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.)
Probe correction is required to accurately determine the far-field pattern of an antenna from near-field measurements. At Raytheon Primary Standards Laboratory (PSL) in El Segundo, CA, data acquisition hardware, instrument control software, and a mechanical positioning system have been developed and used with an HP Network Analyzer/Receiver system to perform these measurements. Using a three antenna technique, the on-axis and polarization parameters of a linearly (or circularly) polarized probe are calibrated. The relative far-field pattern of the probe is then measured utilizing the two nominal, orthogonal polarizations of the source antenna. All measurements are stepped in frequency and use a time domain gating technique. The probe and the source antenna are optically aligned to the interface and unique, kinematic designed interface flanges allow repeatable mounting of the antennas to the test station.
The technique is based on a non-redundant and non-uniform representation of the near-field on the measurement plane and performs an estimation of the fields samples outside the measurement region. Thanks to the non-uniformity distribution of the samples, also the estimation of a limited number of them allows a significant improvement in the far field reconstruction. The numerical and experimental investigation presented in this paper confirms the effectiveness and flexibility of the technique, which requires a low computational effort.
The backtransformation in (planar) Near Field processing is often claimed to be a very powerful tool for antenna diagnostics. Less known is a kind of defocusing effect which is introduced by the processing. Selecting the visible space in the Far-Field domain has a similar effect as a bandfilter in the frequency domain of an electric signal. In that analogous case it is better known that after the transform to the time domain, one has to deal with sin(x)/x behavior, limiting the resolution. The mathematics and convolution effects of both the onedimensional time-frequency transform as the two dimensional Near-Field Far-Field transform will be explained. Some measurement procedures are proposed, including S/N requirements. It turns out that the back transformation technique has some nasty properties which limit the use for alignment purposes. Some alternatives are discussed.
The Sirius 2 telcommunication satellite was build in France by Aerospatiale. As a subcontractor Saab Ericsson Space (SES) developed the telecommunication antenna for direct television broadcast. The satellite was successfully launched November 13, 1997. Three antennas were manufactured by SES: a quality model (QM), a flight model (FMl) and a flight spare (FM2). Each antennas consists of a 1.4 meter in diameter shaped main reflector fed by a shaped subreflector and a dual polarized feed horn. For the test of the antennas, spherical near-field antenna test ranges located at Ericsson Microwave System (EMW)/SES in Sweden and at the Technical University of Denmark (DTU) were used. Each of the three antennas was measured twice. Between the two measurements mechanical and thermal tests were performed. The paper presents the measurements on the satellite antennas together with a discussion of the advantages of using the spherical near-field technique for this type of measurements. Compared to a far-field range the advantages are evident: At both SES and DTU a measurement distance of ten and six meters respectively were used on the indoor ranges. On a far-field range a measurement distance in the excess of 250 meters must be applied. To decrease the measurement time the near fields were only measured in a certain region on the near field sphere. The influence of this truncation will be discussed. Coordinate systems for the antennas were defined using mirror cubes. The RF measurements as well as the optical measurements on the cubes were performed without dismounting the antenna from the antenna positioner. The radiation patterns are therefore precisely decined with respect to the coordinate systems of the cubes.
This paper is concerned with the measurement and analysis of a circularly polarized, flat plate patch array receiving antenna at 12.5 GHz. Input impedance and far field pattern measurements of the antenna over the frequency band from 10 to 15 GHz were performed. The small Compact Range (CR) facility of the Ohio State University Electro Science Laboratory OSU/ESL was used to measure the gain pattern. Gain pattern measurement of the antenna was done by using the gain comparison method. A broadband (2-18 GHz), constant phase pyramidal horn antenna was used as a reference. The data were analyzed to determine the radiation efficiency of the antenna.
Ths paper demonstrates the plane wave, pattern subtraction method for performing range compensation of full-sphere antenna patterns measured on a fixed line-of-sight far-field range. The range field is measured on the surface of a sphere and a plane wave model of the range field illuminating the antenna under test (AUT) is determined. The range compensation algorithm uses information contained in both the plane wave model and the AUT pattern measurement to estimate the error pattern that is added to the measured AUT pattern by an extraneous source. This estimated error pattern is subtracted from the antenna pattern measurement to obtain a compensated pattern. The compensated pattern and estimated error pattern are improved iteratively. This paper demonstrates the technique using measured data. The AUT is measured in a far-field anechoic chamber that contains a secondary horn antenna located 20 degrees off-axis from the range antenna, which is used as an extraneous source. The AUT is a 474 element planar array operating at a frequency of 9.33 GHz.
