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.)
Plane wave generator (PWG) for Over The Air (OTA) characterization of beamforming millimeter wave devices, provides an attractive solution comparing to conventional measurement techniques (Compact Antenna test Ranges (CATR) and Far-field chambers). MVG’s Plane wave generator for 5G NR FR2 applications ([1]-[4]) is an innovative tool which permits the user to measure the radiating elements with low to medium directivity radiation characteristics with excellent precision. Conventional CATR systems are not suited for stationary DUT (with / without person) measurement scenario. In this paper, experimental results are presented for a dual-polarized PWG system, covering the 3GPP bands n257, n258 and n261 (24.25-29.5 GHz). System measurement results show good comparison with simulations and measurements of the PWG alone. Another advantage of PWG presented here, is that we can modify the size of the QZ. Results from a pre-production unit for a 15cm QZ shows amplitude variation of less than ±1 dB and achieve more precision for smaller DUT. Measurement results from the pre-production unit with a quiet zone of up to 38cm sphere diameter, show amplitude variations of less than ±2dB. This variation is compatible with the DUT + phantom or human measurement application. Pattern results for Antenna Under Test (AUT) with low to medium directivity (6dBi up to 17dBi) compare well with simulations and measurements from other systems. For a given AUT, the impact of different positioning mast is also evaluated. Excellent stability of patterns, when the AUT is placed at different positions inside the QZ, is observed. These results confirm that the dual-polarized PWG system presents an attractive solution for FR2 characterization of low to medium directivity radiating elements.
The characterization of antennas is a time-consuming task. Its acceleration leads often to large and sensitive numerical problems. Therefore, special care must be taken of the choice of the parameters, the optimization, and the stability of the employed resolution methods. Based on Huygens’ principle, the radiation operator can be defined from an equivalent surface enclosing the Antenna Under Test (AUT). The discretization of this operator leads to the so-called radiation matrix. An expansion basis of the fields radiated from the equivalent to the measurement surface is constructed by the Singular Value Decomposition (SVD) of that matrix. The Reduced-Order Model (ROM) is the compressed representation of this basis obtained by truncating the SVD. The truncation order, T, is computed by inspection of the singular value distribution and is strongly linked to the number of degrees of freedom of the radiated fields. Several practical and technical aspects are studied in this article to provide a systematic, efficient and reliable procedure for the characterization of the radiated fields using the ROM. Analytical criteria are used to define the dimensions of the radiation matrix enabling a stable determination of the compressed basis. The truncation order, T, is the key-point of this method as it determines the size of this basis. Therefore, its variation is studied with respect to the discretization step and the geometry of both equivalent and measurement surface. Finally, the Randomized SVD (RSVD) is used in order to significantly reduce the computation time with negligible impact on the accuracy. To illustrate our procedure, it is applied to various scenarios and experimental results of spherical measurements. Estimations of the time savings by using the RSVD are also provided.
The novel extension of the synthetic aperture radar (SAR) technique to the terahertz (THz) spectrum has emerging short-range applications, especially in an indoor environment. One of the key applications is the generation of a high-resolution indoor environment map in emergency scenarios such as a burning or smoky building, where optical technology might not provide any relevant information. The THz SAR map enables precise localization, classification, and material characterization of concerning objects which can assist in identifying the danger from electrical cables located in the walls and ceilings, and the structural integrity and failure of the walls/ceilings. Hence, the investigation of through-wall sensing at the THz spectrum is of vital importance. This paper addresses the through-wall sensing at the THz spectrum by employing the SAR technique. A miniature version of the wall using gypsum plasterboards is constructed, where the plasterboards are mounted on a frame. Two types of frames are considered, where one frame is of wood and the other is of metal. Additionally, electrical cables are placed between the plasterboards. This miniature version is quite similar to a practical environment. Besides, some of the considered components of the wall are in a burned state. For through-wall sensing, a vector network analyzer (VNA) based testbed is implemented and measurements are recorded in both transmission and reflection modes for three frequency spectrums, which are 75-100 GHz, 220-330 GHz, and 325-500 GHz. At the THz spectrum, the penetration capabilities are always of concern. Therefore, foremost, penetration losses among different components of the wall are investigated with transmission measurements. Further, to evaluate the sensing capabilities behind the wall, transmission measurements are recorded by considering the whole structure of the wall. Besides, relative attenuation among different frequency spectrums is presented. The addressed evaluation is also of significant interest in the area of wireless communication such as 6G and security. Lastly, in reflection mode, a 2D SAR trajectory is implemented and a 3D image of the wall is reconstructed. It is analyzed for identification and precise localization of the cables and frame-blocks. The identified components are further processed for burned state detection.
