of a zero-mean Gaussian distribution along each axis whose standard deviations are osp-ephh, osp-ephc, andosp-ephl.The probability that a satellite’s position error is not characterized by this a priori ephemeris model is less than 10-4 per hour. • Monitoring Approach • Compare broadcast vs precise in HCL to ensure sigmas are less than 1m, 2.5m, 7.5m for Radial, Cross Track, and In Track

  6. Ephemeris – Radial PRN: 31 2010-11-20 03:45:00 1.44416876193 PRN: 29 2010-12-16 02:45:00 -1.28223089084 PRN: 05 2010-10-23 23:45:00 -1.25456410331 PRN: 25 2010-12-23 04:45:00 -1.40505207557 PRN: 25 2010-12-24 06:00:00 -1.29037947749 PRN: 11 2010-10-15 00:00:00 -2.16619174736

  7. Ephemeris – In Track PRN: 11 2010-10-15 04:00:00 14.0549074655 PRN: 25 2010-12-24 07:45:00 14.6629813657 PRN: 27 2010-11-07 20:15:00 8.39537775327 PRN: 30 2010-12-30 04:15:00 8.86779042477 PRN: 24 2010-12-30 01:30:00 -15.7248821707

  8. Ephemeris – Cross Track PRN: 17 2010-12-01 06:45:00 -2.73555408015 PRN: 21 2010-12-31 00:00:00 -2.57918952043

  9. RIC Outliers

  10. Ionospheric Threat Model Monitoring • Assertion The values of and iono adequately protect against worst case undersampled ionosphere over the life of any ionospheric correction message, when the storm detectors have not tripped. • Monitoring Approach • Monitor for Chi^2 values greater than 1 in the four regions • CONUS > 1% • Alaska > 2% • Caribbean > 10% • Other > 3%

  11. Monitoring Regions

  12. Total Chi2 values ≥ 1 from all regions for 2010 at ZLA (35.07% zeros)

  13. 2010 Total Chi2 Values Over 1

  14. Antenna Monitoring • Assertion The position error (RSS) for each WAAS reference station antenna is 10cm or less when measured relative to the ITRF datum for any given epoch. (Mexico City is allowed 25cm). The ITRF datum version (realization) is the one consistent with WGS-84 and also used for positions of the GPS Operational Control Segment monitoring stations.

  15. Purpose • Accurate antenna positions needed to support DGPS applications • Correct for Time Dependent Process • Tectonic Plate Movement • Subsidence • Correct for Shift Events • Seismic • Maintenance • WIPP Review for integrity issues • Greater than 10 cm WIPP should review • Greater than 25 cm WIPP must review • Special case for Mexico City (25 cm for review) • Project the need for a WAAS Antenna Coordinate Update

  16. Survey Details • Survey Date • 2011-01-28 • Cross Compared Against • CSRS-PPP • WFO-R2 • Coordinates Projected to six months beyond WFO WFO Release 3 • 2012-05-01

  17. Results • Against CSRS-PPP • All sites less than 5 cm. • Against WFO-R2 • All sites less than 5 cm.

  18. Code Carrier Coherence • Assertion The a priori probability of a CCC failure is less than 1x10-4 per set of satellites in view per hour for GPS satellites and 1.14x10-4 for GEO satellites.

  19. CCC monitoring approach • Anik, Galaxy 15 and all GPS satellites are monitored for CCC trips for Q4 2010 (last data for CCC data for Galaxy 15 was on 2010-12-15). • AMR is not currently monitored (not used as ranging source, UDRE floor=50m) • All CCC monitor trips are investigated whenever a trip occurs to determine source of trip • Minimum data sources used in correlation and analysis: • CCC test statistic • UDRE threshold value • CMCI measurements from NETS SQA • WAAS Iono calculation • L1/L5 Iono GUST calculation • published planetary Kp and Ap values • Chi2 values

  20. Reported CCC trips for Q4 2010 Date GEO PRN C&V 2010-10-24 06:38:25 138 ZLA 2010-10-24 06:38:30 138 ZDC 2010-10-25 19:23:54 138 ZTL 2010-10-25 19:24:05 138 ZLA 2010-12-08 11:55:40 138 ZDC ZTL 2010-12-08 14:03:56 138 ZDC ZLA 2010-12-11 22:55:44 138 ZDC ZTL 2010-12-12 02:45:52 138 ZDC ZTL 2010-12-12 15:24:27 138 ZDC ZTL

  21. CCC plots

  22. CCC plots

  23. Signal Quality Monitor Assertion The a priori probability of a signal deformation (SD) failure is less than 2.4x10-5 per set of satellites in view per hour for GPS or GEO satellites. The worst-case range errors due to nominal signal deformations are more than 25cm on any satellite signal relative to the other satellites in view. Monitoring Approach All SQM Trips will be monitored for and investigated Max and Median data for each metric will be plotted by Requested UDRE Monitoring for discrepancies between satellite Plots are for the first 4 days of every week for the entire quarter Plots were made using the tools from HMI Build 299

