부산대학교 도서관

This work aims to propose and optimise a non-redundant spherical spiral near-to-far field (NTFF) transformation for elongated AUTs from spiral near-field (NF) data acquired over the upper hemisphere due to the presence of an infinite perfectly electric conducting (PEC) ground plane. Such a technique properly exploits the principle of image and the theoretical foundations of spiral scan for non-volumetric AUTs to develop the non-redundant representation along the sampling spiral in presence of PEC ground plane and to synthesise the voltage NF data which would be acquired over the spiral wrapping the lower hemisphere. Once these voltage NF data have been synthesised, then an efficient 2-D optimal sampling interpolation scheme allows the recovering of the NF data required by the classical NTFF transformation. In the hypothesis that the AUT and its image exhibit a predominant dimension as compared to the other two ones, a prolate spheroidal source modeling is here adopted. Numerical tests show the accuracy of the developed non-redundant spherical spiral NTFF transformation.