NAVSTAR performance analysis

Description

The two dimensional, real time navigational accuracy available from NAVSTAR/GPS is investigated, with emphasis on GPS performance during the partially implemented phase (1980 to 1987), and on the ocean areas surrounding Canada. The mathematical models used in this study to compute NAVSTAR satellite positions, check for satellite visibility, and compute the covariance matric of two dimensional position are described. The length of the semi major axis of the standard error ellipse represented by this covariance matrix is used as the GPS performance indicator. A computer program implementing these models is listed. Results based on four error models are presented. These error models are P-code ranging only (with 4 metre tang errors), C/A-code ranging only (16 metre range errors), P-code assisted by Loran-C and C/A-code assisted by Loran-C. For this analysis Loran-C range to Cape Race, Angissoq and Sandur were assumed to have standard deviations of 140 m. In all cases it was assumed that the user’s clock could be kept synchronized to GPS time independently from the GPS measurements, to within 0.3 microseconds. Based on these assumptions, it was found that in the Davis Strait area, combined P-code GPS and Loran-C should provide 150 metre positioning about 11 hours per day, with the present (1980) orbital configuration of six GPS satellites. GPS performance with only six satellites is also a function of both latitude and longitude. In general high latitude (60 ° and above) have poorer performance. The performance at low and middle latitude depends on the relationship between the observer’s meridian and the meridians travelled by the GPS satellite subtracks. Plots of the variation in the length of the GPS error ellipse semi major axis over a 24 hours period are presented for 14 different locations. Compete output listings are presented for three locations and several error models. A full GPS constellation of 24 satellites was simulated. The two dimensional, P-code positioning accuracy in Davis Strait, using all visible satellites was uniformly of the order of five metres. It is concluded that it is feasibly to use GPS in its present limited deployment as an operational survey positioning systems in the eastern Canadian artic, provided the requirements of the survey are met by 150 metres or better positioning for about 11 hours a day.