Fast TheoBR: A method for long data set stability analysis

Published

September 1, 2010

Author(s)

Jennifer A. Taylor, David A. Howe

Abstract

TheoBR (Theo1 with Bias Removed) is a high-confidence statistic that evaluates frequency stability at long \tau values. However, for real-world data sets that are thousands of points and longer, the calculation of the bias with respect to the Allan Variance is very slow, taking hours, days, or even weeks on a typical personal computer. In order to make TheoBR more accessible for use on these types of data sets, a method of averaging points together within the data set is developed. The error introduced into the bias calculation by this technique is analyzed and compared to the exact bias value and a correction formula developed to minimize this error for FM noise types. Finally, this ''Fast TheoBR" technique is applied to a real data set of 223,120 points and determines stability at the longest \tau values in seconds as opposed to weeks.

Citation

IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control

Volume

57

Issue

9

Pub Type

Journals

Download Paper

Local Download

Keywords

Allan variance, flicker noise, frequency, noise, phase, stability, TheoBR, TheoH, time, white noise

Citation

Taylor, J. and Howe, D. (2010), Fast TheoBR: A method for long data set stability analysis, IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=50609 (Accessed April 25, 2024)