Adaptive Estimation of Altitude Bias in Terrain Referenced Navigation

Brian Copp and Kamesh Subbarao

Peer Reviewed

Abstract:	Terrain navigation under conditions of a time-varying altitude bias is described using an algorithm based on multiple model adaptive estimation. In this method, a bank of nonlinear filters is used, where each filter is assigned to a discrete bias hypothesis. The individual filters represent a discrete (gridded) approximation to the two-dimensional probability density for a particular bias hypothesis. The filter weights are recursively updated using Bayes rule, and position estimates are formed as a weighted sum of the filter estimates. To estimate a time-varying bias, the probability densities for each filter are periodically combined using a parametric mixture model. The mixture model represents a 'best estimate' of the probability density based on all the measurements collected to that point. This technique has the advantage that it treats altitude bias as a nuisance parameter, and is therefore not dependent on a particular bias model. Results indicate that positioning errors comparable to the Cramer-Rao lower bound can be attained even under these non-ideal conditions, and that false fixes can be avoided if a sufficient number of bias hypotheses are used.
Published in:	Proceedings of the 28th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2015) September 14 - 18, 2015 Tampa Convention Center Tampa, Florida
Pages:	2115 - 2126
Cite this article:	Copp, Brian, Subbarao, Kamesh, "Adaptive Estimation of Altitude Bias in Terrain Referenced Navigation," Proceedings of the 28th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2015), Tampa, Florida, September 2015, pp. 2115-2126.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In