ABSTRACT A spectral interpretation using the frequency sensitivity of the Allan variance (Avar) and Thêo-Hybrid (ThêoH) is used to determine fα noise, or ``power-law noise.'' ThêoH has narrower chi-square confidence than Avar; consequently, ThêoH provides significantly better determination of fα noise types at long term. Furthermore, ThêoH has even narrower confidence than chi-square. Because the algorithms used to calculate these confidence intervals are computationally intensive, we have constructed an empirical formula that approximates confidence intervals as the percent error for ThêoH.
Digital Object Identifier 10.1109/TUFFC.2007.260
© 2007, by The Institute of Electrical and Electronics Engineers, Inc. All rights reserved.