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.)
Since the early days of antenna pattern recorders, advances in instrumentation and computers have enabled measurement systems to become highly automated and much more capable. Automated systems have provided higher productivity, more efficient use of test facilities, and the ability to acquire more data in less time. In recent years, measurement speeds of microwave receivers and vector network analyzers have advanced considerably. However, to take full advantage of these speed improvements, the measurement system architecture must be carefully considered. Small differences in instrument timing that are repeated many times can make large differences in system measurement time. This paper describes a general method of calculating system measurement time based on the primary factors that affect system timing, including position trigger detection, frequency switching time, multiplexer switching time, receiver measurement time, and timing overhead associated with triggers, sweeps, and measurements. It also shows how key features of instruments available today can be used along with improved antenna measurement system architectures to optimize system throughput.
A comparison of antenna gain and radiation pattern for R-band (1.7 – 2.6 GHz) antennas has been performed between Korea Research Institute of Standards and Science (KRISS) and eight domestic participants including private companies and public institutes. Its purpose was to check equivalences between KRISS and participants in gain and radiation pattern measurements for antennas, particularly at R-band, to support antenna manufacturers and end users in Korea as a proficiency test program of the ‘Antenna Measurement Club’ of KRISS. This comparison uses three traveling standards (a general purpose antenna (pyramidal standard gain horn antenna), a fan-beam antenna (a sector antenna for mobile base stations), and a small antenna (a sleeve dipole antenna) and measurement parameters are the power gain and radiation pattern of the traveling standards. Gain comparison method, extrapolation method, and cylindrical near-field measurement method are used in this comparison.
ABSTRACT This work deals with the experimental validation of a direct near-field – far-field transformation with cylindrical scanning for electrically long antennas, which requires a minimum number of measurements. Such a transformation is based on a nonredundant sampling representation making use of a flexible source model-ling suitable to deal with electrically long antennas and allows the evaluation of the antenna far field directly from the acquired near-field data without interpolating them. The good agreement between the so recovered far-field patterns and those obtained via the classical cylindrical near-field – far-field transformation assesses the effectiveness of the approach.
Mathematical Absorber Reflection Suppression (MARS) has been used successfully to identify and extract range multi-path effects in a great many spherical [1, 2], cylindrical [3, 4], and planar [5, 6] near-field antenna measurement systems. This paper details a recent advance that enables the MARS measurement and post-processing technique to be used to correct antenna pattern data from far-field or compact antenna test ranges (CATRs) where only a single great circle pattern cut is taken. This paper provides an overview of the measurement and novel data transformation and post-processing chain that is utilised to efficiently correct far-field, frequency domain antenna pattern data. Preliminary results of range measurements that illustrate the success of the technique are presented and discussed.
Azimuthal mode (µ mode) truncation of a high-order probe pattern in probe-corrected spherical near-field antenna measurements is studied in this paper. The results of this paper provide rules for appropriate and sufficient µ-mode truncation for non-ideal first-order probes and odd-order probes with approximately 10dBi directivity. The presented azimuthal mode truncation rules allow minimizing the measurement burden of the probe pattern calibration and reducing the computational burden of the probe pattern correction.
In this work we study the near-field antenna measurement using cubical surface scanning and related near-field to far-field (NF-FF) transformations. The cubical surface scanning is a fascinating idea because it can be realized using widely used planar scanning on six surfaces of a cube, and it provides the possibility to determine the complete 3-D pattern instead of the pattern in a limited angular region as in traditional planar scanning. The NF-FF transformation presented in this paper is based on spherical vector wave expansion (SWE). The most important issue of this paper is to introduce the azimuthal mode decomposition technique to be applied as a part of the NF-FF transformation allowing a reduction in the computational burden of the transformation.
Multi-layer dielectric rod (MLROD) antenna has been shown to provide wideband, dual-polarization, symmetric patterns, and stationary phase center. The key challenge in designing a MLROD antenna is to choose proper thickness and dielectric constant of each layer, and shape of radiation tip to meet desired VSWR, pattern, and phase center requirements. This becomes even more difficult as the number of layers increases for achieving greater bandwidth. This paper discusses a design optimization procedure of a UWB 3-layer MLROD using Genetic algorithm with novel fitness functions for simultaneously controlling reflection coefficient, phase center, and pattern three key characteristics. The final design exhibited excellent desired performance throughout the desired frequency range.
Antenna pattern measurement in the presence of obstacles requires an accurate characterization of the antenna-scatterer interaction in order to retrieve the multipath effects that distort the antenna pattern. In this contribution, a new approach based on the Sources Reconstruction Method is proposed. The idea is to characterize the Antenna-Under-Test (AUT) and the scatterers through a set of equivalent currents that radiate the same fields as the original problem. Thus, it is expected that the equivalent currents retrieved on the surface enclosing just the AUT will provide the AUT radiation pattern. A comparison with modal expansion of the fields on the reconstruction domains is also performed.
