一种面向水面纹理的毫米波LFMCW雷达成像算法

Imaging Algorithm of Millimeter-wave LFMCW Radar for Water Surface Texture Detection

利用毫米波线性调频连续波(LFMCW)雷达探测水面时,静止目标的回波以及噪声通常会淹没水面本身的回波信号,导致采用传统成像算法得到的结果中难以看到水面波浪纹理。针对这一问题,该文提出一种面向水面纹理的毫米波LFMCW雷达成像算法,该算法在距离向上采用Dechirp技术进行距离压缩,在方位向上进行分块处理。方位向分块处理过程中,首先根据静止目标与运动目标回波多普勒频率不同的特性,在频域去除静止目标回波的干扰;然后基于水面电磁散射特性,采用最大似然估计方法估计方位向频谱参数,计算水面回波能量。采用该算法对实测数据进行处理,结果显示该算法能够获得水面纹理信息,成像结果优于传统成像算法。

In the application of millimeter-wave Linear Frequency Modulated Continuous Wave （LFMCW） radar for water surface detection, the echo of water surface itself is always covered by the echo of stationary targets and noises, leading to the result that water surface texture can hardly be seen in the figures obtained by the conventional imaging algorithm. To solve this problem, an imaging algorithm of millimeter-wave LFMCW radar for water surface texture is proposed, the Dechirp technique is adopted to complete the range compression in range direction, and the data is divided into blocks to be dealt with separately in azimuth direction. During the processing in azimuth direction, interference from static targets is removed in frequency domain according to the fact that stationary targets and moving targets have different Doppler frequencies; then, based on the electromagnetic scattering characteristic of water surface, a maximum likelihood estimation method is used to estimate azimuth spectrum parameters to calculate the energy of water surface echo. The proposed algorithm is used to process measured data, and the results show that water surface texture can be obtained, which means that the proposed algorithm is superior to the traditional one.