Compressive Sensing Based Joint Wall Position Detection and Multipath Exploitation in Through-the-Wall Radar Imaging

In through-the-wall radar imaging, ghost targets can arise because of multipath reflections. Having a model that compensates for the multipath returns solves this problem; however this requires the knowledge of the room geometry. If wrong information about the room geometry is used in the model, the obtained image will be denser than the image of the true scene. This paper proposes a multipath exploitation method that utilizes this fact. A greedy search is performed to find the wall positions that produce the sparsest image of the desired scene. Hence, detecting the true wall positions and obtaining a ghost-free image. Simulation results demonstrate the effectiveness of the proposed method.