Far-Field prediction from a reduced number of phaseless bipolar Near Field data

This paper proposes an effective phaseless nonredundant Near-Field-Far-Field (NF-FF) transformation with a planar bipolar scan. The here developed characterization approach relies on the nonredundant sampling representation of the squared amplitude of the electromagnetic field, derived by assuming the source as enclosed in an oblate spheroid. A remarkable reduction (about $90 \%$) of needed NF samples, as compared to standard $\lambda / 4$ sampling, has been reached. Furthermore, by exploiting the available a priori information on the antenna under test, an effective representation of the unknown has been devised, to improve the performance of the solution algorithm. A numerical analysis has been carried out to assess the effectiveness of the proposed characterization approach.