| 摘要: |
| 为解决现有的正弦调频（SFM）信号参数估计方法运算复杂度高、受信噪比限制等问
题，提出了一种基于循环自相关的SFM信号参数估计新方法。首先分析了SFM信号循环自相关
函数特征，推导了信号调制频率的估计表达式；然后对信号延时相乘以去除其正弦调制特性，
得到单频信号并估计信号载频。最后，利用信号频率调制的周期性，对下变频至零频的信号
进行周期累加以减少噪声影响，通过对累加后的信号进行瞬时频率计算得到调制指数估计值
。仿真表明，信噪比（SNR）大于6 dB时，各参数估计值的均方根误差小于-18 dB
。该算法计算量较小，为同等条件下利用卡森准则（CR）方法的16%，便于工程实现。 |
| 关键词: 正弦调频信号 循环自相关 参数估计 周期累加 |
| DOI： |
|
| 基金项目: |
|
| A new sinusoidal FM signal parameters estimation algorithm based on cyclic autocorrelation |
| HUANG Hao,LI Yun-ha,ZHU Jun |
| () |
| Abstract: |
| A new approach for Sinusoidal Frequency Modulation (SFM) signal based
on cycle autocorrelation function is proposed to lower the computational complex
ity and adapt lower Signal to Noise Ratio (SNR). First, by computing the cyclic
autocorrelation function of SFM signal and analyzing its traits, the estimation for
modulated frequency is given. By delay-and-compute the SFM signal, its sinus
oidal modulation character is eliminated and it becomes a single-frequency signal
from which the carrier frequency can be estimated. Utilizing the periodicity of
the frequency modulation, the zero frequency signal after down conversion is per
iodically accumulated to decrease the effect of noise. Then, the FM coefficient
can be obtained by estimating the instantaneous frequency of the accumulated signal.
The simulation results demonstrate that when SNR≥6 dB，the Root Mean Squar
e Error(RMSE) is lower than -18 dB. The computation of the algorithm which is jus
t 16% of the one of Carson Rule (CR) under the same conditions shows that the al
gorithm is easy for engineering realization. |
| Key words: sinusoidal FM cyclic autocorrelation parameter estimation periodic accumulation |
