基于矩阵填充的随机步进频雷达高分辨距离-多普勒谱稀疏恢复方法  

Matrix Completion-based Range-Doppler Spectrum Estimation for Random Stepped-frequency Radars

作　　者：胡雪瑶 梁灿 卢珊珊 王在洋 郑乐[1,2,3] 李阳 HU Xueyao;LIANG Can;LU Shanshan;WANG Zaiyang;ZHENG Le;LI Yang(Radar Research Lab,School of Information and Electronics,Beijing Institute Technology,Beijing 100081,China;Chongqing Innovation Center,Beijing Institute of Technology,Chongqing 401120,China;The Electromagnetic Sensing Research Center of CEMEE State Key Laboratory,Beijing 100081,China;Xi’an Electronic Engineering Research Institute,Xi’an 710100,China)

机构地区：[1]北京理工大学信息与电子学院雷达技术研究所,北京100081 [2]北京理工大学重庆创新中心,重庆401120 [3]CEMEE国家重点实验室电磁感知研究中心,北京100081 [4]西安电子工程研究所,西安710100

出　　处：《雷达学报（中英文）》2024年第1期200-214,共15页Journal of Radars

基　　金：国家自然科学基金(62388102);国家重点研发计划(2018YFE0202101,2018YFE0202103)。

摘　　要：随机步进频雷达通过合成大带宽,能在较低硬件复杂度下获得距离高分辨效果,同时由于其每个脉冲的载频随机捷变,因而具有强的抗干扰、电磁兼容能力,在复杂电磁环境高精度探测领域具有重要的应用价值。然而,由于其波形在时频域稀疏的感知形式,造成回波相参信息有所缺失,因而传统匹配滤波方法在估计高分辨距离-多普勒时会演化为欠定估计,导致估计谱中产生起伏高旁瓣,严重影响探测性能。为此,该文提出一种基于Hankel重构矩阵填充的随机步进频雷达高分辨距离-多普勒谱低旁瓣稀疏恢复方法。该方法采用低秩矩阵填充思想补全波形在时频域稀疏感知时造成的缺失采样,恢复目标连续相参信息,可以有效解决欠定估计问题。文章首先构建了随机步进频雷达的慢时间-载频(时-频)回波欠采样数据矩阵;然后,重构待恢复数据矩阵为双重Hankel型,并分析证明了矩阵满足低秩先验特性;最后,利用ADMM算法补全未采样时频数据,恢复相参信息,保证了高分辨距离-多普勒谱低旁瓣稀疏恢复。仿真和实测试验证明了该文所提方法的有效性和优越性。Random Stepped Frequency(RSF)radars can achieve high-range resolution with relatively low hardware complexity by synthesizing a wide bandwidth.Moreover,because of the random frequency agility of each pulse,the radars possess robust anti-interference and electromagnetic compatibility capabilities,rendering them invaluable for high-precision detection in complex electromagnetic environments.However,the inherent sparsity sensing of the radar waveform in the time-frequency domain,causes a lack of echo coherence information,leading to an underdetermined estimation of the traditional matched filter,which results in fluctuating high side lobes in the estimation spectrum and adversely deteriorating detection performance.This paper proposes a sparse recovery method based on Hankel matrix completion for the high-resolution range-Doppler spectrum of the RSF radars.Using the low-rank matrix completion concept,this method fills in the missing samples caused by sparse sensing for RSF radars,thereby restoring continuous coherence information and effectively addressing the underdetermined estimation issue.First,an undersampled data matrix of a single coarse-resolution range for RSF radar is constructed.Subsequently,the time-frequency data matrix is reconstructed into a double Hankel form,and its low-rank prior characteristics are analyzed and proven.Finally,the Alternating Direction Method of Multipliers(ADMM)algorithm is applied to restore the unsampled time-frequency data,ensuring sparse recovery of the high-resolution range-Doppler spectrum with low sidelobes.Simulations and real tests demonstrate the effectiveness and superiority of the proposed method.

关 键 词：随机步进频 相参处理 高旁瓣 稀疏恢复 矩阵填充 

分 类 号：TN958.6[电子电信—信号与信息处理]