GPS Satellite-to-User Range Accuracies: A Calibration Experiment

Abstract

From 20 October to 17 November 1986, three Contracting and three Government institutions took part in a cooperative evaluation of the accuracy of the Global Positioning System (GPS). In effect, acting as independent GPS users employing CIA-code single-frequency or P-code dual-frequency receivers, each organization computed satellite-to-user range prediction errors.Each institutional user computed statistical summaries of these User Range Errors (URE). The root-mean-square (rms) of the URE's across all satellites, and over all 30-days of the test, as reported by each user ranged from a high of 4.8 meters to a low of 3.1 meters. The C/A-code users experienced the largest rms URE. The GPS is required to provide 6 meter rms URE.For the same test period, the GPS Operational Control System (OCS) computed URE performance measures of 2.0 meters (rms) and 2.5 meters (rms). This suggests that these GPS performance measures might be improved.Throughout the 30 day test, users reported that the daily rms URE were approximately steady-state. However, just prior to the test, the rms URE did appear correlated over several days with rms values near 6 meters. This URE degradation appeared to improve as NAVSTARS 3,6, and 10 were less and less eclipsed by the Earth shadow.

The Global Positioning System Geodesy Odyssey

“…A 30-day evaluation of OCS accuracy for Block I satellites was conducted in 1986 with UREs of 3 to 5 m estimated independently by several organizations [22]. The WGS 84 gravity field model replaced the WGS 72 model in 1987.…”