Frequency estimation via weighted multipoint interpolated DFT

Abstract

A Weighted multipoint interpolated discrete fourier transform (WMIpDFT) method for high accuracy frequency estimation of a multi-frequency signal component is presented. This method uses the maximum sidelobe decay windows. The relationships to estimate the frequency when an H-term maximum sidelobe decay window (H≥2) is used for (2J+1)-point interpolation (1≤J≤H+1) are given. The effectiveness of the method proposed was analysed by means of computer simulations for a multi-frequency signal without noise as well as with quantisation noise.