MIMO through-wall radar 3-D imaging of a human body in different postures

This paper focuses on three-dimensional (3-D) imaging for the human body with different postures to enhance the situation awareness ability of through-wall radar. First, we design a two-dimensional MIMO array where the receiving antennas are uniformly and sparsely arranged in horizontal and vertical directions and four time-division multiplexing transmitting elements are placed at the four array corners. As a result, the equivalent aperture is closed to four times the size of the physical array. Second, an identical time-delay wall compensation algorithm attached to back-projection algorithm is presented to easily and approximately correct the image distortion caused by wall penetration, which employs an identical time-delay offset in the whole 3-D pixel-to-pixel imaging process to partially balance out multiple different focusing time-delay errors corresponding to multiple transmit-receive channels in each pixel. Finally, we develop a human body model with three typical postures in the FDTD simulation. Based on the MIMO array and the timedelay compensation algorithm, the simulated images created via the back-projection algorithm clearly provide the identifiable posture characteristics.