The proposed method is based on fiber-optic link connected to the antenna under test in order to measure the radiation pattern of electrically small antennas without any metallic cable effects. The experimental radiation characterization of electrically small antennas reveals significant effects of the RF cable strongly disturbing the antenna gain and radiation patterns for both polarization components. With a very simple use and a low cost, the presented systems can characterize radiation of both narrow and UWB compact antennas in the 0.1-4 GHz frequency band. The proposed test-bench is firstly validated then its performances are compared to classical cable measurements and antenna simulation results. Both measured and simulated results are compared and the agreement is excellent (±1 dB) for co polarisation in the different cut planes. Limitations of the proposed method are also addressed. For very low gain levels (for the presented antennas at 950 MHz and 473 MHz, a cross polarization level below -30 dBi) the 42 x 20 x 23 mm3 autonomous optical receiver is too large and alters the antenna under test radiation. In this last case, and for a using at higher frequency, it is recommended to miniaturize the optical receiver whose size is mainly governed by battery performances (6 hours).
Fig. 1: Cylindrically conformal antenna array system This paper discusses an example of a unique cylindrical conformal array design approach. This exemplary design is composed of six subarrays of series-fed microstrip patch antenna operating in X-band. The diameter of the conducting cylinder is 7.5 inches in diameter. Each subarray contains seven patch antennas connected in series configuration and is connected to a single coaxial probe located at the center element. Such feed arrangement greatly reduces the number of feeding cables, increases the feed-line isolations, and minimizes the feed-line radiation. The elevation pattern of each subarray is controlled by the number of patches and impedance matching tapering via varying the width of connecting microstrip lines. The azimuth pattern is controlled by the subarray height above cylinder surface, substrate width, cylinder radius, and surface treatments between subarrays. Measurement results exhibited good impedance matching and broad antenna coverage in X-band.
The Enhanced Antenna Subsystem (EAS) is a 12 beam, receive only antenna which uses a combination of switched elements and phase delay to accomplish independent beam steering. The upper portion of the domeshaped antenna is populated with 45 circularly polarized antenna elements in an icosahedron pattern along with 15 additional circularly polarized elements along the cylindrical skirt extension. The antenna was tested in our 35’ by 35’ by 65’ compact range. Pattern testing was accomplished by mounting the antenna to a roll positioner atop a high load tower, which was then mounted to an azimuth turntable. The range has a 20’ by 20’ reflector providing an 8’ quiet zone. Using a switching network, we simultaneously characterized 11 statically pointed beams while tracking the range source antenna with the 12th beam. Post processing of the data was performed to separate the beam data and calibrate out losses through the switching network.
In this paper, a beam-steering computer design is explored for a large space-fed phased-array antenna. GTRI previously developed a beam-steering computer for a smaller phased-array antenna which accomplished spherical propagation focusing and multiple phase-only beam-broadening modes. In a subsequent effort, the beam-steering computer design was scaled for a large phased-array antenna to accomplish similar tasks. To verify the design, a series of far-field measurements was initiated to characterize the performance of the antenna by comparing with past measured near-field data and modeled results. One of the primary responsibilities of the beam-steering computer was the focusing of the spherical propagation wave front. A measurement technique is discussed to accomplish this focusing for the large space-fed phased-array antenna by correcting measurement errors in the spherical propagation routine of the beam-steering computer. Additional patterns were taken using the updated feed horn focal point for spherical propagation correction. By correcting the phase errors caused by spherical propagation defocusing in the original beam-steering computer, significantly better antenna performance was obtained, including higher peak gain, reduced nearby sidelobe levels, and removal of beam-pointing errors. Another important responsibility of the beam-steering computer was the ability to realize multiple antenna modes, including a focused pencil beam as well as defocused broadened-beam modes. A stochastic gradient descent algorithm was utilized to obtain several phase tapers to accomplish beam-broadening for the antenna modes. These modes were implemented in the beam-steering computer and tested on a far-field range. The antenna patterns were compared with modeled results and with previous measured data to ensure validity of the implementation.
Calibration of probes for planer near field range measurements is generally required to obtain accurate cross-polarization (xpol) data; however, probe calibration is costly and time consuming. Using analytical models in place of calibration is generally much more cost effective, but may result in larger measurement errors. In a previous paper [1], we showed that simple models of copol probe patterns with zero xpol can give accurate measured results, provided that the probe xpol is much better, generally 10-15 dB better, than the Antenna Under Test (AUT). The next question is “Can a lower performing (and cheaper) probe be used if both the copol and xpol probe patterns are modeled?” In this paper, we compute AUT xpol measurement errors that result from probe xpol errors, and we compare far field AUT patterns processed using probe models with patterns processed with calibrated probe files.
One of the most promising bands for long-range radiocommunications is the Ka-band (25-40 GHz). This is due to the existence of a natural radio transmission window around 30 GHz. Both terrestrial and satellite transmission systems are planned on this frequency band. For satellite applications, circular polarization is needed and the antennas or antenna arrays must frequently exhibit specially tailored radiation patterns. This paper proposes an efficient planar element for Ka-band telecom and remote sensing applications. The element has reduced losses (and hence good efficiency), while providing circular polarization (AR better than 3 dB) and good matching (better than 10 dB) in the 25.5-27.0 GHz frequency band. The element is fed by an integrated low loss transmission line (suspended strip line, SSL). This modular design allows an easy grouping into high efficiency subarrays, which include beam forming networks (BFNs) built in the same SSL technology and an innovative transition to the final standard waveguide feeder.
