Welcome to the AMTA paper archive. Select a category, publication date or search by author.
(Note: Papers will always be listed by categories. To see ALL of the papers meeting your search criteria select the "AMTA Paper Archive" category after performing your search.)
We introduce a new type of planar near-field measurement technique for testing small antennas which, heretofore, have been traditionally tested via spherical or cylindrical scanning methods. Field acquisition in both these procedures is compromised to a certain extent by the fact that probe movement induces change in relative geometry with respect to, and thus interaction with, the anechoic chamber enclosure. Moreover, obstructing equipment, such as antenna pedestals, may significantly impede, or even reduce the available angular scope of any given scan. Our proposed procedure, by contrast, minimizes both the residual interaction contaminant and the threat of obstruction. We have in mind here a variant, a hybrid version of planar scanning wherein, on the one hand, we limit severely the size of the acquisition rectangle (and thus minimize the contaminating influence of a variable probe/chamber interaction), while, on the other, we really do collect near-field data throughout a complete range of solid angle around any candidate AUT, front, back, above, below, and on both sides. Such completeness is achieved through the mere stratagem of undertaking six independent planar scans with the AUT suitably rotated so as to expose to measurement, one by one, each of the faces of an enclosing virtual box. In the meanwhile, the inevitable AUT pedestal per se remains immobile and removed from any occupancy conflict with the scanned probe. We have accordingly named our new planar near-field data acquisition scheme the “Boxed Near-Field Measurement Procedure.” With subsequent use of our Field Mapping Algorithm (FMA), elsewhere reported, we obtain the entire field exactly, everywhere, both interior and exterior to the surrounding (virtual) box. In particular, we achieve enhanced accuracy in the far-field patterns of primary interest by virtue of the completeness of data acquisition and its relative freedom from spurious contamination. The angular completeness of data acquisition conferred by our procedure extends in principle to antennas of arbitrary size, provided, of course, that due provision is made for the necessary scope of measurement rectangles. The benefits are seen to be especially valuable in the case of narrow-beam antennas, whose back lobe pattern details, usually deemed as inaccessible and hence automatically forfeited during conventional (i.e., utilizing a “onefaced box,” in our new way of thinking) planar near-field testing, are thrust now into full view. Our new, full-enclosure planar acquisition technique as now described has been verified by analytic examples, as well as by hardware measurements, with excellent results evident throughout, as we are about to demonstrate.
Outdoor far- field antenna test ranges have declined in popularity due to the advent of alternative test methods, e.g., Near-Field Antenna ranges and Compact Antenna Test Ranges. They are also costly to maintain. A natural consequence of that trend is that far-field ranges are either shut down or rendered dormant for long periods of time. The latter was the situation for the NGAS (Northrop Grumman Aerospace Systems) CTS 10K Far-Field range. The Far-Field was an outdoor range with a 10,000’ range length, open transmit site and radome enclosed receive site. It had been dormant for 7 years and was needed for a unique test before the test site was vacated completely. This paper provides a brief description of the range, the upgrades made to address equipment obsolescence and the checkout process to ensure that the range would meet performance requirements. The range needed to operate from 100 MHz to 18 GHz. Therefore, range diagnostics were performed at various frequency points and swept measurements also executed. A Range Readiness report was created and presented internally. Elements of that report are shared in this paper.
The radiation characteristics of antennas can be deter-mined by measuring amplitude and phase data in the ra-diating near-field followed by a transformation to the far-field. Accurate phase measurements especially at high frequencies are very demanding in terms of the required measurement equipment and tolerances. Phaseless mea-surement techniques have been proposed, which often deal with a second set of amplitude only measurement data in order to compensate the lack of phase information. In this paper the concept of phaseless spherical near-field measurements will be addressed by presenting a phaseless near-field transformation algorithm for spherical antenna measurements, working with amplitude only data on two spheres. In particular the measurement of a patch antenna is considered to demonstrate the utility of the technique for low gain antennas. To address the issue of robustness, inaccurate measurement distances as well as spherical rotation angles are considered in order to evaluate the accuracy of the method against probe positioning errors. Furthermore noise contributions are introduced to emu-late measurement inaccuracies in general.
