Measurement of frequency sensitivity coefficient and evaluation of type-B uncertainty for atomic frequency standard based on statistical correlation of noise

TL;DR

This paper introduces a statistical correlation method to measure frequency sensitivity coefficients and evaluate type-B uncertainty in atomic frequency standards, enabling real-time analysis during device operation.

Contribution

The paper presents a novel noise correlation approach for FSC measurement and u_B evaluation, integrating uncertainty analysis with ongoing device monitoring.

Findings

Method validated by simulations and experiments.

Allows simultaneous effect evaluation during operation.

Provides a criterion to compare with standard methods.

Abstract

Precise measurement of frequency sensitivity coefficients (FSCs) of some physical effects contributing to uncertainty is an important work in type-B uncertainty (u_B) evaluation of atomic frequency standards. We proposed a method based on statistical correlation of noise to measure FSCs and evaluate u_B for atomic frequency standards, and giving a full statistical expression of FSC as K_I=COV_A(y,x_I)/(sigma(x_I))^2 , where y and x_I are fractional frequency of atomic frequency standards and noise independent variable of Ith physical effects, respectively. The confidence value and the effect of time interval have also been discussed. Except for theoretical derivation, the method also has been validated by numerical simulation and demonstration experiments. Comparing with standard method, the statistical correlation of noise method combines u_B evaluation with type-A uncertainty (u_A), measures FSCs not by special experiments, but from the monitoring data when the device is running. It not only gives a criterion to test the standard method, but also can evaluate many effects at the same time when the device is operating. The method can be extended to other precision measurement fields and is available to comprehend the physical effectiveness of the least square method.