High-resolution Radar Imaging Using Two-dimensional Superresolution Techniques

雷达成像是识别雷达目标的一项重要技术。然而，由于采用了二维傅里叶变换，分辨率有限。本研究旨在建立高解析度雷达成像方法，并发展相关技术。研究期间取得的主要成果如下：1。提出了一种基于二维MUSIC算法的超分辨率雷达成像技术。它具有高分辨率功能。二维MUSIC在信号较多的情况下不可靠。我们引入了限定区域的空域门控来减少信号的数量。在我们使用MUSIC方法之前，先应用门控。为了检验二维MUSIC的高分辨率能力，我们进行了室内无线电传播实验。我们测量了阶跃频率系统中未调制正弦波的幅值和相位。我们在不同的接收天线位置重复此测量。结果表明，雷达成像中每个多径分量的到达角和传播时间的估计相当于每个散射中心的估计。实验表明，二维MUSIC算法可以分辨出传统傅里叶技术无法分辨的两个位置相近的信号。MUSIC算法在散射中心不具有频率依赖性的情况下能够正常工作。我们修改了MUSIC方法，使其可以应用于具有幅频依赖性的信号。这些考虑对于包括频率相关散射中心在内的雷达目标的高分辨率估计是有用的。

Radar imaging is an important technique to identify a radar target . Resolution is, however, limited because the two-dimensional Fourier transform has been employed. The purpose of this study is to establish a method whichgives high-resolution radar imaging, and to develop related techniques. Major results obtained in the research period are as follws :1.A superresolution technique using a two-dimensional MUSIC algorithm was developed for the radar imaging . It has high-resolution capabilities.2.The two-dimensional MUSIC is not reliable when there are many signals. We introduced spatial-domain gating which bounds a region to reduce the number of signals. The gating is applied before we use the MUSIC method.3.In order to examine the high-resolution capabilities of the two-dimensional MUSIC,we carried out indoor radio propagation experiments. We measure an amplitude and phase of an unmodulated sinusoidal wave in a stepped frequency system. We repeat this measurement at different receiving antenna positions. It was shown that estimation of the arrival angle and propagation time of each multipath component is equivalent to that of each scattering center in the radar imaging. From the experiments, it was shown that the two-dimensional MUSIC algorithm can resolve two close located signals which the conventional Fourier technique cannot resolve.4.The MUSIC algorithm works properly when scattering centers do not have frequency dependence. We modified the MUSIC method in such a way that it can be applied to signals with amplitude frequency dependence. These considerations will be useful for high-resolution estimation of a radar target including frequency dependent scattering centers .