Offset parabolic reflector Compact Ranges are limited for cross polarization measurements in comparison to compensated dual reflector systems. This means that, in some cases, the crosspolar measurements at low levels show a significant content of the compact range reflector cross polar. An investigation has been carried out at INTA to reduce the crosspolarization measurement errors levels to those of a compensated dual reflector system by the application of vector deconvolution techniques. Results are shown of the validation of the algorithm in a far-field range where a crosspolar field is introduced by depointing the transmitter antenna.
We are developing a new indoor bistatic measurement technique for scale model targets. This procedure will collect far-field data at bistatic angles from 60° to nearly 180° and near-field data over a 10' high, 10' radius cylinder surrounding the target. A stationary parabolic reflector illuminates the target while a duplicate parabolic reflector, rotated to its bistatic position, acquires far-field data. The independent, concentrically mounted near-field scanner gathers comparison data. Most compact range reflectors employ shaped edges to avoid edge diffracted signals entering the measurement volume. We report results of using shaped absorber material over otherwise unmodified reflector edges to reduce diffraction. High-resolution 3D images of sample structures demonstrate the practicality of this approach.
Due to the high cost of constructing anechoic chamber, the multi-usage of a chamber in various applications is very effective in terms of cost as well as space. In this paper, we describe an anechoic chamber, currently used at SK Telecom in Korea. This is designed for the measurements of both far/near field antenna and EMC/EMI in the identical chamber. This anechoic chamber and measurement system support antenna test in the frequency range of 150 MHz to 40 GHz and satisfy the requirement of ANSI C63.4 and CISPR16.1for EMC/EMI. The near field measurement system supports planar, cylindrical and spherical methods to test various types of antennas. For the far field and EMC/EMI measurement, the planner near field scanner is hidden by movable absorber wall. The AUT positioner is foldable and can be stored under the chamber floor. Brief description of the chamber and the measurement system with measured results are also provided.
The DREO-DFL Antenna Research Lab (DDARLing), contains far-field and planar near-field antenna measurement ranges. Measurements can be made on both ranges from 1.0 to 62.5 GHz. In the early implementation stages of our antenna measurement ranges, most of our energy was absorbed in mastering the mechanics of the positioners and the intracies of the operation of the software, and addressing component failures. To make useful measurements, it is necessary to minimize system errors. Early experience and frustration has led us to the development of an ordered series of standardized procedures that are aimed at careful set-up, calibration, and operation of the ranges. Within these procedures, attention is paid to the identification and minimization of errors due to alignment, equipment calibration, linearity, leakage, multipath, and drift. Following a brief description of the two ranges in the DDARLing facility, the paper provides details of one of these procedures.
The objective of this study is to evaluate the measurement errors of a near-field range at in order to develop some techniques to minimize them. Measurements were performed on a standard gain horn as references. The methodology presented demonstrates that it is feasible to calculate the far-field radiation from near-field measurement with one deconvolution that will include all the errors introduced by the instrumentation
A general approach for phase-retrieval is discussed. The representation is based on an advanced non-redundant sampling representation and is able to explicitly take into account geometrical characteristics of the source, like the overall dimension and the general shape, as well as a priori inforn1ation on the near-field and far-field.
The Plume Measurement Facility is a new outdoor facility providing the capability to characterize tactical rocket motor plumes. Radar cross section of the plume is measured by both a near field and a far field radar. Infrared/ultraviolet/visible (IR/UVNIS) charac teristics are measured by numerous instruments recording spacial, temporal, and spectral data. All instrumentation is calibrated and adjusted to realtime standard day meteorological data and all data is recorded on a common synchronized time base.
Multipath on far-field ranges causes distortion of pattern measurements. The multipath components can be removed by illuminating the antenna under test with short-duration pulses and applying a time domain gate. Equivalently, the measurements can be made in the frequency domain and transformed to the time domain with the Fourier transform. After gating, the time-domain data are transformed back to the frequency domain, yielding improved CW patterns at discrete frequencies. Virginia Tech has recently added time-domain gating capability to its far-field antenna range. The data acquisition and processing software is implemented using the LabVIEW language, which makes the data acquisition and time-domain processing very easy to control. Practical guidelines for selecting a gate are given. Results are presented for an open-ended waveguide and conical dipole. With wideband antennas, gated patterns show significantly improved symmetry and null depth.
