一种频域插值变化迭代频率估计方法

针对现有频率估计算法存在的复杂度高、频率估计能力弱、估计结果均方差大等缺点，在固定迭代AM（Aboutanios—Mulgrew）无偏频率估计算法基础上，提出一种频域插值变化迭代频率估计算法，推导了不同迭代参数实现无偏估计的充分条件，证明了有偏估计时本算法的收敛性和偏离度，通过设置不同迭代参数，可以实现无偏或有偏估计。仿真分析表明：当具有较高信噪比时，在整个频率估计范围内，该方法均方误差接近CRLB（Cramer-RaoLowerBound，克拉美一罗下限）；当FFT（FastFourierTransform，快速傅里叶变换）粗估计残余频率接近0．5时，该方法的均方误差优于CRLB，为CRLB的96％。

A Frequency Estimator with Variable Iteration by Interpolation in Frequency Domain

Based on fixed-iterative AM （Aboutanios Mulgrew） non-bias frequency estimator algorithm, a frequency estimator with variable iteration by interpolation in frequency domain is proposed as existing frequency estimators have drawbacks such as high complexity, poor frequency estimate performance and significant rms deviation of esti- mate results. Sufficient conditions for unbiased estimation are derived for different iteration parameters. The conver- gence and bias of biased estimation are demonstrated. The estimator can be unbiased or biased by setting different it eration parameters. Simulation results show that the Mean Square Error （MSE） of the proposed estimator is close to CRLB （Cramer-Rao Lower Bound） over the frequency estimation range when signal-to-noise ratio is high. The MSE of the proposed estimator is 0.96 times CRLB when the residual frequency after FFT （Fast Fourier Transform）- based estimation is near 0.5.