A new frequency estimator of sinusoid based on interpolated FFT

A new frequency estimator of sinusoid based on interpolated FFT amplitudes is proposed. Firstly, zeros of length N are padded to the sinusoid sampled data of length N in the time domain. Next, 2N-point FFT is performed and the coarse estimation is made by searching the location of the spectrum line with maximum amplitude. Finally, the fine estimation is made by utilizing the amplitudes of the spectrum line with maximum amplitude and its two neighbors. Simulation results show that the root mean square error of frequency estimation of proposed estimator is close to the Cramer-Rao Lower Bound when the signal frequency locates anywhere between two neighboring FFT spectral lines. The proposed method has low computation complexity and outperforms the existing estimators.