Wideband channel measurements have been used extensively to determine path loss and time dispersion characteristics of radio channels (e.g., [1], [2], [7]). The principles used to temporally resolve individual received signal components for wideband propagation measu rements can be applied to antenna pattern measu rements to achieve more accurate results. Multipath, a propagation phenomenon which occurs when reflecting or scattering objects exist in an environ ment, causes inaccuracies in measured patterns when narrowband signals (e.g. continuous wave) are used to perform far-field antenna measu rements. Using the wideband technique described in this paper, the effects of multipath can be completely eliminated from pattern measurements. The method described here is especially useful when antenna range dimensions are limited in space or when multipath signal components caused by distant reflectors are irreducible.
Coherent processing using measurements on two probe scan planes with different antenna under test (AUT)-to-probe separations reduces the effects of coupling between the AUT and the probe or, alternatively, reduces the effects of room scatter. The results of these doublet scans can be coherently combined to mitigate one or the other (but not both) of these error terms. For either case, the extraneous signals cancel when the far field patterns from the two planes are coherently combined. The new "quadrille" scan technique coherently combines four separate scan planes which will cancel in one set of pattern measurements both the AUT-probe coupling error and the room scatter error. If either the coupling or the room scatter is much larger than the other, the error reduction attained by the quadrille may not merit the additional measurement time; however if the two terms are comparable the quadrille may be needed to attain precise measurements.
A technique to reduce the scan length in near field antenna measurement is presented. In the technique, the original scan length is selected for a critical angle of 30° 35°. The measured near field probe data is then extrapolated beyond the available probed region. The extended near field probe data is next used to predict the far field pattern of the AUT. The extrapolation is carried out by estimating the aperture distribution from the measured probe data. The aperture size, the separation between the AUT and the probed plane and the orientation of the probed plane with respect to the AUT are selected such that the aperture distribution leads to the minimum error between the measured near field probe data and the near field due to the aperture distribution.
Andrew Corporation, founded in 1937 and headquartered in Orland Park, Illinois, has evolved into a worldwide supplier of communication products and systems. To develop a superior, high performance line of base station products for a very competitive marketplace, several new antenna measurement systems and upgrades to existing facilities were implemented. This engineering project developed an indoor test range facility incorporating design tool advantages from among Andrew Corporation's other antenna test facilities. This paper presents a 22-foot vertical by 5-foot diameter cylindrical near-field measurement system designed by Nearfield Systems Incorporated of Carson, California. This system is capable of measuring frequencies ranging from 800 MHz to 4 GHz, omnidirectional and panel type base station antennas up to twelve feet tall having horizontal, vertical or slant (+/- 45 degree) polarizations. Far-field patterns, near-field data and even individual element amplitude and phases are graphically displayed.
A sophisticated software package FARANA (FAR-field ANAiysis) is presented for transforming planar near-field test data to far-field antenna patterns, including enhanced analysis of far-field results. FARANA is coded in MATLAB version 5.0. MATLAB (MATrix LABoratory) is an interactive mathematical modelling tool based on matrix solutions without dimensioning. Using MATLAB, numerical engineering problems can be solved in a fraction of time of time required by programs coded in FORTRAN or C. FARANA operates with a state-of-the-art graphical user's-interface, is intuitive to use and features high speed and accuracy. This paper addresses an assessment of the program, discusses its use and enhanced far-field analysis capabilities.
We describe a technique which uses field intensity patterns formed by the interference of an unknown test antenna and a known referenceantenna - holograms in the classical optical sense - for determining the far-field pattern of the unknown antenna. The field intensity is measured by acquiring an infrared picture of the tem perature distribution on a resistive screen heated by incident microwave energy. The output of the camera is processed to yield the electric field intensity on the surface of the resistive screen. Required measurements are the field patterns of the unknown antenna and two holograms taken with relative phase differences between the reference and unknown antennas of 0° and 90°. In addition, the amplitude and phase of the reference field at the measurement plane are needed. These can be obtained from a separate measurement of the reference using standard near-field techniques. The algorithm gives the complex near field of the antenna under test which can then be processed to obtain the far-field pattern of the antenna under test. We present results showing far-field patterns which acceptably reproduce the main beam and near sidelobes. Such techniques will allow rapid testing of certain antenna types.