A spherical near-field measurement range at Nearfield Systems Inc. has recently been used to measure gain, pattern and polarization of a multi-element helix array operating in the UHF band. Verification of gain performance over the operating band was of primary importance and so major efforts were made to obtain the best possible gain results and to quantify the gain uncertainty through a complete error analysis. A single element helix gain standard was first calibrated and the estimated uncertainty in this calibration was 0.35 dB. A double ridged horn was to be used as the probe for the spherical near-field measurements and so the patterns of the horn at all test frequencies were measured on the spherical range using identical horns as the AUT and the probe. From these measurements, probe pattern files were generated that could be used to perform the probe correction in the measurements of the helix gain standard and the multi-element array. The helix gain standard was then installed in a new spherical near-field range at NSI with the double ridged horn as the probe. The range used a specially designed phi-over theta rotator that could support and rotate the array and maintain the required position accuracy. The chamber was lined with 36 inch absorber. Spherical measurements were then performed and the data processed to provide the far-field peak amplitudes at each frequency that were necessary for gain measurements. The far-field peak values are equivalent to the far electric field for the gain standard and are compared to the same parameter for the multi-element array to produce the final gain results. The helix array was then installed in the spherical range and a series of measurements were performed to produce the far-field gain, pattern and polarization results and also to provide the data for the complete 18 term uncertainty analysis. The uncertainty in the gain measurements was 0.45 dB and the axial ratio uncertainty was 0.11 dB.
Pattern distortion due to finite range measurement of antennas in close proximity to electrically large metallic media is examined. The ubiquitous yet arbitrary 2D2/. distance requirement cannot be blindly applied to scenarios where antennas couple to nearby structures. A C-band standard gain horn antenna is analyzed near a circular metallic plate at 6 GHz using the commercial software FEKO. The near-fields are computed at various radii, which are set to multiples of D2/., where D is defined as the largest dimension of the complete structure. The radiating near-field patterns are normalized and compared to the far-field pattern. Results indicate that measurement at 2D2/. may not be necessary. Increasing fractions of D2/. results in a diminishing measurement error that may be tolerable, depending on the intended application.
This paper elaborates on certain aspects of a new measurement process that permits an assessment of spherical near-field (SNF) measurement errors based on a set of practical tests that can be done as part of any SNF measurement. It provides error bars for a measured radiation pattern in an automated fashion.
– far-field transformation with cylindrical scanning are efficiently determined by using an optimal sampling interpolation algorithm. The comparison of the far-field patterns reconstructed from the acquired irregularly distributed measurements with those obtained from the data directly measured on the classical cylindrical grid assesses the effectiveness of the approach.
The Georgia Tech Research Institute (GTRI) analyzed a phased-array antenna for the purpose of testing phase-only defocusing methods. The array is defocused with the objective of broadening its beam at the cost of lower antenna gain. A design for the beam-steering computer is accomplished which adds the capability of focusing a beam, steering in azimuth and elevation, and performing beam defocusing using only element phase. Widening of the beam is accomplished using only 180° phase shifts in the elements, and it is compared with widening accomplished using gradual phase tapers. The antenna is measured in a near-field range to obtain amplitude and phase information as a function of each element in the array. Near-field testing of the antenna is also used to verify the capability of the beam-steering computer; two-dimensional antenna patterns and near-field hologram projections are compiled to prove this functionality. A software model is designed to mimic the behavior of the phased array antenna in its operational modes; it is also used to predict antenna gain and beamwidth prior to near-field testing. Measured and modeled antenna patterns are compared using focused and defocused modes. Metrics are performed on the near-field data to infer statistics of the individual phase shifters and on the computed far-field patterns to characterize the entire antenna. The defocusing methods under analysis are phase-only methods, due to the inability to control amplitude weighting of elements in this antenna. One method discussed uses only 180° shifting of elements in the antenna to achieve a desired beamwidth. This is compared with another method which gradually spoils the beam by applying a phase taper across the aperture. The results from near-field testing compare the defocusing methods and characterize the relationships between gain, beamwidth, and sidelobe levels for both defocusing methods.
Near-field antenna measurements are accurate and common techniques to determine the radiation pattern of an antenna under test. The minimum near-field sampling rate is dictated by the electrical size of the antenna and usually equidistant sampling is applied for planar, cylindrical, and spherical measurements. Certain applications either rely on or benefit from near-field sampling on irregular grids. To handle irregular measurement grids near-field transformation algorithms like equivalent current methods or the multilevel fast multipole accelerated plane wave based technique are required which do not rely on regularly sampled data. In this contribution the plane wave based near-field transformation is applied to spherical, cylindrical, and “combined” near-field measurements employing irregular sampling grids. The performance is assessed by various simulated near-field measurement scenarios.