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 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.
This paper examines antenna measurement errors attributable to instrumentation, and their effect on measurement uncertainty.
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 comparison of three near-field position error correction techniques has been performed on simulated near-field data. The purpose of this study was to evaluate the allowable positional tolerances required for planar near-field scanners. Simple k-correction, extended k-correction, and Taylor series correction were applied to computed near-field data contaminated with various kinds of errors, including position errors in one and three dimensions, and random electrical noise. Ideal and error contaminated near-field data were computed for small-size, mid-size, and large-size arrays. Probe position errors up to one-quarter wavelength in each axis and one wavelength in a single axis were used. Probe position error correction was performed using all three methods, and the results were evaluated
The mechanical rotator must be correctly aligned and the probe placed in the proper location when performing spherical near-field measurements. This alignment is usually accomplished using optical instruments such as theodolites and autocollimators and ideally should be done with the antenna under test mounted on the rotator. In some cases it may be impractical to place the alignment mirrors on the AUT or optical instruments may not be available. In these and other cases, it is desirable to check alignment with electrical measurements on the actual AUT and probe. Such tests have recently been developed and verified. Appropriate comparison and analysis of two near-field measurements that should be identical or have a known difference yields precise measures of some rotator and probe alignment errors. While these tests are independent of the AUT pattern, judicious choice or placement of the antenna can increase the sensitivity of the test. Typical measurements will be presented using analysis recently included in NSI software.
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.
Scientific-Atlanta has recently begun work on a large 55 ft.(W) x 45 ft.(H) compact range reflector. The reflector is a Model 5738 with a 45 ft. focal length and a 38 ft. diameter by 38 ft. long cylindrical quiet zone. Due to the large size of the reflector, it is necessary to form the surface as several large, independent sections and assemble and align the reflector at the installation site. The 5738 reflector is shown in Figure 1 with the 38 ft. quiet zone superimposed. Figu re 1. Front View of 5738 Reflector Showing Sections The independent and predictable behavior of large sections proves to be very beneficial for performing an electrical alignment of the reflector based on field probe phase data. This paper discusses the required alignment tolerances and analytic tools developed to predict the effects on quite zone performance due to alignment errors in the sections of the reflector.
Phase space representation (PSR) is introduced as a diagnostic tool for near-field antenna measurements. The PSR of a linear scan is defined as a two dimensional function of position and wavenumber. This combined spatial-wavenumber distribution can reveal features which are not directly visible in either spatial- or wavenumber-domains. It is shown that PSR is useful for both error diagnostic and compensation of certain errors. In particular, the benefits of PSR in identifying aliasing, spatial errors, multiple AUT-probe reflections, and random errors in amplitude and phase will be demonstrated. The capability of the PSR in compensation of contaminated measurements is demonstrated by examples.
Calibration of monostatic radar cross section (RCS) has been studied extensively over many years, leading to many approaches, with varying degrees of success. To this day, there is still significant debate over how it should be done. In the case of bistatic RCS measurements, the lack of information concerning calibration techniques is even greater. This paper will present the results of a preliminary investigation into calibration techniques and their suitability for use in the correction of cross-polarization errors when data is collected in a bistatic configuration. Such issues as calibration targets and techniques, system stability requirements, etc. will be discussed. Results will be presented for data collected in the C and X bands on potential calibration targets. Recommendations for future efforts will also be presented.
The image-based near-field to far-field transformation is based on a reflectivity approximation that is commonly used in ISAR imaging. It is a limited but computationally efficient transform whose accuracy, for appropriate targets, rivals that of computationally more intense transforms. Previous results include applications of the transform to lOA. long wire and lOA. long conesphere numerical data. Here, 1-D and 2-D versions of the transform are applied to conesphere near-field measurements data and the results are compared to corresponding far-field measurements data. Transform errors obtained for these data are compared to corresponding results obtained using newly generated near-field and far-field numerical data. The image-based transform is believed to be especially applicable to the far-field correction of near-field measurements of complicated targets like aircraft or vehicles that are too large or too poorly defined to be simulated numerically.
