Correlation Structure of Ionospheric Estimation and Correction for WAAS

Andrew Hansen, Eric Peterson, Todd Walter and Per Enge

Abstract:	The ionosphere is a distributed medium that causes dispersive delay in the propagation of transionospheric radiowaves such as GPS signals. One of the fundamental components of the Federal Aviation Administration's Wide Area Augmentation System for GPS is the estimation and correction of this geographically varying delay. Because GPS is based on satellite ranging determined by propagation time, any uncorrected advance or delay in the propagation time such as residual ionospheric delay represents an error source in the resulting differential position solution. If this residual delay exceeds the confidence bound given on the range correction the error may constitute a threat to the navigation user. Paramount in aviation applications using WAAS for navigation guidance is the protection of the position solution, whereby we mean the containment of the position error within the confidence bound in the position domain. One approach to protecting the user from residual error in the differential navigation solution is to inflate the confidence bound on the differential range corrections which in turn inflates the confidence bound in the position domain. The penalty for taking this approach however is quite high because the system may only be used for guidance if the confidence bound on the position, which we call Horizontal and Vertical Protection Levels (HPL and VPL), are smaller than the corresponding Horizontal and Vertical Alert Limits (HAL and VAL) for any given phase of flight. In aviation applications of WAAS, the ionospheric corrections are only required for the final approach phase of flight when Instrument Precision approach with Vertical guidance (IPV) or Category I precision approach (CATI) conditions apply. These conditions have the most stringent requirements specified in the WAAS Minimum Operational Performance Standard (WAAS MOPS) and are very challenging to achieve. Thus the inflate approach is not satisfactory. One approach to protecting the user from residual error in the differential navigation solution is to inflate the confidence bound on the differential range corrections which in turn inflates the confidence bound in the position domain. A better approach is to consider the correlation structure of the ionospheric measurements taken by the WAAS reference network and in turn the correlation structure of the residual error after the resulting ionospheric correction is applied to the user's ranging measurements. With such a characterization, the confidence bounds on the ionospheric corrections can be minimized to increase availability and yet still protect the user's navigation solution. We undertake the characterization process with real measurements from the WAAS reference station network over North America. We have utilized both ionospheric models and real measurements to build an additive correlation model of the ionosphere which can be used in the estimation process for constructing WAAS ionospheric corrections. This correlation model is critical to the ionospheric estimator as part of the covariance propagation we use to determine the ionospheric correction confidence bounds or so-called Grid Ionospheric Vertical Error (GIVE). The paper includes a description of this model, its implementation in the correction algorithm, and an analysis of the performance achieved by the overall WAAS correction in terms of accuracy, integrity, and availability.
Published in:	Proceedings of the 2000 National Technical Meeting of The Institute of Navigation January 26 - 28, 2000 Pacific Hotel Disneyland Anaheim, CA
Pages:	454 - 463
Cite this article:	Hansen, Andrew, Peterson, Eric, Walter, Todd, Enge, Per, "Correlation Structure of Ionospheric Estimation and Correction for WAAS," Proceedings of the 2000 National Technical Meeting of The Institute of Navigation, Anaheim, CA, January 2000, pp. 454-463.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In