In previous papers [1]-[4] we have presented formulations for the circular and linear near field-to-far field RCS transformations (CNFFFT and LNFFFT, respectively). These formulations assumed that the target did not have significant extent above or below a central (waterline) plane, and that the circular or linear near field scans lied in this waterline plane. In this paper, the CNFFFT and LNFFFT formulations are generalized to scans that lie in a plane parallel to and above or below the waterline plane. These scans correspond to conical or great circle RCS cuts, respectively, in the far field at elevation angles other than 90°. We will show that the generalization can be accomplished by modifying just the frequency domain processing steps that are common to both algorithms, while leaving the spatial processing portions (apart from a minor variable redefinition) unchanged. The paper focuses on the mathematical derivation and numerical implementation of the algorithms; examples of numerical and experiment results are deferred to future papers.
Inverse equivalent current method has recently gained popularity in the applications of near-field far-field (NFFF) transformations especially when near-field (NF) measurements are carried out on irregular measurement grids around the arbitrarily shaped object under test. Usually low order (LO) Rao-Wilton-Glisson (RWG) basis functions or even point based low order basis functions are used for the discretization of the unknown surface current densities on the triangular discretization elements. Better accuracies are achievable when equal number of higher order (HO) basis functions is employed to represent unknown surface current densities. Nearly-orthogonal hierarchical vector basis functions complete to full first order with respect to the curl space are therefore utilized for the discretization of inverse equivalent surface currents defined on flat triangular domains. Various numerical examples are presented and comparison is made with the results of LO discretization.
Accurate calibration of near-field measurements requires the probe used for the measurement be well characterized. The determination of the absolute gain of rectangular open-ended waveguide probes is difficult due to the broad beamwidth in both the E-plane and H-plane which increase the likelihood of multi-path affecting the accuracy of the measurement. Multi-path may be minimized by reducing the separation distance, but at the price that far-field conditions may no longer apply. A variation of the two matched antenna method is to use a large reflecting plate to form an image of the probe. Use of the entire bandwidth of the probe, and time-gating the results to isolate the signal reflected from the plate allows the gain to be determined. The procedure also allows the determination of the aperture reflection coefficient used by theoretical probe models used for pattern compensation in the near-to-far-field transformation.
A technique to diagnose faulty elements present in patch antenna array from either measured far field radiation pattern or return loss characteristic is suggested. A linear array consisting of eight square patch elements with uniform excitation and ./2 spacing between them is considered. A method is developed using Artificial Neural Networks to detect one or two faulty elements present in the array. A neural network is trained with one third of the possible faulty radiation patterns and tested with two thirds of faulty patterns. ANN is implemented with Radial Basis Function neural network (RBF) and Probabilistic neural network and their performance is compared.
An accurate and efficient numerical model is developed to simulate the far field of an antenna under test (AUT) measured in a Compact Antenna Test Range (CATR), on the basis of the known quiet zone field and the theoretical aperture field distribution of the AUT. The comparison with the theoretical far-field pattern of the AUT shows the expected measurement accuracy. The numerical model takes into account the relative movement of the AUT within the quiet zone and is valid for any CATR and AUT of which the quiet zone and aperture field, respectively, are known. The antenna under test is the Validation Standard Antenna (VAST12), especially designed in the past for antenna test ranges validations. Simulated results as well as real measurements data are provided.
An accurate and efficient numerical model is developed to simulate the far field of an antenna under test (AUT) measured in a Compact Antenna Test Range (CATR), on the basis of the known quiet zone field and the theoretical aperture field distribution of the AUT. The comparison with the theoretical far-field pattern of the AUT shows the expected measurement accuracy. The numerical model takes into account the relative movement of the AUT within the quiet zone and is valid for any CATR and AUT of which the quiet zone and aperture field, respectively, are known. The antenna under test is the Validation Standard Antenna (VAST12), especially designed in the past for antenna test ranges validations. Simulated results as well as real measurements data are provided.
ABSTRACT In this work, a probe compensated near-field – far-field transformation technique with spherical spiral scanning suitable to deal with elongated antennas is developed by properly applying the unified theory of spiral scans for nonspherical antennas. A very flexible source modelling, formed by a cylinder ended in two half-spheres, is considered as surface enclosing the antenna under test. It is so possible to obtain a remarkable reduction of the number of data to be acquired, thus significantly reducing the required measurement time. Some numerical tests, assessing the accuracy of the technique and its stability with respect to random errors affecting the data, are reported.