Spherical Near-Field (SNF) antenna measurements are one of the most accurate antenna characterization methods. However, despite the high accuracy, they have several drawbacks. One of them is the requirement/challenge to acquire not only the amplitude but also the phase of the near-field measurements in order to reconstruct the far-field of the antenna. However, in comparison to amplitude, phase measurements require expensive equipment and are more prone to inaccuracies, particularly at higher frequencies. Moreover, in specific scenarios such as over the air measurements, a reference phase is unavailable or inaccessible. To overcome this, phaseless approaches are of interest and different methods such as using different probes or the two-spheres technique are investigated. The latter has gained more interest in recent research by introducing phase retrieval algorithms such as Wirtinger-Flow, PhaseLift or Gerchberg-Saxton to SNF. We consider an SNF setup with a roll-over-azimuth positioner for the Antenna Under Test (AUT) and a probe mounting with 90° rotation in azimuth. Measuring two spheres with different radii necessitates translating the measurement coordinate system on the AUT-probe axis. This can be done either via an offset of the probe mounting or via a linear axis, whereby the latter is the more flexible solution in terms of measurement automation. In our approach, the considered measurement setup is part of a hybrid test range, consisting of an SNF setup and a Compact Antenna Test Range (CATR). The AUT positioner, part of both CATR and SNF, is mounted on an additional linear axis. Since the linear axis has been designed to measure at different positions in the quiet zone of the CATR, it is not aligned with the near-field probe. Thus, when measuring two spheres, the probe misalignment must be compensated in post-processing by means of an additional rotation of the probe’s spherical mode coefficients. This method allows using arbitrary oriented linear axes to vary the measurement radius and hence increases the flexibility in choosing different radii for the two-sphere technique, which is a critical parameter in the phaseless reconstruction of the far-field.
A custom radar kit that integrates with a portable computer (laptop) for assembly and operation by students and researchers has been developed at MIT Lincoln Laboratory. The assembled radar kit uses two low-cost cylindrical metal cans that serve as the antennas, one for transmitting and one for receiving radar signals. The antennas operate as linearly polarized openended circular waveguides (10.5 cm diameter) fed with a thinwire monopole probe. Over the 2.4 to 2.5 GHz band, the measured reflection coefficient is less than −10 dB, the peak realized gain is greater than 7 dBi, and the half-power beamwidth is approximately 70 degrees in both the E- and Hplanes. FEKO method of moments simulations of the antenna are compared with the measured data and good agreement is demonstrated.
Radar sensors are an essential component in the automotive sector and take over safety-relevant functions in the field of autonomous driving. Therefore, the need for validation of automotive radar systems is increasing. Within this paper, a measurement setup for automated static and dynamic tests of integrated radar sensors is set up in the robot-based measurement chamber available at the Institute of High Frequency Technology, RWTH Aachen University. The system parameters two-way pattern, range and speed resolution as well as angular resolution and separation capability are measured and analyzed for an integrated automotive radar sensor. The measured results show the expected performance of the radar system and point out the high variability of the built setup.
