基于稀疏贝叶斯学习的多径抑制的高分辨穿墙雷达成像技术研究

Through-the-wall radar imaging(TWRI) is a newly developed technique that can form an image of region of interest which is behind all obstacles．Owing to its capabilities of detection，imaging，positioning，tracking and recognition of the target which is behind the wall，TWRI has been involved in a wide range of applications, such as street battle，anti-terrorist, disaster relief and hostage rescue．In the TWRI, the observed scenario is often a closed environment and the propagation of electromagnetic wave in such an environment would lead to refractions from the interior walls. Therefore, the received echo includes the direct echo, which is from the transmitter direct to the target and back, and multipath echoes (indirect echoes), which are firstly from the transmitter to the interior walls, and then refraction to the target and back. These multipath echo components in the received echo would lead to ghosts in the formed image. These ghosts would result in the high false alarm in the detection, and it would have seriously adverse effect on radar detection performance. In this project, the research on high-resolution TWRI with multipath suppression will be launched. The echo model based on the geometry configuration of the multipath propagation is firstly analyzed, and then high-resolution TWRI imaging algorithms with multipath suppression will be proposed in the sparse Bayesian learning framework, based on these aspects including sparsity of targets, group sparsity of the scattering coefficients, the correlation of scattering coefficients and the spatial continuity of the target of interest.

穿墙雷达成像是最近发展起来的一种对障碍物后的区域进行微波成像的技术，它能够对隐藏在室内或者建筑物后的目标进行探测、成像、定位、跟踪与识别等，因此在城市巷战、反恐安保、灾害及人质救援等方面有着广泛的应用。在穿墙雷达成像过程中，由于观测场景往往是封闭的室内环境，电磁波不仅会在目标与雷达天线之间传播，还会在天线-墙体-目标之间进行多次折射，产生多径传播现象。在成像结果上目标的多次折射回波会使得图像上出现多个虚假目标（虚像），这些虚像的出现会给目标检测引入很多虚警，严重影响雷达系统检测性能。本课题主要研究多径传播环境下的高分辨穿墙雷达成像技术，分析多径传播的几何模型及其回波信号模型的特点，从目标稀疏性、多径传播组稀疏结构、散射系数关联性学习以及目标内在的空间分布结构等角度出发，探索和挖掘目标在多径传播中的内在结构和特性，在稀疏贝叶斯学习框架下研究并构建多径抑制的高分辨成像新方法。

穿墙雷达成像是最近发展起来的一种对障碍物后的区域进行微波成像的技术，它能够对隐藏在室内或者建筑物后的目标进行探测、成像、定位、跟踪与识别等，因此在城市巷战、反恐安保、灾害及人质救援等方面有着广泛的应用。在穿墙雷达成像过程中，由于观测场景往往是封闭的室内环境，电磁波不仅会在目标与雷达天线之间传播，还会在天线-墙体-目标之间进行多次折射，产生多径传播现象。在成像结果上目标的多次折射回波会使得图像上出现多个虚假目标（虚像），这些虚像的出现会给目标检测引入很多虚警，严重影响雷达系统检测性能。本课题主要研究多径传播环境下的高分辨穿墙雷达成像技术，分析多径传播的几何模型及其回波信号模型的特点，从目标稀疏性、多径传播组稀疏结构、散射系数关联性学习以及目标内在的空间分布结构等角度出发，探索和挖掘目标在多径传播中的内在结构和特性，在稀疏贝叶斯学习框架下研究并构建多径抑制的高分辨成像新方法。课题组通过三年的技术攻关，提出了基于组稀疏结构的高分辨穿墙雷达成像新方法，并进一步挖掘组稀疏结构组内散射系数的关联性，提出了联合组稀疏结构和组内关联性学习的高分辨穿墙雷达成像新方法，通过充分利用观测场景中目标的空间连续分布特性，提出联合组稀疏结构和空间分布结构的新方法，最后主要通过仿真实验验证了算法的有效性。该部分研究成果可直接应用有反恐安保、地震救援和室内监视等领域。