Numerical Analysis of Mask-Based Phase Reconstruction in Phaseless Spherical Near-Field Antenna Measurements

Area of Science:

• Electromagnetics and Applied Physics
• Antenna Theory and Design
• Signal Processing

Background:

• Phase retrieval is crucial for reconstructing complex signals from amplitude measurements.
• Spherical near-field antenna measurements often lack phase information, hindering the determination of antenna characteristics.
• Recovering complex coefficients (e.g., equivalent currents, spherical modes) is essential for obtaining the antenna's far-field pattern.

Purpose of the Study:

• To apply and evaluate a mask-based phase recovery technique for phaseless spherical near-field antenna measurements.
• To investigate the effectiveness of different mask types in improving the accuracy of antenna parameter reconstruction.
• To assess the impact of mask parameters, oversampling, and algorithm choice on reconstruction performance.

Main Methods:

• Mathematical formulation of the mask-based phase recovery concept.
• Introduction and discussion of various mask types (random distributions, ϕ-rotations, probes).
• Performance evaluation using simulations with multiple antennas under test (AUTs) and the Wirtinger flow algorithm.

Main Results:

• The mask-based approach demonstrated potential for improving reconstruction error in phaseless spherical near-field measurements.
• Reconstruction accuracy is influenced by factors such as the number of masks used, the degree of oversampling, and the specific mask type employed.
• Simulations provided quantitative insights into the performance of different masks and their impact on retrieving antenna coefficients.

Conclusions:

• Mask-based phase recovery is a viable technique for enhancing antenna measurements where phase information is absent.
• The choice of mask type and measurement configuration significantly affects the success of the phase retrieval process.
• This method offers a promising avenue for more accurate characterization of antennas from amplitude-only near-field data.