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.)
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.
When a planar near-field measurement is done, errors are introduced due to imperfections in the mechanical and electrical parts of the measurement equipment. In order to identify the characteristics of different types of errors, a MatLab program that simulates the near-field from an antenna has been developed. The near-field is transformed to far-field and the errors are evaluated. This paper looks into four different error types: 1) Truncation errors (if the measurement surface is to small the near-field will be truncated before it reaches adequately low levels), 2) Probe-AUT distance errors (fluctuations in the probe AUT distance over the measurement surface), 3) Zigzag errors (due to data being acquired during both travel directions of the probe), 4) I,Q amplification errors (different amplification for the I and Q channels in the receiver). The results are presented in plots which illustrate where in space the largest antenna pattern errors occur.
A series of measurements to validate the performance of a Vertical Planar Near-field Antenna Test Range located at the Hughes Space and Communications Company (HSC) was performed. These measurements were made as part of a task to provide validation of this particular range for detailed Production Antenna Testing. This validation was required in preparation for measuring a particular flight antenna. The range validation consisted of a series of self comparison tests and far-field range pattern comparison tests using an offset reflector antenna as the validation antenna. This antenna had been previously measured on a far-field antenna range which is in constant use to test flight antennas. This paper describes the range validation tests and presents some of the results. Comparisons of some far-field patterns measured on the validation antenna at both the far-field and near-field ranges is presented.
This paper presents a new technique for performing range compensation of full sphere antenna patterns measured on fixed line-of-sight antenna ranges where pattern measurements are made over a spherical surface. Such ranges include far-field, compact, and spherical near-field ranges. A plane wave model of the range field illuminating the antenna under test (AUT) is determined as described in another paper. This plane wave model consists of a small, selectable number of plane waves. Equations are given describing the transformation of range coordinates to AUT coordinates. This allows the response of an AUT to a plane wave from an arbitrary direction to be defined using only the far-field pattern of the AUT. The error pattern added to the pattern measurement by the extraneous plane waves is then estimated using the plane wave model and the measured pattern. This error pattern is subtracted from the antenna pattern measurement to obtain a compensated pattern. The compensated pattern and error pattern are improved iteratively. This paper demonstrates the technique using simulated data. The rotation of the spherical AUT grid with respect to the range grid during the measurement requires an interpolation of the measured fields to estimate the error pattern. Investigations of interpolation error are presented. The computational complexity of the compensation algorithm, excluding the plane wave model, is on the order of the number of measurement points on the spherical measurement grid. K
Cellular and PCS basestation antennas are basically arrays with highly directive elevation patterns and broad azimuth patterns. This causes measurement problems because they are large but not directive in both principal planes. As a result, the pattern measurements of these antennas that have been performed outside have been unreliable in many cases because they are very receptive to interference and range clutter. Thus, one wants to move inside but the antenna size can significantly impact the overall range cost. This paper describes a very practical solution to this problem. Since basestation antennas are long and narrow, one can use a near field scanner approach to deal with the length. In fact by using a sectorial horn probe, the narrow dimension of the antenna-under-test is illuminated by a cylindrical wave. Thus, the scanner need only probe the field along the antenna length. This linear scan data can then be transformed to generate the desired far field elevation pattern. The details of this novel design will be described as well as the results, to illustrate the system capability and accuracy.
A technique was developed to recover the near-field function on a larger data set than the one that is measured. It requires the preliminary determination of functions containing the information relating the two data sets. The simplest way of obtaining such a function is to measure the near-field function on the larger and the smaller data set. This seems to be a drawback to the technique. However. after making one such pair of measurement it is therefore non necessary to do so again and the field of the antenna can be obtained, from the smaller data set measurement, with comparable accuracy. The technique is somewhat different when compensating for a sampling rate reduction. However, in both cases an analytical extension is required to fill the desired domain of definition, followed by a division. In the case of the sampling area the division of the spectral functions f2 by f1 is made in the spectral domain while in the case of the sampling rate the division of the near-field functions E' by E is made in the near-field domain. An experiment was performed to demonstrate the applicability of the above technique. A full near-field measurement of a linear array antenna was performed and processed, then after displacement of the antenna, measurements were done, in one case, on a truncated smaller scan area, and in another case with a larger sampling interval. The technique was applied to recover the complete far-field characteristics of the antenna from the smaller data set. The far-field characteristics of the antenna obtained by this technique were shown to be very similar to the results obtained from a more complete near field measurement.