High precision antenna measurements can be carried out in different ways, for example directly in the far-field, in a compact antenna test range (CATR) or in a near-field range. Far-field measurements have the disadvantage that they require a lot of space and are sensitive to environmental influences. Although these effects can be avoided with a CATR, these systems are demanding and expensive. For this reason, simplifying near-field measurement setups are often preferred, whereby three main methods have become established: planar near-field (PNF), cylindrical near-field (CNF) and especially spherical near-field (SNF) measurements. Current measurement setups are usually optimized for one of these techniques and are therefore only of limited use for the other methods. Recent developments have focused on robotic measurement systems, which are able to overcome the boundaries between the different measurement techniques. Due to the high number of degrees of freedom and an almost unlimited positioning and orientation possibility in space, these systems offer numerous advantages due to their flexibility. This allows the same measurement system to be used seamlessly to perform PNF, CNF and SNF measurements. Even for more demanding measurement procedures, for example based on compressed sensing, robotic systems create the possibility to efficiently approach the required sampling points. At the Institute of High Frequency Technology at RWTH Aachen University, such a robot-based measurement setup is currently being established. In addition to the six-axis robot arm, the system has two additional axes, specifically a linear axis on whose slide a rotary axis is mounted. In the currently used configuration, the antenna under test is mounted on the axis of rotation, while the probe antenna is installed at the robotic arm. This results in a total of eight degrees of freedom combined in a novel test range design, which distinguishes the measurement setup from existing ones. Further details on the measurement setup, the implementation of antenna and radar measurements and the operation of the entire system will be explained.
The standard definitions of terms for antennas and radio wave propagation are provided in IEEE Std. 145-2013 and IEEE Std. 211-2018, respectively. These documents define the accepted and harmonized terminology used in the fields of antennas and propagation. Arbitrary use and misuse of fundamental terms are still observed in technical papers and presentations at conferences.
Phase uncertainty in antenna measurements introduces significant errors to the amplitude of the transformed pattern in Spherical Wave Expansion (SWE). To get a better understanding of the impact of phase errors, the measured phase error of a Low Noise Amplifier (LNA) is synthesized as a random phase error and subsequently added to the measured antenna patterns of three different antennas during the SWE. The resulting erroneous patterns are compared with the measured reference patterns and the error magnitude and probability distribution are studied. It is proven that the introduced errors to the transformed far-field patterns can be substantial. Furthermore, the relation between the antenna type and the error level and distribution is elaborated. The error level is different for the three antennas and the error level distribution is dependent on the mode spectra of the antennas.
A near-field far-field transformation (NFFFT) technique with a plane-wave synthesis is presented for predicting two-dimensional (2D) radar cross sections (RCS) from multistatic near-field (NF) measurements. The NFFFT predicts the FF of the OUT illuminated by each single source, then the plane-wave synthesis predicts the FF of the OUT each illuminated by each plane-wave by synthesizing the FFs given in the NFFFT step. The both steps are performed in the similar computational procedure based on a single-cut NFFFT technique that has been proposed previously. The method is performed at low cost computation because the NF and source positions are required only on a single cut plane. The formulation and validation of the method is presented.
The significant measurement standards in the antenna measurement community all present suggested error analysis strategies and recommendations. However, many of the factors in these analyses are static in nature in that they do not vary with antenna pattern signal level or they deal with specific points in the pattern, such as realized gain, side lobe magnitude error or a derived metric such as on-axis cross polarization. In addition, many of the constituent factors of the error methods are the result of analyses or special purpose data collections that may not be available for periodic measurement. The objective of this paper is to use only a few significant factors to analyze the error bounds in both magnitude and phase for a given antenna pattern, for all levels of the pattern. Most of the standards metrics are errors of amplitude. However, interest is increasing in determining phase errors and, hence, this methodology includes phase error analysis for all factors.
A simulation-supported measurement campaign was conducted to collect monostatic radar cross section (RCS) data as part of a larger effort to establish the Austin RCS Benchmark Suite, a publicly available benchmark suite for quantifying the performance of RCS simulations. In order to demonstrate the impact of materials on RCS simulation and measurement, various mixed-material targets were built and measured. The results are reported for three targets: (i) Solid Resin Almond: an almond-shaped low-loss homogeneous target with the characteristic length of ~10-in. (ii) Open Tail-Coated Almond: the surface of the solid resin almond's tail portion was coated with a highly conductive silver, effectively forming a resin-filled open cavity with metallic walls. (iii) Closed Tail-Coated Almond: the resin almond was manufactured in two pieces, the tail piece was coated completely with silver coating (creating a closed metallic surface), and the two pieces were joined. The measured material properties of the resin are reported; the RCS measurement setup, data collection, and post processing are detailed; and the uncertainty in measured data is quantified with the help of simulations.
