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 major drawback of Spherical Near-Field (SNF) measurements is the comparatively long measurement time, since the scanning of a whole sphere enclosing an Antenna Under Test (AUT) is required to calculate the Spherical Mode Coefficients (SMCs) required for the computation of the far field. Since the SMCs prove to be sparse under certain conditions, efforts have been made to apply compressed-sensing techniques to reduce the measurement time by acquiring a smaller number of sampling points. These approaches have been successfully tested in simulation using classically acquired measured data. This decouples the measurements from practical problems, such as basis mismatch due to the finite precision of the mechanical positioner and environment effects. In this paper, results from a sparse data acquisition performed with a physical system are reported. To decouple the error introduced by the approach itself from the error introduced by non-idealities in the measurement system, an AUT is measured using both traditional near-field sampling and compressed near-field sampling. The classically acquired data is used both as reference and as source to simulate a synthetic compressed measurement. The effects introduced by real considerations are calculated by comparison between the synthetic compressed measurement and the acquired one, while the error of both is evaluated by comparison to the reference measurement. The results further demonstrate the viability of this method to accelerate SNF measurements and pave the way for further research.
This paper introduces the new Portable Laser Guided Robotic Metrology (PLGRM) system at the National Aeronautics and Space Administration's (NASA) Glenn Research Center. Previous work used industrial robots in fixed facilities to characterize antennas and required fixtures that do not lend themselves to portable applications. NASA's PLGRM system is designed for in-situ antenna measurements at a remote site. The system consists of a collaborative robot arm mounted on a vertical lift and a laser tracker, each on a mobile base. Together, they enable scanning a surface larger than the robot's reach. To accomplish this, the robot first collects all points within its reach, then the system is moved and the laser tracker is used to relocate the robot before additional points are captured. The PLGRM implementation will be discussed including how safety and planning are combined to effectively characterize antennas. Software defined triggering is a feature, for flexible integration of vector network analyzers and antenna controllers. Lastly, data will be shown to demonstrate system functionality and accuracy.
The Three-antenna polarization measurement technique is used to determine the axial ratio, tilt angle and sense of polarization of three antennas from measurements on each of three antenna pairs. The three antennas are generally nominally linearly polarized and the measurement data consists of the change in amplitude from the initial antenna orientation where they are co-polarized to the orientation where one of the antennas is rotated about its axis to the null amplitude position. The sign of the phase change is also noted and the phase change at the null position is known from theoretical calculations to be either plus or minus 90 degrees. The correct sign is determined from the sign of the phase change. For antennas with axial ratios in the range of 50 to 80 dB that will be used as near-field probes or as feeds for reflector antennas, it is imperative to measure the polarization parameters as accurately as possible. The primary source of uncertainty in the measurement is due to scattered signals in the measurement range that arise from multiple reflections between the two antennas and from the absorber on the chamber walls. For antennas with very large axial ratios, the scattered signals can be larger than the true measurement signal. These scattered signals can change the sign of the phase and produce large errors in the amplitude at the null. If the separation distance between the antennas is adjusted after rotating to the null to produce a maximum amplitude, the scattered signal is in phase with the true measurement signal. If the distance is adjusted for the minimum at the null, the scattered and true signals are out of phase. Measurements at these two positions will produce the best measurement of the phase sign and the true amplitude. But if measurements are being performed at a number of frequencies, the maximum and minimum amplitude positions will be different for each frequency, and this will complicate automated multifrequency measurements. New improvements have been developed in the details of the measurements that greatly improve the determination of the phase sign and the amplitude at the null for multiple frequency measurements and these will be described and illustrated in the following paper. With these improvements, the estimated uncertainty of a 60 dB axial ratio is on the order of 1.8 dB. A new technique has also been developed to improve the source correction of the pattern data for probes with large axial ratios that guarantees that the on-axis polarization of the pattern data will be identical to the results of the Three-antenna measurement. The probe correction processing will then produce the highest accuracy results for the polarization of the AUT.
