Separating the Variances of a Two-Component Clock Model by Sequential Minque

Charles Greenhall

Abstract:	Minimum norm quadratic unbiased estimation, or MINQUE, is a method for improving the variance estimates of noise components in a Gauss-Markov least-squares problem. This study treats a simple special case: estimating the two noise levels of a clock whose phase noise is the sum of white FM and random walk FM. Given prior estimates of the noise levels, perhaps from an Allan deviation plot, MINQUE calculates new estimates and their uncertainties. Although the original MINQUE calculation on N data takes O(N2) space and O(N3) time, it can be done sequentially in bounded space and O(N) time. The method is applied to data from a simulation and from a comparison of two hydrogen masers.
Published in:	Proceedings of the 40th Annual Precise Time and Time Interval Systems and Applications Meeting December 1 - 4, 2008 Hyatt Regency Reston Town Center Reston, Virginia
Pages:	275 - 286
Cite this article:	Greenhall, Charles, "Separating the Variances of a Two-Component Clock Model by Sequential Minque," Proceedings of the 40th Annual Precise Time and Time Interval Systems and Applications Meeting, Reston, Virginia, December 2008, pp. 275-286.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In