基于FMCW MIMO毫米波雷达的压缩感知三维成像  

Compressed sensing 3D imaging with FMCW MIMO millimeter wave radars

作　　者：刘东东 张远辉 刘康 王书楠 LIU Dongdong;ZHANG Yuanhui;LIU Kang;WANG Shu'nan(College of Mechanical and Electrical Engineering,China Jiliang University,Hangzhou 310018,China;Shanghai Andar Electronic Technology Co.,LTD.,Shanghai 200000,China)

机构地区：[1]中国计量大学机电工程学院,浙江杭州310018 [2]上海岸达电子科技有限公司,上海200000

出　　处：《中国计量大学学报》2024年第3期496-508,共13页Journal of China University of Metrology

基　　金：中德无人驾驶中心毫米波雷达联合实验室基金项目(No.03103-211179),浙江省自然科学基金项目(No.LY19 F010007)。

摘　　要：目的:针对调频连续波(frequency modulated continuous wave,FMCW)多输入多输出(multiple input multiple output,MIMO)雷达采用传统压缩感知方法由于观测矩阵过大导致重构困难的问题,在不影响压缩感知重构算法的性能下,提出了一种基于压缩感知重构理论的分区压缩感知重构成像方法。方法:此方法在传统的压缩感知重构方法的基础上,对包含水平角度、俯仰角度与距离信息的雷达阵列信号张量进行多维度的联合观测,在对数据进行压缩采样的条件下极大保留目标的三维特征;引入图像压缩感知领域的分块压缩感知(block compressed sensing)思想,大幅度降低了观测与重构时的内存占用;对观测矩阵采用非随机的构造方法,使其更易于在硬件上实现;与传统的直接成像算法相比,基于压缩感知重构理论的分区压缩感知重构算法能够在保证成像效果的同时,大幅降低数据采样率与数据传输压力。结果:通过仿真验证,分区压缩感知重构算法能够在压缩采样比为0.4的极端条件下,平均结构相似度达到79%;通过雷达成像实验表明,此算法在通用环境下依旧能够实现复杂目标的高分辨成像。结论:所提出的算法能够在较少采样的条件下实现雷达的高分辨成像,相较于传统重构算法更易于实现。Aims:For frequency modulated continuous wave(FMCW)multiple input multiple output(MIMO),a MIMO radar uses the traditional compressed sensing method to solve the difficult reconstruction problem due to the large measurement matrix.Without affecting the performance of the compressed sensing reconstruction algorithm,this paper proposes a partition compressed sensing reconstruction imaging method based on the compressed sensing reconstruction theory.Methods:Based on the traditional compressed sensing reconstruction method,the multi-dimensional joint observation of radar array signal tensors containing horizontal angle,pitch angle and range information was carried out;and the three-dimensional characteristics of the target were greatly retained under the condition of compressed data sampling.The idea of block compressed sensing in the field of image compressed sensing was introduced to greatly reduce the memory occupation during observation and reconstruction.A non-random construction method was used to construct the measurement matrix,which made it easier to implement in hardware.Compared with the traditional direct imaging algorithm,the partitioned compressed sensing reconstruction algorithm based on the compressed sensing reconstruction theory could greatly reduce the data sampling rate and data transmission pressure while ensuring the imaging effect.Results:Simulation results showed that the partitioned compressed sensing reconstruction algorithm could achieve an average structural similarity of 79% under the extreme condition of a compression sampling ratio of 0.4.Radar imaging experiments showed that the algorithm could still achieve high resolution imaging of complex targets in the general environment.Conclusions:The proposed algorithm can achieve high-resolution radar imaging under the condition of less sampling,which is easier to implement than the traditional reconstruction algorithm.

关 键 词：MIMO雷达 压缩感知重构 阵列信号处理 成像雷达 高分辨成像 

分 类 号：TN957.51[电子电信—信号与信息处理]