Reflectivity Reconstruction from Phaseless, Non-Redundant Near-Field, Plane-Polar Data

Image-based approaches allow estimating the reflectivity of a target from near-field (NF) measurements. We propose a new approach to reconstruct the reflectivity of a planar target from amplitude only data, whereby monostatic non-redundant NF data are used. A plane-polar scanning strategy is considered. The phaseless imaging problem is dealt with as a quadratic inverse one and requires two sets of independent amplitude measurements of the scattered field, collected on two different scanning surfaces. A non-redundant sampling representation of the squared amplitude of the NF data is worked out. A big reduction of about 94% of the requested NF samples, as compared to classical conservative λ/8 phaseless measurements, is achieved. No loss of accuracy is experienced with respect to the case when the reconstructions are performed by complex (i.e., amplitude and phase) data. Notably, the overall samples are even lower than those necessary to carry out a complex standard NF procedure on a solitary surface. The only available a priori information on the scatterer is the size of the enclosing box. This is accounted for by an effective representation of the reflectivity profile, thus improving accuracy and reliability of the phaseless technique. Numerical and experimental results assessing the effectiveness of the approach are presented.