A dual linear polarized 3D printed magneto-electric phased array antenna for various 5G New Radio (NR) frequency bands is proposed and its beam steering performance is investigated. The magneto-electric radiating element exhibits a well-defined stable pattern quality, low variation in the impedance over a wider bandwidth and high port to port isolation in a dual polarization configuration. The analog beamforming network (BFN) of the array is also designed. The fabricated board will be combined with the 3D printed array aperture for experimental verification of the scan performance.
There are certain applications where the use of electro-optical (EO) probes to acquire near-field measurements can provide major advantages as compared to conventional RF measurement techniques. One such application is in the area of high power RF measurements that are required for calibration and test of active electronically scanned arrays (AESA). The family of EO probes presented herein utilizes the Pockels effect to measure the time-varying electric fields of the antenna under test (AUT). The use of a non-invasive, broadband EO probe facilitates measurement of the tangential electric field components very close to the AUT aperture in the reactive near-field region. This close proximity between the AUT and the measurement probe is not possible with conventional metallic probes. In this paper, the far field gain patterns acquired using the EO probe will be compared to the corresponding gain patterns obtained from conventional far-field and near-field methods. The measurement results, along with the advantages and disadvantages of the EO system configuration, will be presented.
Radar scattering is typically represented as the RCS of the test object. The term RCS evolved from the basic metric for radar scattering: the ratio of the power scattered from an object in units of power per solid angle (steradians) normalized to the plane-wave illumination in units of power per unit area. The RCS is thus given in units of area (or effective cross-sectional area of the target, hence the name). Note that the RCS of the test object is a property of the test object alone; it is neither a function of the radar system nor the distance between the radar and the test object, if the object is in the far field. Because the RCS of a target can have large amplitude variation in frequency and angle, it is expressed in units of decibels with respect to a square meter and is abbreviated as dBsm (sometimes DBSM or dBm2). In terms of this definition, the RCS of a radar target is a scalar ratio of powers. If the effects of polarization and phase are included, the scattering can be expressed as a complex polarimetric scattering (CPS) matrix. The measurement of the RCS of a test object requires the test object to be illuminated by an electromagnetic plane wave and the resultant scattered signal to be observed in the far field. After calibration, this process yields the RCS of the test object in units of area, or the full scattering matrix as a set of complex scattering coefficients. This paper describes the planned upgrades to the old IEEE Std 1502™-2007 IEEE Recommended Practice for Radar Cross-Section Test Procedures [1]. The new standard will reflect the recent improvements in numerical tools, measurement technology and uncertainty estimates in the past decade.
The aim of the paper is to address a relevant issue in the Near-Field (NF) measurements: the reduction of the measurement time. Generally speaking, for a given hardware, two main directions can be pursued. The first requires the adoption of an optimal field sampling strategy that reduces the number of sampling points, and the length of the scanning path, without impairing accuracy. The second strategy adopts an optimized control system able to exploit at the best the available hardware (scanning system and measurement instrument). Indeed, the latency of the instrument defines the maximum probe velocity during the field acquisition. Accordingly, unlike the conventional continuous scanning, an optimized controller can speed up the scanning by moving the probe along the measurement trajectory with a variable velocity, accelerating and decelerating between two consecutive sampling points, to increase the average speed. However, the use of an optimized controller is fruitful only when the optimized sampling scheme allows large distances between two consecutive sampling locations, to increase as much as possible the maximum probe speed. In this paper, by suitably using both the above strategies, it is proposed a fast NF system, implemented on a microcontroller Arduino Due, an extremely cheap and off the shelf hardware, that is able to handle the scanner and realize the synergy between the optimized sampling and the optimized control strategy. The simulation and experimental results show a dramatic reduction of the measurement time (up to one order of magnitude) with a high tracking precision (also in accordance with the proposed methodology), and of the costs with respect to standard solutions.
