Cylindrical MIMO Array-Based Near-Field Microwave Imaging

Microwave/millimeter-wave imaging has a wide prospect in applications of concealed weapons detection. The main problem of the existing techniques lies in the contradiction between the imaging speed and the system cost. In this respect, a cylindrical sparse multiple-input-multiple-output (MIMO) array topology is proposed, which has no mechanical scanning so as to achieve fast sampling process. Furthermore, it can provide more angular observations than the mainly used systems with a planar aperture. To speed up the image reconstruction process, a phase center approximation-based transformation is made to convert the multistatic sampling scheme to its monostatic counterpart for which there are more fast and tractable imaging algorithms. Then, a phase calibration method is presented after we analyze the error of such an equivalence. In so doing, the cylindrical MIMO topology is transformed into an approximate monostatic one. Finally, a wavenumber domain algorithm is employed to attain the fast image reconstruction. Numerical simulations are provided to demonstrate the performance improvement of the proposed technique.