一种基于DFT的次优高精度频率估计算法与实现

A suboptimal accurate frequency estimator and implementation

在快速傅里叶变换(FFT)粗估计的基础上,通过曲线拟合,得到一种实现简单的次优高精度频率估计算法。现有的精确估计算法多采用FFT输出的幅度信息,或是FFT的复数输出进行精确估计。本文提出了利用幅度平方信息做精确估计的算法,有效地简化了运算复杂度,实现结构简单。通过仿真验证了本算法在低信噪比下也具有较高的估计精度。

This paper presents a suboptimal accurate frequency estimator via fitting the refined result. Previous accurate fre- quency estimators use the magnitude or complex outputs from FFT （Fast Fourier Transform）. However, the suboptimal estimator is based on the magnitude-square outputs, which can simplify the hardware implementation. The simulation results indicate that the suboptimal frequency estimator has a good performance at low signal-to-noise ratio.