Full probe compensation techniques for Spherical Near Field (SNF) measurements have recently been proposed [1-5]. With such techniques, even antennas with more than decade bandwidth are suitable probes in most systems. The abolition of otherwise frequent probe changes during multi-service campaigns is a highly desirable feature for modern measurement applications such as automotive. In this paper, a standard dual-ridge horn with 15:1 bandwidth is investigated experimentally as probe in a SNF automotive range. The accuracy of the probe compensation technique is validated by comparison to standard single probe measurement.
When designing an anechoic chamber, determination of the extent and quality of the quiet zone is crucial. While rigorous simulation methods can be used for this in principle, in practice such methods quickly become too computationally expensive with increasing frequency. In this paper, the authors evaluate a couple of asymptotic approaches based on ray tracing, and quantify their value for anechoic-chamber design.
Spherical near-field measurements are regarded as the most accurate technique for the characterization of an Antenna Under Tests (AUT) radiation. The AUT's far-field radiation characteristics can be calculated from the Spherical Mode Coefficients (SMC), or spherical wave coefficients, determined from near-field data. The disadvantage of this technique is that, for the calculation of the SMC, a whole sphere containing the AUT must be Nyquist-sampled, thus directly implying a longer measurement time when only a few cuts are of interest. Due to antennas being spatially band-limited, they can be described with a finite number of SMC. Besides, the vector containing the SMC can be proved sparse under certain circumstances, e.g., if the AUT's radiation pattern presents information redundancy, such as an electrical symmetry with respect to coordinate system of the measurement. In this paper, a novel sampling strategy is proposed and is combined with compressed-sensing techniques, such as basis pursuit solvers, to retrieve the sparse SMC. The retrieved sparse SMC are then used to obtain the AUT's far-field radiation. The resulting far-field pattern is compared for both simulated and measured data. The reduced number of points needed for the presented sampling scheme is compared with classical equiangular sampling, together with the estimated acquisition time. The proposed sampling scheme improves the acquisition time with a reasonable error.
The National Institute of Standards and Technology (NIST) has measured a WR-8, 3D printed horn at 112.25, 118.75, and 125.25 GHz using the near-field spherical scanning method. The data were processed with both the NIST standard software and the probe-position compensation software. We conclude that the positioning capability of the NIST Configurable Robotic Millimeter-wave Antenna System is so accurate that probe-position compensation is negligible at these frequencies.
This paper discusses design and fabrication of a low cost, combined pressure / thermal test-bench engineered for environmental tests of UAV mounted antennas. Both test-beds are mainly made of commercial of-the-shelf (COTS) parts and in-house made frames. They occupy small space and do not require specific professional skills for operation or high maintenance cost. Measurement setup is designed to reliably reproduce temperature and pressure corresponding to altitudes from sea level to 6000 m (20000 ft) with dynamic load equivalent for 200 m/s (400 knots) of air speed. Experimental results of radome enclosed wideband antenna are presented.
An experimental case of spherical probe-corrected phaseless near-field measurements with the two-scans technique is presented, based on magnitude measurements at two surfaces of the VAST12 reflector antenna performed at the DTU-ESA Facility. Phase retrieval using strictly the directly measured near-field magnitude was unfeasible in this setup, due to the small sphere separation allowed by the probe positioner, which led to incorrect and excessively slow convergence. Phase retrieval with larger separation between spheres has shown remarkable results. For these tests a measured magnitude was used in combination with calculated near-field magnitudes at different (larger and smaller) spheres with larger separations than allowed by the experimental setup. It has been seen that larger separation between measurement spheres improves accuracy of phase retrieval. A measurement with a backprojected measurement with 3 m sphere separation is of particular interest because it can be potentially replicated in the DTU-ESA Facility assuming such range of movement was allowed, while being accurate down to an error of less than-35dB. Measurements with larger spheres show even better accuracy. These good results were obtained with the normal spatial sampling rate for complex measurements and with a very simple Hertzian dipole initial guess, and show the superior performance of spherical phaseless measurements with the two-scans technique, compared to a planar setup.
