固有模态函数（IMF）积检测器——以低信噪比情况下超宽带雷达信号检测为例

An IMFs-Product Detector for the UWB Radar Signal with Low Signal-to-Noise Ratio

首次提出了一种固有模态函数积检测器。首先通过经验模式分解（EMD）把带噪信号分解成有限个固有模态函数（IMF）。检测的基本思路是，对各个IMF分量的绝对值作逐点乘积，用于抑制噪声并凸现信号，最后进行滤波和判决。本文以UWB信号为例，数据源于UWB雷达实验系统。在低信噪比（SNR），UWB脉冲与噪声波形相似，且噪声概率密度函数（PDF）未知情况下，进行实验。结果表明，当峰峰信噪比低于5dB时，该检测器性能优于Teager能量算子（TEO）。

We develop an intrinsic mode functions（IMFs） product detector for UWB radar. The UWB noisy data are decomposed into finite number of intrinsic mode functions（IMFs） by empirical mode decomposition（EMD） method. For enhancing signal peaks and suppressing noise, the detection method is based on the IMFs point-wise product. Filtering and judging are implemented in the last. The proposed method was tested in the case of low signal-to-noise ratio（SNR）, the large similarity between target signal and background noise, and absence knowledge of noise＇s probability density function（PDF）. The test data comes from the UWB radar experimental system. While peak-to-peak SNR is under 5dB, we compared the performance of the proposed detector with that of Teager energy operator（ TEO）, and showed the superiority of the proposed detector over TEO.