A SVD approach to compensate probe positioning errors in a non-conventional plane-rectangular near-to-far-field transformation

An efficient singular value decomposition (SVD) technique to correct known positioning errors in a non-redundant plane-rectangular near-to-far-field transformation (NTFFT) using an innovative scan, named planar wide-mesh scanning, is experimentally assessed. This transformation employs a non-redundant sampling representation of the voltage detected by the probe, got by considering the antenna as contained in an oblate ellipsoid, to precisely reconstruct the near-field data necessary to execute the classical plane-rectangular NTFFT from the voltage values at the sampling points specified by the representation. These last samples are unknown and are effectively recovered from the positioning errors affected ones by applying a SVD based procedure. Some experimental results obtained at the University of Salerno and assessing the efficacy of the proposed technique are reported.