  24. SQM max plot

  25. GEO Signal Quality • Assertion The WAAS SIS satisfies the requirements for code-carrier coherence and fractional coherence stated in sections 3.1.4.2 and 3.1.4.3 of the [draft] system specification FAA-E-2892c • Monitoring Approach • Collect WAAS SIS data from each GEO using GUST receivers connected to dish antennas • Compute and plot the metrics outlined in sections 3.1.4.2 and 3.1.4.3 of FAA-E-2892c • Examine plots, tabulate max metric values and pass/fail states, analyze failures in further detail to identify possible causes

  26. Performance Summary

  27. Summary • Only issue: fly-by of CRW apparently affected measurements of both CRE and CRW • Twice-daily elevated noise on some days • When CRW most N and S; bigger effect as CRE, CRW got close • Different sites saw effects of different magnitude • Additional periods of noise on Nov 12 (closest approach) • PR oscillation correction now included in processing • Mitigates systematic error in GUST receiver • L1 and L5 PR corrections mapped using GUST receiver and prototype SIGGEN in Zeta lab • Applicable to any WAAS GUST/G-II receiver • Dependent on PRN code, PR(t) and PR(t-1) • Allows more accurate evaluation of received signal

  28. Example of Elevated Noise: 19 Nov 2010 OKC PRN 135 LTN PRN 135 • Elevated noise at approx. 2:21 and 14:19 UTC (N and S extremes of CRW orbit) -- close to but not at zero Doppler at either OKC or LTN (or APC Primary, which showed little to no effect at either time this day) caused higher than normal max CCC values • Effect worse when CRW was close to CRE (effect also seen on CRE) • Since different sites show different effects, probably not on SIS; will monitor as CRW returns to nominal orbit slot

  29. Example of PR Correction Effect: 19 Nov 2010 OKC PRN 138 before correction OKC PRN 138 after correction • PR oscillation correction generally benefits L1 more than L5 and PRN 138 more than PRN 135 (oscillation signatures have different magnitudes) • CRE performance marginal without correction but well below spec limits with correction (nominally; not including CRW fly-by effects) • Since oscillations are a systematic receiver effect, mitigation allows better evaluation of received signal

  30. PRN 135 Short-term CCC CRW passes CRE Nov 12 Note: missing values indicate days with switchovers or incomplete data

  31. PRN 135 Long-term CCC Note: missing values indicate days with switchovers or incomplete data

  32. PRN 135 Short-term CC Note: missing values indicate days with switchovers or incomplete data

  33. PRN 135 Long-term CC Note: missing values indicate days with switchovers or incomplete data

  34. PRN 138 Short-term CCC Note: missing values indicate days with switchovers or incomplete data

  35. PRN 138 Long-term CCC Note: missing values indicate days with switchovers or incomplete data

  36. PRN 138 Short-term CC Note: missing values indicate days with switchovers or incomplete data

  37. PRN 138 Long-term CC Note: missing values indicate days with switchovers or incomplete data

  38. Code Noise and Multipath (CNMP) Overbounding • Assertion • The Code Noise and Multipath (CNMP) error bound is sufficiently conservative such that the error in linear combinations of L1 and L2 measurements is overbounded by a Gaussian distribution with a sigma described by the Root Sum Square (RSS) of L1 and L2 CNMP error bounds except for biases, which are handled separately.3 • Monitoring Approach • Bounding for L1, IFPR, Delay • Aggregate and WRE Slices • All bounding failures analyzed in further detail

  39. Equations Used • Cumulative distribution function (CDF): • For examining the behavior at larger values of x: • Pass is Δx > 0 for all |x|>0.25

  40. Aggregate Plot of CNMP Delay

  41. Aggregate Plot of CNMP IFPR

  42. Aggregate Plot of CNMP RDL1

  43. CNMP Tabular Results from Poor Performing WRE Slices *This is a subset of sites as an example

  44. Summary • Quarterly monitoring results continue to support specific assertions called for in the HMI document. • All antenna positions are within 5 cm. • The CCC Test Statistic for the GEOs is ????

  45. Offline Monitoring Document • Report format is separated into 3 hierarchical reading levels: • Level 1: Executive summary • 2-3 page overview of the events • Level 2: Main body • ~30 pages of technical briefings, limited number of graphs • Level 3: Materials and Methods • Supplemental information, including: • Additional Figures • Details of the tool configuration (build no, flag settings, etc.) • Data filenames and location (to possibly re-run in the future) • OLM coding standards and guidelines • First draft is scheduled to be released on March 31st

  46. Offline MonitoringData Types and Standards • Standards: • Slicing requirements – data from different sources are examined separately and not aggregated • UDRE index • PRN • Binning requirements – different bin sizes are used for different analyses (0.01, 0.001, etc.) • 4 File Formats: • (1) Histogram files – histogram of raw counts of the metric (not probabilities), can be compiled together

  47. Offline MonitoringData Types and Standards • (2) Statistics files – each column in histogram file has a list (rows) of 15 descriptive statistics associated with it: • Counts • Mean • Standard deviation • Minimum, Maximum, Absolute maximum • Sigma over-bound (zero centered), Sigma over-bound (mean centered) • 1st quartile, Median, 3rd quartile • Mean and standard deviation of absolute value • RMS • Variance • (3) Time series files: variable data over time • Time represented in WAAS time, UTC time (HHMMSS) and seconds into the day • Files can be concatenated together to form multi-day sets • (4) Quantity files: two-dimensional slices of any particular quantity (ex. UDREI/GPS PRN of |CCC metric|)

More Related