An STD Reliability Test for Kinematic GPS Processing: A Case Study Based on KARS and Applanix POSGPS Software

Naci Yastikli, Charles Toth, and Dorota Grejner-Brzezinska

Abstract:	Practical experiences have shown that the standard deviations (STD) obtained from kinematic GPS processing software may not always reflect the actual error due to the lack of information about functional correlation. Neglecting the physical correlation between epochs, systematic errors and improperly modeled parameters may cause incorrect estimation of the standard deviation. A number of the GPS processing software consider the variance-covariance matrix of the observations without correlation or simply deal with the diagonal components only. As a consequence, the a posteriori standard deviations of the estimated coordinates could be too optimistic and may not represent the actual quality of the estimated coordinates. This is particularly important when the improperly estimated STDs are used as a priori values for GPS/IMU (Global Positioning System/Inertial measurement unit) Kalman filtering, and consequently defining the georeferencing error budget of imaging sensors, such as airborne digital camera, LiDAR (Light Detection and Ranging) and IfSAR (Interferometric synthetic aperture radar). The primary objective of the research presented in this paper is to assess the STD reliability of the kinematic GPS processing using the NGS KARS and Applanix POSGPS software packages. In this research, the scale factor (SF) was used as a reliability indicator for software generated STDs. In order to assess the STD reliability of kinematic GPS processing, two data sets from the San Andreas and San Jacinto Faults LiDAR Mapping project were used. Since there is no easily available absolute reference for an airborne trajectory, in our investigations one base station was used as a rover, called simulated rover, to create kinematic solutions with respect to several other reference stations with varying separation to simulated rover. The two daily data sets from the LiDAR Mapping project were processed in kinematic mode using L3 (ion-free) observations. The single baseline solution from NGS KARS and Applanix POSGPS software and multi-baseline (network) solutions from Applanix POSGPS were obtained, and the scale factor was computed by comparing the software generated STD, called a formal error, to the weighted standard deviation. The reliability of STD, computed from NGS KARS and Applanix POSGPS were analyzed.
Published in:	Proceedings of the 63rd Annual Meeting of The Institute of Navigation (2007) April 23 - 25, 2007 Royal Sonesta Hotel Cambridge, MA
Pages:	321 - 326
Cite this article:	Yastikli, Naci, Toth, Charles, Grejner-Brzezinska, Dorota, "An STD Reliability Test for Kinematic GPS Processing: A Case Study Based on KARS and Applanix POSGPS Software," Proceedings of the 63rd Annual Meeting of The Institute of Navigation (2007), Cambridge, MA, April 2007, pp. 321-326.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In