An Enhanced Frequency Estimation Algorithm Using a Three-Point Spectral Interpolation Method

Abstract

Frequency estimation of sinusoidal signals is a critical task in various signal processing applications, including control systems, monitoring, radio broadcasts, and more. The Fast Fourier Transform (FFT) is a widely employed technique for signal analysis; however, it suffers from spectral leakage issues. To mitigate this problem, windowing functions are utilized, aiming to enhance frequency estimation accuracy through the combination of an optimal window and a precise frequency correction formula. In this study, a novel frequency estimation algorithm based on a three-point spectral interpolation method is proposed and compared with the Jacobsen algorithm. Simulation results demonstrate that the proposed algorithm exhibits superior performance in terms of frequency estimation errors. Specifically, the maximum frequency estimation error for the proposed algorithm, when using the Nuttall window, was found to be 0.001, representing a 29-fold reduction compared to the error of 0.029 for the Jacobsen algorithm. This improvement highlights the effectiveness of the proposed interpolation-based algorithm for accurate frequency estimation in signal processing applications.