Novel Interpolation Method of Multi-DFT-Bins for Frequency Estimation of Signal With Parameter Step Change

Abstract

The interpolation discrete Fourier transform (IpDFT) method is one of the most commonly used nonparametric methods. However, when a parameter (frequency, amplitude, or phase) step changes in the DFT period, the DFT coefficients will be distorted seriously, resulting in the large estimation error of the IpDFT method. Hence, it is a key challenge to find an IpDFT method, which not only can eliminate the effect of the step-changed symbol but also can sufficiently eliminate the fence effect and the spectrum leakage. In this article, an IpDFT-based method is proposed to estimate the frequency of the single-tone signal with the step-changed parameters in the sampling signal sequence. The relationship between the DFT bins and the step-changed parameters is given by several linear equations. At most six different DFT bins are used to eliminate the effect of the symbol.