Authors: Bernd Hofmann ,Thomas F. Eibert
Source: FERMAT, Volume 34, Communication 2, Jul.-Aug., 2019
Abstract: The influence of noise on the minimum number of measurement samples for a sparse recovery in a spherical antenna near-field to far-field transformation (NFFFT) is investigated. To this end, several modified phase transition diagrams are determined. These diagrams show the minimum number of samples required to achieve a certain accuracy in the reconstructed far-field. With this approach, the reconstruction by the basis pursuit algorithm and its quadratically constrained version are shown to be sufficiently robust against measurement noise for actual measurements. In consequence, sparse recovery can be applied to the spherical NFFFT with predictable accuracy.
Index Terms: Gram matrix, Rao-Wilton-Glisson functions, preconditioner, matrix inversion
View PDF
Sparse Recovery with Predictable Accuracy in Noisy Spherical Antenna Near-Field Measurements