Search

In many applications, including radar, radio and television, medical, industrial, and others, frequency estimate of a periodic sinusoidal signal is a crucial step in the signal processing process. Due to its simplicity of usage in digital systems, the interpolation-based signal frequency estimation algorithm is now frequently employed. Because of its performance speed, interpolation methods in the analysis rely on the fast Fourier transform (FFT). This paper proposes an approach for single-tone frequency estimate utilizing DFT interpolation with Parzen windowing in order to increase the accuracy of frequency estimation. In addition, compared to Li and Dian algorithm, the proposed method has a lower computing complexity and more steady performance. To minimize undesirable effects brought on by spectrum leakage from the FFT procedure, suitable windows have been investigated. To investigate the viability of the suggested method, three windows Flattop, Parzen, and Bohman, were applied to the simulation signal. When compared to the other windows, the Parzen window with the proposed algorithm outperformed them with a maximum frequency estimation error of 0.00003 compared to 0.0001 and 0.0002 for the Dian and Li algorithms, respectively, when the Sample Size was 8192. [ABSTRACT FROM AUTHOR]