Study Of Some Key Problems Of Phaseless Near-Field Antenna Measurement

In order to reduce the cost of near-field antenna measurement at high frequency range and also for the reason that it is quite difficult to obtain the phase distributions on the scanning planes with high-frequency in test.This article presents a method,we called planer phaseless near-field antenna measurement,which the kernel of the method is to collect some groups of magnitudes data on the scanning planes and retrieving the phase distributions at each sampling points by algorithms.Then,the far-field pattern of the antenna under test can be constructed using the retrieved phase data,sampled magnitudes data and near-field to far-field transformation theorem.Firstly,the theorem of traditional antenna measurement is stated,the core of the theorem is reconstructing the far-field pattern using stationary phase method and plane wave spectrum method.On this basis,the probe receiving directivity,which is defined as probe receiving spectrum,is taken into account for far-field pattern reconstruction.Then,the near-field antenna measurement theory with probe compensated is established.Secondly,the theorem of phaseless planer near-field antenna measurement is presented based on the traditional antenna measurement theory.Two sets of scanning planes are placed paralleled in the near field region of the antenna under test and magnitudes data are collected on these two planes respected to the Nyquist sampling theorem.After that the far-field pattern is constructed using NF-FF transformation whose input data are magnitudes data from measurement and the phase data from the algorithm called iterate Fourier transformation.What's more,the phaseless measurement strategy developed for scanned bean antenna is also presented.The beam pointing angle which is determined firstly by Differential Evolution Algorithm on the first step and the initial phase distribution can be determined by this angle at the same time.In the next step,the accurate phase distributions and the far-field pattern can be found by iterate Fourier transformation and NF-FF transformation.This article also investigates two initial conditions which we called the random iterate phase distributions and the phase distributions based on the shape of the antenna aperture,which have different influence on the results of phase retrieving.And we further point out that a better far-field pattern,a faster convergence rate and a higher precision can be obtained on the condition of the phase distribution based on the shape of the antenna apertureThirdly,the convergence of the phase retrieval method is discussed.The convergence of the Differential Evolution Algorithm is proved by applying Markov chain theory at the beginning,and we point out that the convergence of the DEA has a weak dependence of the initial conditions as this method is a global optimization algorithm.The convergence of Iterate Fourier Transformation is proved by set theory,norm theory and alternating projection theorem and we point out that iterate Fourier transformation is a local optimization algorithm,so it has a strong dependence of the initial condition.We then further presents that the differential evolution algorithm can provide a good initial condition for iterate Fourier transformation algorithm.At last,a theoretical basis about the phase retrieval strategy which combines with differential evolution algorithm and iterate Fourier transformation is established.Finally,we prove our theory by the simulation using symmetrical array antenna and horn antenna,and the influence of truncation-error,sampling rate,distance between two scanning planes and probe position error on the phaseless near-field measurement is investigated.After all,a parameter selection strategy is given to instruct the actual antenna measurement.