Investigation on Diversity Effects for Phaseless Near-Field Far-Field Transformations

Phase retrieval algorithms for phaseless near-field (NF) measurements are known to benefit from diversity (i.e., diversity in the sampling locations or type of probes) in a measurement setup. A numerical reconstruction method is utilized to find the number of independent measurement samples in a phaseless NF antenna measurement setup. Certain magnitude and phase difference measurement samples can be linearly reconstructed from previous phaseless measurements. The reconstruction is possible if the corresponding bilinear form for the new sample is linearly dependent from the bilinear forms corresponding to the previous measurement samples. Any linearly independent bilinear form corresponds to an independent measurement value. It is shown that the diversity in the measurements is continuously increasing with an increasing number of measurement distances until a saturation distance is reached. Furthermore, it is shown that the far-field (FF) magnitudes can be approximately reconstructed by the linear reconstruction procedure from magnitude-only NF measurements. The approximation error decreases if more NF measurement distances are considered.