This paper describes measurements performed at the National Physical Laboratory (NPL) and Near Field Systems Inc (NSI) on Open Ended Waveguide (OEWG) probes that are typically used for near-field measurements. The effect of the size and location of the absorber collar placed behind the probe was studied. It was found that for some configurations, the absorber collar could cause noticeable ripples in the far-field patterns of the probe and this in turn could affect the probe correction process when the probe was used in near-field measurements. General guidelines were developed to select an absorber configuration that would have minimal effect on the patterns, polarization and gain of the probes.
A broad band interferometer antenna was designed and manufactured by Saab Avitronics. Saab Aerotech has installed a test facility for calibration of the interferometer antenna. The main purpose of the facility is to measure the interferometric function of the antenna. The interferometric function of the antenna can be measured directly but this method puts very high demands on the test range performance. An alternative method where each element is centered on a short far-field range is evaluated and compared by measurement with a large compact range at Saab Microwave Systems. The paper also describes the design aspects when measuring broad band, broad beam interferometer elements together with the actual design of critical components such as positioners, RF-system and absorber treatment.
This paper analyzes the reduction of the noise effect in spherical near-field antenna measurements. Two techniques have been evaluated: the first one is based on the mode truncation and the second one consists of a spatial filtering after a diagnosis process. The antenna under test (AUT) used for this evaluation is the 12 GHz Validation Standard antenna (VAST12). The VAST12 measurements have been performed in the Spherical Near-Field Antenna Test Facility of the Technical University of Madrid (UPM). These measurements have been corrupted adding a White Gaussian Noise (WGN) with different levels. First, the effect of the number of spherical modes considered in the near-to-far-field transformation has been evaluated, analyzing also the error due to the mode truncation versus the reduction of the noise uncertainty associated to each spherical mode. Second, a diagnosis process based on a holographic technique has been carried out. A spatial filtering including the AUT aperture has been applied and then, the far-field is reconstructed and compared with the uncorrupted far-field. Several results illustrate the signal In this paper, the effect of White Gaussian Noise (WGN) in spherical near-field measurements and the improvement of the signal to noise ratio (SNR) through mode truncation and spatial filtering are evaluated employing simulations and measurements. Section 2 and 3 respectively explain the mode truncation and spatial filtering to minimize the noise. In section 4 and 5 the results achieved when applying both techniques are illustrated. Finally, section 6 summarizes the conclusions drawn.
We describe two theoretical bases for an algorithm for back-projection. The first is (1) Fourier inversion of the mathematical expression for the far electric field components in terms of the aperture electric field. The second is (2) Fourier inversion of the complete vectorial transmitting characteristic of Kerns' scattering matrix. It is this characteristic that results from the standard process of planar near-field (PNF) scanning and the ensuing reduction of the PNF transmission equation. We demonstrate that the theoretical approaches (1) and (2) yield identical back-projection algorithms. We report on back-projection measurements of an 18 inch X-band flat plate phased array using the far-field obtained from both planar and spherical near-field scanning. The spherical measurements were made on a large arch range.
A new simple approach is presented to calibrate the gain of standard gain horn antennas operating in the millimeter-wave frequency band. In terms of calibration, it is difficult to accurately measure the gain of standard gain horn antennas in the far-field region due to the space limitation. Therefore, near-field measurement methods are generally used to calibrate the gain of standard horn antennas. The extrapolation range method is one of the most accurate measurement methods in the near-field region. In the conventional extrapolation range method, a moving average process is applied to remove multiple reflections between antennas. Moving average can only remove multiple reflections between antennas. Therefore, electromagnetic absorbers are required to remove other reflections increasing measurement uncertainties. The time-domain gating method in extrapolation range allows us to remove all reflection waves, and achieve accurate antenna gain calibration without absorbers. The time-domain gating also reduces the number of measurement positions in the extrapolation ranges and obtains the gain of antennas in wide frequency ranges. In this paper, we compare the theoretical value with the time-domain gating method without absorbers by measuring three W-band standard gain horn antennas.
