Abstract
Based on the coded frequency-selective surface (FSS) loaded with PIN diodes and compressive sensing theory, we proposed a new method of millimetre wave compressive sensing imaging. The FSS is fed by a simple patch antenna right under of it, the diode in FSS unit is set “off” to be transparent to the incident millimetre wave, whereas it is set “on” to reflect the incident wave. By switching the state of each PIN diode randomly, we can get different radiation patterns with low correlation. Such a random-switching imaging mask can compress the information on the physical layer and provide enough effective measurements that are important for compressive sensing. At last, we use the sparsity adaptive matching pursuit (SAMP) algorithm to reconstruct the scene. Simulation results show that the aperture working at 37.5 GHz can accomplish the scene reconstruction with a PSNR = 22 dB and the compression ratio is 4:1.