Reduced Azimuthal Sampling for Spherical Near-Field Measurements

This paper investigates on the use of under-sampling over the azimuthal dimension to reduce measurement time on spherical near-field scanning. This means that the number of angular φ samples is reduced, which allows to reduce the number of positioner steps, obtaining measurement time savings virtually proportional to the number of sample reduction. Of course, this under-sampling introduces an error, which can be interpreted as an aliasing term over the retrieved Spherical Wave Expansion of the Antenna Under Test (AUT). The axial symmetry of most antennas allows the application of significant under-sampling ratios with small aliasing errors. However, this information is not a priori known due to the lack of a reference AUT radiation pattern, or in the case of malfunctioning antennas with degraded symmetry. An “aliasing error estimator” is introduced, which estimates the error caused by the under-sampling without any a priori knowledge of the AUT. This estimator can be used as a stopping criterion of the iterative measurement procedure when the desired accuracy is achieved. The proposed technique will be demonstrated using different antennas, showing considerable reductions in measurement time with low errors in the transformed far-field pattern, and with the guarantee that the error is below a given threshold thanks to the derived estimator.