A new look at phaseless planar near-field measurements: limitations, simulations, measurements, and a hybrid solution

In this paper, we have revisited the phase-retrieval problem for planar near-field antenna measurements. It will be shown that the complexity of retrieval procedures is a function not only of the independency of different sets of measurements, but also of the characteristics of the antenna under test (AUT). Antenna-pattern features, such as its beam direction, will have a profound effect on the success of phase-reconstruction algorithms. The failure of a well-known phase-retrieval method, the iterative Fourier technique, is investigated for a case where the antenna has a scanned beam. It is found that this is due to a less-than-optimum choice of the initial guess. To alleviate the deficiency of the iterative Fourier technique, a simple but effective initial guess is sought by the use of the differential evolutionary algorithm. This global search technique tries to find the best initial phase guess that minimizes an error criterion. Subsequently, this best guess is fed to the iterative Fourier phase-retrieval routine for further phase refinements. Having done this, the far-field can subsequently be constructed. The improvement in the phase-reconstruction algorithm is examined through a series of simulations and measurements.