基于谱分析的分布式无源雷达非均匀成像  

Distributed Passive Radar Nonuniform Imaging Based on Spectral Analysis

作　　者：韩裕生 张延厚 王硕 姜兆祯 HAN Yusheng;ZHANG Yanhou;WANG Shuo;JIANG Zhaozhen(Department of Information Engineering,Army Academy of Artillery and Air Defence,Hefei 230031,China;Anhui Province Key Laboratory of Polarization Imaging Detection Technology,Hefei 230031,China)

机构地区：[1]陆军炮兵防空兵学院信息工程系,安徽合肥230031 [2]偏振光成像探测技术安徽省重点实验室,安徽合肥230031

出　　处：《雷达科学与技术》2020年第2期187-193,199,共8页Radar Science and Technology

基　　金：安徽省自然科学基金(No.1808085MF163)。

摘　　要：分布式无源雷达空间构型不规则,导致目标回波采样在空间谱域分布不均匀,目标反演质量较差。首先构建三维投影模式下的二维成像模型;对非均匀谱域进行量化分析,建立空间谱支撑域、扭曲度和RMSE三项指标评价空间谱填充非均匀度;然后提出一种快速高斯网格NUFFT的成像方法,通过高斯卷积核对非均匀采样回波在空间谱上进行平滑处理并进行过采样FFT,同时对高斯核进行分裂处理实现快速重构。仿真实验表明,本文评价指标合理评估了空间谱非均匀性,成像方法进一步改善成像质量,且大幅提高计算速度,验证了空间谱非均匀度对成像质量的影响。The spatial configuration of the distributed passive radar is irregular,resulting in uneven distribution of the target echo sampling in the spatial spectral domain and poor target inversion quality.Firstly,a 2D imaging model is constructed under 3D projection mode.Then,the nonuniform spectral domain is analyzed quantitatively,and three indexes,including spatial spectral support domain,distortion degree and RMSE,are established to evaluate the spatial spectral filling nonuniformity.Then,a fast Gaussian gridding NUFFT algorithm is proposed.The Gaussian convolution kernel is used to perform smoothing and over-sample of the nonuniform sampling echoes on the spatial spectrum,and the Gaussian kernel is split to achieve fast reconstruction.The simulation experiments show that the evaluation index proposed in this paper can reasonably evaluate the spatial spectrum nonuniformity.The imaging method further improves the imaging quality,greatly increases the calculation speed,and verifies the influence of spatial spectrum nonuniformity on the imaging quality.

关 键 词：分布式无源雷达成像 非均匀空间谱分析 评价指标 非均匀快速傅里叶变换 快速高斯网格 

分 类 号：TN958.97[电子电信—信号与信息处理]