GPS World - January 2008 - (Page 66)

INNOVATION | Avionics & Transportation Release 6/7 Performance Fielded in September 2007. Significant availability improvement results from Mexico/Canada reference stations and ionospheric algorithm improvements. Release 5 Performance Fielded in November 2006. Includes Alaskan reference stations, third C&V, and third GEO satellite. 20m VPL 25m VPL 30m VPL 35m VPL 40m VPL 45m VPL 50m VPL 65m VPL 99% VPL LPV200 LPV 99% Vertical Protection Limit (VPL) depicted. Orange indicates the VPL ≤ 35 meters 99% of the time. An LPV approach requires a VPL less than 50 meters. Red indicates the VPL ≤ 50 meters 99% of the time. An LPV200 approach requires a VPL less than 35 meters. Black indicates the VPL > 50 meters 99% of the time. Release 5 VPL Sensitivity Release 6/7 VPL Sensitivity 100% of CONUS has 100% LPV availability 88% of CONUS has 100% LPV availability 79% of CONUS has 100% LPV200 availability 34% of CONUS has 100% LPV availability p FIGURE 6 WAAS performance summary for Release 5 (November 2006) and Release 6/7 (September 2007) as determined using shadow test systems every aspect of system performance. To facilitate testing new reference stations, data from Alaska, Mexico, and Canada were sent to Raytheon’s facilities many months before the new reference stations were cut over into the operational system. This allowed the data from the reference stations to be tested without any impact on the operational system. Any time a shadow system disagrees with the operational system or abnormal system behavior is observed, an anomaly is entered into a database and tracked until it is resolved. A WAAS build has generally been tested for four or more months on a shadow system prior to fielding. Test engineers practice all cutover procedures on the shadow systems prior to fielding a change. This rigorous testing methodology ensures that there will be no surprises when a release goes to the field. Each release is fielded 66 GPS World | January 2008 Percent of Alaska Percent of Alaska 60% of Alaska has 95% LPV200 availability 88% of Alaska has 95% LPV200 availability without interrupting WAAS service. In September 2007, a series of algorithm improvements were implemented in the system that significantly improve availability by lowering the HPL and VPL. The development of these improvements was a joint effort among the members of the WAAS Integrity Performance Panel (WIPP). The WIPP is a group of experts assembled by the FAA to provide technical oversight on the development of the WAAS algorithms. The WIPP is composed of members from Stanford University, Jet Propulsion Laboratory, Mitre Corporation, Zeta Associates, and Raytheon. The most significant of these enhancements affect the computation of the ionospheric corrections and integrity bounds. The WAAS GIVE monitor computes the ionospheric delay corrections at the WAAS grid points. Along with the correction, an integrity bound (GIVE) is calculated. The accuracy of the correction affects the accuracy of the user’s position solution. The size of the GIVE impacts system availability. Large GIVEs result in large HPLs and VPLs. When the HPL or VPL exceeds the HAL or VAL at a particular location, the service becomes unavailable. If this happens during an approach, the pilot must execute a missed-approach procedure. The GIVE monitor computes ionospheric delay estimates at a particular grid point by fitting a plane to delay measurements at the ionospheric pierce points surrounding the grid point as observed by WRSs. (A pierce point is the location of the intercept of a satellite-receiver ray path with a thin shell representation of the actual ionosphere.) The uncertainty in the fit is a function of the geometry of the pierce points, the measurement noise, and the nominal planar fit decorrelation observed during quiet ionospheric conditions. The planar model is tightly coupled with the WAAS irregularity detector, which performs a “goodness of fit” test and inflates the GIVE to 45 meters when a threshold is exceeded. A 45-meter GIVE is not useful to LPV users, so in most cases trips of the irregularity detector cause a loss of availability. To protect users from conditions when the irregularity detector is “near tripping,” the uncertainty in the planar fit is multiplied by an additional inflation factor. The irregularity detector does an excellent job of testing the planarity of the ionosphere in regions with lots of pierce points. However, many satellite geometries exist where an ionospheric storm may not be sampled by WAAS but may be sampled by a user. The WAAS undersampled threat model inflates the GIVE when the ionospheric grid point is not well sampled. The model is generated from a conservative analysis of the worst ionospheric storms witnessed during a solar maximum period (near the peak of the approximately 11-year sunspot cycle). The irregularity detector is an integral part of the model. When the irregularity detector trips, the data is pruned out of the model. Without the irregularity detector, the model would be significantly worse. A more detailed dewww.gpsworld.com Percent of CONUS Percent of CONUS http://www.gpsworld.com

Table of Contents Feed for the Digital Edition of GPS World - January 2008

GPS - January 2008 Contents Out in Front Expert Advice The Money-Go-Round u-Nav Latest Acquisition Apples to Apples Global SBAS 2008 GPS Receiver Survey Advertisers Index & Company Directory The Manufacturer's Road Year of the Who Working Indoor Up and Down Good, Better, Best Marketplace Classifieds Seen + Heard

GPS World - January 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.