One approach to getting near-field ISAR measurement costs down is to dispense with track or turntable, and instead mount the SAR antenna on a vehicle and simply 'drive by' the target of interest, which might be a vehicle or aircraft standing on tarmac. In this situation the antenna path will depart from a perfect straight line or circular arc, and there will also be vibrational wobble at the antenna phase center. These effects can defocus the images obtained. One way to overcome the focus problem is to mount strategically placed corner reflector(s) in front of the target, each in a different range cell. These then act as phase references, used to refocus the image. However it is not strictly necessary to employ reflectors - a good focus can normally be obtained by suitably processing just the target return itself. This paper will describe autofocus procedures which have been sucessfully used in conjunction with chirp radar data, in the 'driveby' situation.
The DTU-ESA Spherical Near Field Antenna Test Facility in Lyngby, Denmark, which is operated in a cooperation between the Danish Technical University (DTU) and the European Space Agency (ESA), has for an ex tensive period of time been used for calibration of Standard Gain Horns (SGHs). A calibration of a SGH is performed as a spherical scanning of its near field with a subsequent near-field to far-field (NF-FF) transformation. Next, the peak directivity is determined and the gain is found by subtracting the loss from the directivity. The loss of the SGH is determined theoretically. During a recent investigation of errors in the measurement setup, we discovered that the alignment of the antenna positioner can have an extreme influence on the measurement accuracy. Using a numerical model for a SGH we will in this paper investigate the influence of some mechanical and electrical errors. Some of the results are verified using measurements. An alternative mounting of the SGH on the positioner which makes the measurements less sensitive to alignment errors is discussed.
For frequencies above 30 GHz, RCS reference target method is, in general, more accurate than scanning the field by a probe. Application of mechanically calibrated targets with a surface accuracy of 0.01 mm means that the phase distribution can be reconstructed accurately within approximately 1.2 degrees across the entire test zone at 100 GHz. Furthermore, since the same result can be obtained for both azimuth and elevation patterns, all data is available for the characterization of the entire test zone. In fact, due to the fact that the reference target has a well known radar cross-section, important indication of errors in positioning can be obtained directly from angular data as well. In the first place the data can be used in order to recognize the first order effects (+/- 5 degrees in all directions). Applying this data, defocussing of the system reflector or transverse and longitudinal CATR feed alignment can be recognized directly. Furthermore, mutual coupling can be measured and all other unwanted stray radiation incident from larger angles can be recognized and localized directly (using timedomain transformation techniques). Inmost cases even a limited rotation of +/- 25 degrees in azimuth and +/- 10 degrees in elevation will provide sufficient data for analysis of the range characteristics. Finally, it will be shown that sufficient accuracy can be realized for frequencies above 100 GHz with this method.
A waterline bistatic algorithm, based on the exact near field to far field transformation (NFFFT) and previously exercised on numerical data, is here applied to actual measured data taken at a traditional RCS range reconfigured for near field measurements. The resulting far field predictions for a lOA and 20A conesphere were initially worse than expected. Further examination of the data yielded two important observations. First, the data were found to have relative alignment errors from set to set, leading to a significant broadening of the predicted far field peaks. Second, a few data sets exhibited a constant phase offset inconsistent with the other measured data. This paper discusses the detection of the data misregistration issues highlighted above, along with their ad hoc correction. Predictions are give for the waterline bistatic NFFFT algorithm applied to the measured near field data, both before and after the corrections have been applied. The results are compared with analogous results for numerical input data.
The RATSCAT radar cross section (RCS) measurement facility at Holloman AFB, NM is working to satisfy DoD and customer desires for certified RCS data. This paper discusses the low frequency characterization of the RATSCAT VHF/UHF Measurement System (RVUMS). The characterization was conducted on a portable pit with a 30' foam column at the RAMS site. System noise, clutter, backgrounds and generic target measurements are presented and discussed. Potential error sources are examined. The use of background subtraction and full polarimetric calibration are presented. Potential errors, which can occur from using certain cross-pol calibration techniques, are discussed. The phase relationship between each polarization components of the scattering matrix and cross-pol validation techniques are considered.
