Correlation analysis of experimental permittivity data

Abstract

The auto-correlation function Φ_h,h(r) of the relaxation time distribution function h(r) is obtained from the auto- and cross-correlations of real and imaginary parts of the permittivity by inversion of convolution integrals. If the permittivity data are subject to experimental error, the spectrum of the auto-correlation Φ_h,h(r) appears to be most suitable for the determination of the main characteristics of the distribution function h(r). It is necessary to obtain additional information by evaluating the lower order moments of the distribution function.