Antenna Characterization From a Small Number of Far-Field Measurements via Reduced-Order Models

The determination of a small number of samples to characterize the antenna far-field with a controlled accuracy is addressed. The radiation operator that maps the equivalent currents to the far-field is built from the outer dimensions of the antenna under test (AUT) and is discretized into the so-called radiation matrix using the methods of moments. This matrix is approximated up to the measurement accuracy by truncating its singular value decomposition (SVD) accordingly. This operation enables to construct numerically a reduced basis of the radiated fields. The dimension of this basis and consequently the number of field samples is shown to be determined by the area of the equivalent current convex surface, in agreement with the fundamental works on the minimum non-redundant samplings. The influence of the field sampling strategy and noise on the far-field pattern reconstruction is investigated to determine a realistic small number of field samples. Finally, the characterization of two radiating structures from the so-derived reduced number of samples validates experimentally the proposed approach and demonstrates its practical relevance.