Simulation of the antenna measurement errors caused by scattering of range clutters is presented in this paper. The Uniform Geometrical Theory of Diffraction (UTD) based NEC-Basic Scattering Code is used to simulate the measurement of antenna in a far-field range where structure scatterers present. It is known that these errors which come from various directions will impact the antenna under test differently dependent on the characteristics of the antenna under test. With the available computer codes, one can simulate and study various ranges in order to better understand the characteristics of the ranges and properly adjust, modify, and improve the facility such that better measurement results can be obtained.
The development* of a real time electronic system to accurately measure the pattern of high gain, ultralow sidelobe level antennas in the presence of multipath scatterers is described. Antenna test ranges and anechoic chambers contain objects that scatter the signal from the transmitting antenna into the main beam of a receiving antenna under test (AUT), thereby creating a multipath channel. Large measurement errors of low sidelobes can result. The fabrication of a feasibility demonstration model Antimultipath System (AMPS) is complete. This AMPS uses a 10 MHz wide phase-shift-keyed spread spectrum modulated signal to illuminate the rotating AUT to tag each multipath by its delay. The novel receive section of the AMPS sorts out each multipath component by its delay and adaptively synthesizes a composite cancellation waveform (using delay, amplitude, and phase estimates of the scattered components) which is subtracted from the total signal received by the AUT. After subtraction the resultant is the desired direct path signal which produced the free space pattern of the AUT. Laboratory and antenna range test results are presented and show the promise of measuring sidelobe levels 60 dB below the main beam.
Gain calibration of circular horns and radiation pattern integration applying patterns in two principle planes only is accurate and does not require large computational or measurement effort. This technique is thus more practical than the integration over the entire angular domain, required in case of rectangular horns. However, for many types of AUT’s, additional errors may occur due to the differences in aperture size of the AUT and standard gain horn. The AUT will in many cases have physically larger aperture dimensions. Consequently, unknown test-zone field variations across this aperture can result in additional errors in gain determination. The new method uses a flat plate as a reference target. An RCS measurement of the flat plate is used to derive test-zone field characteristics over the same physical area as the AUT. Combined with the accurate gain calibration described above, field information is available over the entire area of interest and the accuracy in gain determination is increased. In this paper, experimental results and practical considerations of the method will be presented.
Planar near-field antenna measurements have developed into a mature science. Nonetheless, unique difficulties arise when measuring some modern antennas, such as high gain satellite antenna systems. In a typical planar near-field measurement, the antenna under test (AUT) has a collimated beam in the near-field which is perpendicular to the scan plane (i.e. the AUT boresight is parallel to the normal of the scan plane). On the other hand, the scan plane is positioned close to the AUT to maximize the valid angular range in the far-zone patterns. Unfortunately, it is not always possible to place the AUT very close to the scan plane and keep the near-field beam perpendicular to the scan plane. An investigation of the benefits and pitfalls of a planar near-field measurement where the AUT beam is not perpendicular to the scan plane is presented. The measurements of antennas tilted 45 degrees with respect to the scan plane normal are used as examples. With this atypical arrangement, some of the usual errors in a near-field measurement are emphasized. Procedures to identify and reduce these errors will be presented.
The standard planar near-field to far-field transformation method requires data points on a plane-rectangular lattice. In this paper we introduce a transformation algorithm in which measurements are neither required to lie on a regular grid nor are strictly confined to a plane. Computational complexivity is O (N log N), where N is the number of data points. (Actual calculation times depend on the numerical precision specified and on the condition number of the problem.) This algorithm allows efficient processing of near-field data with known probe position errors. Also, the algorithm is applicable for other measurement approaches, such as plane-polar scanning, where data are collected on a non-rectangular grid.