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.)
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.
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.
Nearfield Systems Incorporated (NSI) provides antenna measurement systems to domestic and foreign, commercial and government customers with sophisticated requirements that demand custom solutions for RF, mechanical, thermal or software applications. NSI is continuously adapting existing designs to seek cost effective solutions for each customer's demanding specification. This paper discusses numerous near-field scanner designs to meet a variety of applications. Presented are designs for several vertical planar scanners, horizontal scanners, tilted planar scanners, and special scanners designed to attach to structures to test antennas in-situ.
Far-field range testing has been the standard at the Southwest China Research Institute of Electronic Equipment (SWIEE) and at other facilities in mainland China. SWIEE has recently commissioned a new spherical near-field measurement system from Nearfield Systems Inc. (NSI) and Hewlett Packard (HP) to improve its antenna measurement capability. The near-field system provides significant advantages over the older far-field testing including elimination of weather problems with outdoor range testing, complete characterization of the antenna, and improved accuracy. This paper will discuss the antenna types at SWIEE tested with the NSl/HP near-field system, and the results being achieved.
A large horizontal near field measurement facility has been validated and commissioned at Lockheed Martin's Sunnyvale, CA facility. The new measurement facility will be used for characterizing antennas for a variety of satellites over a frequency range of 1 26.5 GHz. A horizontal near field scanner with a 14m x 7.8m (46' x 26') effective scan area has been designed to allow for 9.8m (32') of vertical clearance permitting zenith oriented satellites to be easily positioned within the range and tested in an efficient manner. The facility will soon support the measurement of antennas that are in a vertical orientation. This is accomplished with a novel add-on that allows vertical planar near field scanning on the same range. The vertical scanner has an effective coverage area of 13.6m (45') horizontal x 9m (30') vertical. The system is being used to test commercial communications satellites.
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.
Gabor holography is an appropriate technique for near field measurements at THz frequencies when apertures of the order of thousands of wavelengths are involved. The method permits pattern prediction over a restricted angular range from intensity measurements, providing a direct method of recovering phase which overcomes cable, planarity and atmospheric effects; problematic to conventional near-field phase measurements. We demonstrate the feasibility and convenience of the method with an example planar near-field measurement at 94GHz for a 1.1m Cassegrain reflector and we determine the relationships governing dynamic range and the requirements for sampling. Finally, two-dimensional numerical simulations for a lm antenna at 0.5THz, with a 10m scan distance, will be presented to demonstrate the feasibility of the method for large terahertz antennas.
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.
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.
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.
Measurements of antennas with integrated electronics is an upcoming topic. In many cases the antennas can only work in pulsed mode which requires synchronisation between radar and measurement equipment. Up and down mixing by internal LO's causes additional problems, especially with Near Field measurements where amplitude and phase data is required. Based upon hands-on experience, this paper treats some of the problems and pitfalls related to the Near Field measurements of an active antenna and alignment of the elements by means of backtransformation of the data.
Recently, several deployable, ground-to-ground col lection systems have been developed for the assessment of aircraft RCS on the flight-line. The majority of these systems require bulky rail or scanning hardware in order to collect diagnostic imaging data. The measurement technique described in this paper, while not a "cure-all", does eliminate the need for bulky hardware by allowing the collection system to move freely around the target while collecting radar backscattering data. In addition, a nearfield-to-farfield transformation (NFFFT) algorithm is incorporated in the process to allow the collection of scattering data collected in the near field to be processed and evaluated in the far field. The techniques described in this paper are a part of a data conditioning process which improves the data quality and utility for subsequent analysis by an automated diagnostic system described elsewhere in this proceedings [1]. The techniques are described and demonstrated on numerically simulated and experimentally measured data.
A time domain near-field far-field transformation technique based on a non redundant plane-polar sampling representation of the field is presented. The method allows to obtain the far-field with a minimum number of samples and/or a reduction of the scanning area. Various computational schemes are presented.
A new holographic technique for array diagnosis is discussed. Numerical and experimental results are shown in case of linear arrays and compared with the performance of the “classical” holographic approach based on FFT. Furthermore, the working progress on a sub-optimal algorithm specifically developped for diagnosis of large arrays is presented.
We describe a technique which employs amplitude-only measurements of an unknown antenna combined with a synthetic reference wave to produce a hologram of a near-field antenna distribution. The hologram, which may be recorded by amplitude-only receiving equipment, is digitally processed using an enhanced theory which allows complete removal of the spurious images normally encountered with optical hologram reconstruction. The recovered near-field data are then processed using standard algorithms to calculate antenna far-fields. We present the theoretical formulation and results of measurements obtained on an 1.2 m reflector antenna.
We describe a technique which employs amplitude-only measurements of an unknown antenna combined with a synthetic reference wave to produce a hologram of a near-field antenna distribution. The hologram, which may be recorded by amplitude-only receiving equipment, is digitally processed using an enhanced theory which allows complete removal of the spurious images normally encountered with optical hologram reconstruction. The recovered near-field data are then processed using standard algorithms to calculate antenna far-fields. We present the theoretical formulation and results of measurements obtained on an 1.2 m reflector antenna.
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.
This paper reports on the results of computer simulations of planar near-field test-zone-fields. Techniques for the improvement of the quality of the test fields are presented and demonstrated. These techniques include the use of larger scan areas and the use of window functions applied to the measured near-field data. Test-zone-field quality is measured by the angular spectrum of the error of the test-zone-field as compared to an ideal plane wave test-zone-field. This investigation sought the minimum scan length, L, for a given critical angle, ?c and separation, S. It is shown that significant improvements in test-zone-field quality can be realized if the test zone is extended from the standard length, Ls=D+2S(tan(?c)) by an amount 20?/cos(?c). This scan length is approximately 30? larger, for a critical angle of 50 degree and 60? larger, for a critical angle of 70 degrees, than the standard length. A raised cosine amplitude/quadratic phase window applied to the measured near-field data can significantly reduce scan length requirement while maintaining the increased accuracy of the extended scan length. The recommended scan length with window is given by Lw=D+2S(tan(?c))+2W, where W is the length of the window applied to each end of the scan measurements. The window description and required length are presented.