Metrology and 1/f noise: linear regressions and confidence intervals in flicker noise context

TL;DR

This paper addresses the challenges of analyzing 1/f noise in metrology by deriving confidence intervals for mean and drift parameters, supported by an actual measurement example.

Contribution

It introduces methods for calculating confidence intervals for mean and drift in 1/f noise, enhancing metrological analysis of correlated noise.

Findings

Derived relationships for confidence intervals in 1/f noise

Application to real measurement data demonstrating the methods

Improved understanding of 1/f noise impact on measurement accuracy

Abstract

1/f noise is very common but is difficult to handle in a metrological way. After having recalled the main characteristics of stongly correlated noise, this paper will determine relationships giving confidence intervals over the arithmetic mean and the linear drift parameters. A complete example of processing of an actual measurement sequence affected by 1/f noise will be given.