Loran-C User Position Software (LUPS) Navigation Performance with the August 2001 Cross-Country/Alaska Flight Test Data

摘要

This paper presents the results of post-processing the August 2001 cross-country/Alaska flight test datarnusing the Loran-C User Position Software (LUPS) developed at Illgen, showing the achieved accuracy,rnintegrity, continuity, and availability with respect to Required Navigation Performance (RNP) thresholds.rnThe truth references were the Global Positioning System (GPS) based positions collected during the flights.rnThe achieved performance satisfies RNP-0.3 and RNP-0.5 level requirements in eastern and westernrnCONUS, respectively. The aggregate performance for flight segments east of the 105oW meridian gave arn(Root-Mean-Square, 95%, maximum) horizontal error of (145, 234, 337) meters during 12 hours of normalrnreceiver operation. Using RNP-0.3 thresholds the eastern segments had no Hazardously MisleadingrnInformation (HMI), 0.3% Misleading Information (MI), 99.99% continuity, and 99.99% availability ofrnformal accuracy and fault-free integrity. The aggregate performance for western flight segments (betweenrnlongitudes 123.o75W and 105oW) had a (RMS, 95%, maximum) error of (305, 630, 854) meters during 7rnhours of normal receiver operation. Using RNP-0.5 thresholds the western segments had no HMI, 1.2% MI,rn99.85% continuity, and 99.64% availability of accuracy and fault-free integrity. The significantly poorerrnperformance in the west is attributed to weaker geometry and fewer valid signals. For the (eastern, western)rnpaths the average Horizontal Dilution of Precision (HDOP) was (0.5, 1.1) and the average number of validrnsignals was (18, 9).rnA special feature of the reported processing is the use of master-clock-to-UTC synchronization error of 330rnmeters (1-sigma) in the estimation weighting and error propagation, based on observed differences betweenrntime-of-arrival (TOA) measurements of dual-rated signals. Tests using a 1-sigma synchronization error ofrn30 meters (based on Loran-C signal specifications) gave dramatically poorer results. The inter-chainrnsynchronization errors found were ten times larger than those reported by the U.S. Naval Observatoryrn(USNO) for the period of the flights. Tests using chain-timing corrections from the U.S. Coast GuardrnNavigation Center (USCG-NAVCEN) Time of Transmission Monitor (TTM) and a synchronizationrnstandard error of 50 meters gave accuracy and fault-free integrity results very close to those usingrnuncorrected chain-timing and standard error of 330 meters. However, the much stronger formal geometryrnof the case with corrected chain-timing yields much greater fault-mode integrity compared to the case withrnuncorrected timing.rnThe LUPS ground conductivity database was extended to Alaska using a constant conductivity value of 2rnmS/m. The resulting 95% navigation error for the Alaskan flights was 881 meters with an average of 7rnvalid signals. The low availability of acceptable solutions and high number of HMI and MI cases foundrnsuggest the need for further investigation into the quality of data and modeling in this area.