This paper addresses the MIMO performance of automotive cellular antenna configurations as measured in an OTA (Over-The-Air) chamber. The OTA chamber measurement has been adopted by the cellular communications industry as a way to validate wireless devices such as smartphones, tablets and modems for RF compliance to the Carrier's network requirements. This performance assessment of an automotive cellular antenna in an OTA chamber is an approach to compare to wireless device performance and antenna verses antenna. The parameters that will be compared as a result of the OTA measurements are TRP (Total Radiated Power), TIS (Total Isotropic Sensitivity) and Data Throughput for given signal levels.
In this contribution the signal received by an antenna is understood as an inner product built by the polarization vectors of the involved antennas. By using a suitable unitary transformation the polarization efficiency can be straightforwardly calculated without additional assumptions. By solving an eigenvalue problem given by a unitary operator which represents a rotation, a simple and illustrative interpretation is possible. The formulation is applied to derive the well-known relations of the improved three antenna polarization measurement technique given by Allen C. Newell, which is mainly based on the measurement of relative power levels. Some measurement results and the calculation of the achievable measurement accuracy are presented.
The power transmitted through a waveguide was determined using in-situ atom-based electric field measurements. The field distribution in the waveguide was measured using Rydberg atoms to find the maximum field, which was used to determine the power. For a proof-of-concept, the power of radio frequency fields at 17.86, 19.63, 26.53, and 33.03 GHz were measured in a WR42 waveguide. A section of waveguide was sealed and filled with cesium atoms. These atom-based measurements are self-calibrated and independent of typical power measurement methods.
An algorithm is developed for the extraction of constitutive parameters from bi-layered uniaxial anisotropic materials backed by a conductive layer. A method of moments-based approach is used in conjunction with a previously-determined Green function. Possible challenges related to measurement diversity are highlighted and a possible mitigation path is proposed.
In order to support near-field measurements of automobile antennas in as realistic as possible environments, an equivalent sources based near-field far-field transformation approach for near-field measurements above a possibly lossy dielectric half-space is presented and evaluated. Different possibilities for considering the half-space influence are discussed, where an approach with an appropriate half-space Green's function is found to be most accurate, as expected. The formulation of the equivalent sources transformation approach with the half-space Green's function and a formulation with free-space Green's function together with equivalent sources representation of the half-space influence are discussed and a variety of results are presented in order to corroborate the feasibility of the various approaches.
Non-destructive testing (NDT) is a fundamental step in the production chain of aircraft structural components since it can save both money and time in product evaluation and troubleshooting. This paper presents a reflection-based imaging technique for electromagnetic (EM) testing of composite panels, with the device under test (DUT) being metal backed and both the transmitting and receiving components of the NDT system situated on the same side of the DUT. One of the key properties of the presented technique is the complete redundancy of a reference measurement, thereby making it feasible to retrieve a high quality image of the DUT with only a single measurement. Data for both a proof-of-concept DUT and an industrially manufactured composite panel is provided, and the retrieved images show the applicability of both the measurement technique and the imaging algorithms.
The use of mode filtering to improve the quality of antenna measurements taken in non-anechoic environments is well known, [1, 2, 3, 4, 5]. In the far-field case [6, 7, 8], it has been shown that it is possible to use standard cylindrical near-field theory [8] to implement the necessary mode filtering using a singularly polarized, great circle, far-field pattern cut consisting of amplitude and phase data. The careful verification of this technique using a compact antenna test range (CATR) was reported in [7, 8] however that implementation had, as a prerequisite, the need to acquire the far-field data on a monotonic and equally spaced pattern abscissa. In many instances this is not convenient or perhaps impossible. This paper presents a recent development which allows data to be processed rigorously when having been acquired using an unequally spaced angular abscissa. This paper sets out the novel, far more sophisticated, algorithm together with results of actual range measurements that were processed using this new technique.