Fuzzy interpolation of the average signal steps

Abstract

In this paper is proposed a fuzzy interpolation method of the average signal steps in each processing stage, for extraction of signal drowned in noise. The fuzzy interpolation method increases the Signal-to-Noise-Ratio (SNR) gain for a periodic signal drowned in noise, and may gives good results for different other signal processing applications, such as: extraction of periodic signals combination drowned in noise, signal shape reconstruction etc. Recommended sampling frequency is up to Ns times of frequency given by the Shannon's condition, where number of signal samples on one time stage, Ns, is usually the order of hundreds or thousands. The simulation and experimental results obtained with periodic signals drowned in noise are given using the Matlab© and a digital signal processing (DSP) platform, respectively. The proposed filtering method is compared with other similar methods by computing the SNR gain.