中文摘要：

提出了基于扩展信源熵值理论的超宽带穿墙成像雷达墙体强杂波抑制方法.首先将回波信号离散化,计算离散信源的概率空间并对该离散信源进行扩展,计算得到扩展后含有墙体强杂波和目标回波的新信源的熵值.然后根据墙体杂波熵值与目标信号熵值的差异设定门限,自适应选取最佳门限调节因子,对回波信号进行杂波抑制处理.经过墙体强杂波抑制处理后,利用后向投影方法对目标进行成像.以基于时域有限差分方法（Finite Difference-Time Domain, FDTD）的仿真软件GprMax2D/3D所获得的穿墙雷达数据进行仿真实验,分别通过基于信源熵值的方法与本文所提方法来抑制墙体强杂波并成像,通过对比结果可知,前者的目标-杂波比增量为15.51 dB,后者的目标–杂波比增量为19.74 dB.因此,本文所提方法能够在相同测量方式下得到更为精确的成像,而且可以在保证成像效果的前提下大大减少天线扫描次数。

英文摘要：

Strong front wall clutter has serious impacts on the target detection and imaging in the through-wall radar （TWR） system. A method of robust wall clutter suppression based on the entropy of an expanded antenna source for ultra-wide-band through-wall radar is presented in this paper. The model of TWR scenario consists of four layers. Assume that the first and the third layers are air space, while the second and the fourth layers are composed of uniform flat concrete wall. The circular target, assumed to be a perfect electric conductor, is located in the third layer. Along the measurement line which is parallel to the front wall, the transceiver antenna scans uniformly. The echo signals that come from the target and walls are processed into discrete data at first, so that the calculation of probability space is subsequently implemented and the discrete data are expanded as well. And then the entropy of the expanded data that contain robust wall clutter and echo of target is calculated. Taking into consideration the amplitude of target signal varying in each scan, while that of clutter signal is not, it is evident that the entropy can be utilized to discriminate the signals between the target and wall. According to the difference between the entropy of the wall clutter and that of the target, a certain threshold can be set and the optimum tolerance threshold is adaptively selected on the basis of target-to-clutter ratio. With the optimum tolerance threshold, process of clutter suppression is conducted. Finally, back projection is employed for imaging of target. In this paper, data of through-wall radar for simulation are provided by GprMax2D/3D, based on the finite difference-time domain methsd. The clutter suppression and imaging are separately conducted by the method based on data entropy and the method proposed in this paper. Comparing the results of simulations, it is shown that the gain of target-to-clutter ratio for the former is 15.51 dB, and that for the latter is 19.74 dB. It is obvious that the propose