Backward Projection Imaging of Through-Wall Radar Based on Airspace Nonuniform Sampling

Currently, backward projection (BP) imaging is widely used in different kinds of through-wall radars. This imaging method is used to make the pixel points corresponding to each echo signal consistent in the time domain by compensating for the time delay of information between the transceiver antenna and the detection target and performing coherent superposition to finally project the target in space. Because of this property, the BP imaging algorithm is computationally intense and has a long computation time, which limits its application in practical engineering. In this paper, we propose backward projection imaging of wall-penetrating radar with nonuniform sampling in the air domain to address the problem of the imaging speed of wall penetrating radar. This is a universal optimization algorithm that optimizes the imaging process of radar in scenes with sparse targets to obtain better imaging results in less time. A stepped-frequency continuous waveform (SFCW) ultra-wideband (UWB) multiple-input multiple-output (MIMO) radar is prepared for the experiments, and the imaging time of the algorithm is verified using real-world data. The results show that the method can achieve real-time imaging based on existing technology, laying a firm foundation for the practical application of through-wall radar.