A method is presented for computing far field antenna patterns from spherical near field antenna measurement data. The new method utilizes a novel Uniform Geometrical Theory of Diffraction (UTD) based transformation of spherically scanned antenna tangential electric (or magnetic) near field measured values to more efficiently obtain the antenna far field. Examples illustrating the accuracy and speed of UTD based spherical near to far field transformations for large to moderately large antennas are presented.
The far-field parameters of an antenna are obtained from near-Field measurement with an accuracy that is limited by the sampling area and the sampling rate used to collect the measurement data. It is therefore important to know the relation between the far-field parameters and the sampling parameters. A parametric study of the far field parameters accuracy versus the sampling parameters was made. In order to determine the optimal choice of the sampling parameters to achieve the desired far-field accuracy, planar near-field measurements of a linear array were performed in an anechoid chamber at the Canadian Space Agency. A program performing Fast-Fourier Transform was used to process the data and to obtain spectral domain and reconstruct the far field patterns. A methodology developed in [1] was used to compare different spectral and far field patterns obtained from different sampling conditions. Parametric curves were developed for the far-field parameters such as gain, beam pointing, beam width, sidelobes, etc.
We review the development and recent performance results of a stand-alone fiber optic based EM field sensor system. The sensor heads are miniature (lcm), electrically passive, and are directly coupled to optical fibers at the remote sensing site. Sensor conversion of EM fields to optical intensity is carried out by mounting small antenna structures directly onto high speed lithium niobate electro-optic modulator chips. Optical power to the sensor head is derived from a stabilized laser which is located within a system chassis at a control room location. Sensor and fiber temperature drift effects are compensated by specialized remote bias control electronics. Recent broad spectrum tests have demonstrated a system bandwidth of about 20 GHz, and a minimum detectable field in the lO's of mV/m. Ultra wideband pulse measurements have demonstrated real time pulse signals of about 2 Vpp for 3 KV/m fields. The sensor system is slated for application in EMI effects such as EM compatibility, and for pin-point near-field and far-field mapping of radiation patterns. The technology is readily scaleable to frequencies exceeding 20 GHz.
Ell has been extensively involved in the development of advanced processing techniques (APT) to improve the quality and utility of both indoor and outdoor RCS/ISAR measurements. These include algorithms for removal of clutter, RFI, and targetsupport contamination (including interactions), prediction of far field RCS from near field measurements, suppression of multipath contamination, and extraction of scattering features/components. These techniques have been implemented in a framework based on ERIM International's IMGMANIP signal/ image processing toolbox and stream input-output (SIO) data flow paradigm. This paper describes a recently-developed Graphical User Interface (GUI) which incorporates the most mature and frequently-used APT algorithms.
Based on the method of moments (MoM), a network model algorithm perturbs the linearized electromagnetic interaction model (the admittance matrix) of a simulated target to match an actual measured data set in the least squares sense, resulting in a more accurate interaction model for the physical target. Since the admittance matrix is independent of source location this technique is amenable for use as a near field to far field transform algorithm. In this ansatz, a MoM admittance matrix is perturbed to match a set of near field measurements, then the far zone field is predicted using the perturbed admittance matrix multiplied by the appropriate far field measurement vector. This paper describes the application of the MoM network model technique to measured and numerical data for ten and twenty wavelength conespheres. Initially, a discussion is given of the code modifications necessary to adapt JRMBOR for network model use. A validation is then provided using "perfect" numerically-generated near field data to perturb an admittance matrix rendered inaccurate through a deliberate undersampling of the conesphere geometry. Finally, results are given for the MoM network model algorithm with measured near field data, with the resulting predictions compared to measured far field truth. Algorithm performance is examined as a function of frequency for monostatic near field input data.