Determination of electrical signal masked by random noise using "super-averaged" functions

Electrical and electromagnetic methods often require the determination of the amplitude of a periodic signal, produced by controlled sources and masked by natural and artificial noise. Since noise is usually random and uncorrelated, several techniques based on a stacking procedure and spectral decomposition are applied. In this paper we revise some of these procedures based on Fourier analysis. We propose a technique which uses "super-averaged" functions obtained from the average of the results of the Fourier analysis from the whole data set.