中文摘要：

文章提出了一种适用于高斯白噪声背景下的正弦信号频率估计新算法。在给定的频率分辨率Δf下,以（m-1）fs/Δf＋1,m=1,2…M为起点进行分段,分段长度根据谱估计要求设定。当正弦信号频率fc/Δf为整数时,各分段中期望信号全相关,应用相干平均法可获得M倍信噪比增益,FFT分析后其谱极大值对应的频点即为正弦信号频率。当fc/Δf不为整数时,由于分段平均引入了相位误差,无法正确进行频率估计。因此可由大到小设定若干频率分辨率Δf,在每个Δf下按上述方法进行频率估计,当相邻Δf下的估计结果一致,或者某个Δf下估计的谱极大值远大于次极大时,即可认为频率估计完成。仿真结果表明,基于相干平均的正弦信号频率估计性能优于直接对同样长数据作FFT分析,且计算量小,硬件上更容易实现。

英文摘要：

We present a new method in subsection 1.2 of the full paper,which divides the recorded data into M parts with starting point at（m-1） fs/Δf＋1,m=1,2…M,where Δf is the frequency interval and fs is the sample frequency.It discusses two cases：（1） fs/Δf is an integer;（2） fs/Δf is not an integer.In the first case,the correlation coefficient of the signal in each segment tends to 1,and the signal to noise ratio （SNR） of coherent average has a gain of M; then the frequency corresponding to the spectrum peak is the estimated frequency. In the second ease, we adjust Δf until we get approximately the first case again and get approximately the estimated frequency. The simulation results, presented in Figs. 2 and 3, show preliminarily that frequency estimation/or sinusoidal signals using coherent average has better performance than the FFT （fast Fourier transform） method; in addition, its computation cost is small and it is easy to