Thermal and other random noise sources give rise to an error contribution in spherical near field measurement systems [12]. With modern receivers and sufficient amplification in the system this term often give an insignificant contribution in the overall measurement uncertainty. However, in special cases the uncertainty linked to random noise may be more significant and the proper treatment of this term is needed to evaluate the impact on overall measurement uncertainty. The motivation for this paper comes from observations on spherical near field (SNF) measurement of relatively small antennas using a high degree of oversampling. In a multiprobe system this is generally the case particularly in measurements of small antennas like dipoles as shown in Figure 1. In these cases the near field to far field (NF/FF) processing is performed with data collected from all probes and some truncation of the spherical mode spectrum depending on the antenna size. The term modal filtering is often used to describe the deliberate truncation of the mode spectrum. What can be observed is that the effective signal to noise ratio (S/N) in small antenna measurements in which modal filtering is applied during NF/FF processing are often much better than the apparent S/N in the “raw” near field (NF) data. Parseval’s theorem, which states that power computed in either domains equals the power in the other explains this difference. The “noise power” is spread out on all available spherical modes and therefore reduced when the mode spectrum is truncated by modal filtering at the appropriate order/distance depending on the size of the antenna. In this paper we present a formal discussion on how the residual noise power after NF/FF processing is affected by the processing parameters. It will be shown that the “effective S/N” can be calculated directly from simple formulas from the applied sampling and filtering. The formulas will be validated by an experimental setup.
The Mathematical Absorber Reflection Suppression (MARS) technique is a method to reduce scattering errors in near-field and far-field antenna measurement systems. Previous tests by the authors had indicated that NSI's MARS technique was not as effective for directive antennas. A recent development of a scattering reduction technique for cylindrical near-field measurements has demonstrated that it can also work well for directive antennas. These measurements showed that the AUT shouldbeoffsetfromtheorigin byadistanceatleastequal to the largest dimension of the AUT rather than only 1-3 wavelengthswhich hadbeenusedfor smallerantennasin the earlier MARS measurements. Spherical near-field measurementshaverecently beenconcludedwhich confirm that with the larger offsets, the MARS technique can be applied to directive antennaswith excellent results. The MARS processing has recently been modified to produce significantly improved results. This improvement isespeciallyusefulfor antennaswherethephasecenterof the horns is located inside the horn and varies with frequency like pyramidal Standard Gain Horns (SGH). Fewermodesarerequired for thetranslatedpatternandthe filtering is more effective at reducing the effect of scattering. The improvement is very apparent for pyramidal horns.
We present an innovative Near-Field test range, named Compact Near-Field (CNF) test range, using photonic probes and advanced Near-Field Far-Field transformations (NFFF). The photonic probe allows distances of one wavelength or less between AUT and probe, drastically reducing test range and scanner dimensions, improving the Signal to Clutter Ratio and the Signal to Noise Ratio, and reducing the scanning area and time. The NFFF, properly formulated as a linear inverse problem, further improves the rejection to clutter, noise and truncation error. The advantages of CNF test ranges are numerically foreseen and experimental results are presented under both, planar and cylindrical scanning geometries.
The probe correction of near-field measured data can be considered as being composed of two parts. The first part is a pattern correction that corrects for the effects of the aperture size and shape of the probe and can be analyzed in terms of the far-field main component pattern of the probe. The second part is due to the non-ideal polarization properties of the probe. If the probe responded to only one vector component of the incident field in all directions, this correction would be unnecessary. But since all probes have some response to each of two orthogonal components, the polarization correction must be included. The polarization correction will be the focus of the following discussion. Previous studies have derived and tested general equations to analyze polarization uncertainty12. This paper simplifies these equations for easier application. The results of analysis and measurements for Planar, Cylindrical and Spherical near-field measurements will be summarized in a form that is general, easily applied and useful. Equations and graphs will be presented that can be used to estimate the uncertainty in the polarization correction for different AUT/Probe polarization combinations and measurement geometries. The planar case will be considered first where the concepts are derived from the probe correction theory and computer simulation and then extended to the other measurement geometries.
An effective near-field – far-field transformation technique with spherical spiral scanning tailored for antennas having two dimensions very different from the third one is here proposed. To this end, an antenna with one or two predominant dimensions (as, e.g., an elongated or quasi-planar antenna) is no longer considered as enclosed in a sphere, but in a prolate or oblate ellipsoid, respectively, thus allowing one to remarkably reduce the number of required data. Moreover these source modellings remain quite general and contain the spherical one as particular case. Numerical tests are reported for demonstrating the accuracy of the far-field reconstruction process and its stability with respect to random errors affecting the data.