基于EMD的无载波超宽带雷达目标检测算法

雷达系统的检测能力与回波采样数据中所包含的目标信号的数据量有紧密联系。无载波超宽带雷达体制下,发射脉冲可以由无载波高斯型信号表示,具有周期内持续时间极短、平均功率极低的特点。所以低信噪比就成为了影响其检测性能的重点问题之一。无载波超宽带雷达体制下的回波目标检测不能够继续沿用经典的频域相关接收的检测理论,由于受到系统信号的时域瞬态特性的影响,利用经验模态分解(EMD)算法对原始回波作初步去噪,同时重构目标信号;再通过包络对齐,完成多个回波的非相干积累。仿真结果验证了这种方法的可行性。

A Target Detection Method Based on EMD for Carrier-Free UWB Radar

The detection ability of a radar system has close relationship with the target information con tained in the reflected signals. Because of the short time instant in a PRT（pulse repeat time） and the low av- erage power, a key problem affecting the performance of the carrier-free UWB（ultra wide band） radar is the detection in low SNR. Based on equivalent signal modeling, the EMD（empirical mode decomposition） meth od is used to process the raw echo data to separate the part of the noise by means of reconstructing the target signal in time domain. Then a method of envelope aligning is adopted to accumulate the multiple reflected waveforms. Simulation of the detection performance with statistical experiment proves availability of this de- tection method. The classical detection theory by correlatively receiving in frequency domain is no longer suit- able for the carrier-free UWB radar. Considering the instantaneous feature of the reflected signal, an EMD algorithm is introduced to filter the raw echoes and reconstruct the target signal. Then the envelope aligning is used to make non-coherent accumulation of multiple reflected waveforms. Simulation shows the feasibility of the method.