In 1987 the author built the world's first Personal Near-field antenna measurement System (PNS). This led to the formation of Nearfield Systems Inc. (NSI) a company that became a major manufacturer of commercial near-field antenna measurement systems. After leaving NSI in 2015 several new personal antenna measurement tools were built including a modern updated PNS. The new PNS consists of a portable XY scanner, a hand held microwave analyzer and a laptop computer running custom software. The PNS was then further generalized into a modular electromagnetic field imaging tool called "Radio Camera". The Radio Camera measures electromagnetic fields as a n-dimensional function of swept independent parameters. The multidimensional data sets are processed with geometric and spectral transformations and then visualized. This paper provides an overview of the new PNS and Radio Camera, discusses operational considerations, and compares it with the technology of the original 1987 PNS. Today it is practical for companies, schools and individuals to build low-cost personal antenna measurement systems that are fully capable of meeting modern industry measurement standards. These systems can be further enhanced to explore and visualize electromagnetic fields in new and interesting ways.
The experimental validation of an accurate and fast near-field-far-field (NF-FF) transformation technique with spherical scan, suitable for long antennas under test (AUTs) mounted in offset configuration, is provided in this work. The main feature of such a NF-FF transformation is to require, unlike the traditional spherical (TS) one, an amount of NF samples, which is minimum and results to be practically the same in both cases of offset and onset mount-ings of the AUT. To this end, an optimal sampling interpolation formula , developed by properly exploiting the non-redundant sampling representations and modeling an offset mounted long AUT by a cylinder ended by two half-spheres, is employed to precisely recover the massive input NF data for the TS NF-FF transformation from the collected non-redundant samples. A considerable measurement time-saving can be so achieved. Experimental results assessing the validity and the practical feasibility of such a technique are shown.
An approximation method is developed to remove the source antenna's cross-polarization discrimination (XPD) contribution from the total measured XPD. This modeling is shown to correlate very well on a flat-panel test with a radome's predicted (ideal-source) XPD. Additionally, a mathematical formulation of the theoretical cross-polarization discrimination (XPD) bounds is presented to validate the proposed method. The measured axial ratio should not exceed these bounds. The measured result is within these bounds and thus this model serves as an additional validation step to both the proposed method and the measured results.
Indoor RCS measurement facilities are usually dedicated to the characterization of only one azimuth cut and one elevation cut of the full spherical RCS target pattern. In order to perform more complete characterizations, a spherical experimental layout has been developed at CEA for indoor Near Field monostatic RCS assessment [3]. This experimental layout is composed of a 4 meters radius motorized rotating arch (horizontal axis) holding the measurement antennas while the target is located on a polystyrene mast mounted on a rotating positioning system (vertical axis). The combination of the two rotation capabilities allows full 3D near field monostatic RCS characterization. 3D imaging is a suitable tool to accurately locate and characterize in 3D the main contributors to the RCS. However, this is a non-invertible Fourier synthesis problem because the number of unknowns is larger than the number of data. Conventional methods such as the Polar Format Algorithm (PFA), which consists of data reformatting including zero-padding followed by an inverse fast Fourier transform, provide results of limited quality. We propose a new high resolution method, named SPRITE (for SParse Radar Imaging TEchnique), which considerably increases the quality of the estimated RCS maps. This specific 3D radar imaging method was developed and applied to the fast 3D spherical near field scans. In this paper, this algorithm is tested on measured data from a metallic target, called Mx-14. It is a fully metallic shape of a 2m long missile-like target. This object, composed of several elements is completely versatile, allowing any change in its size, the presence or not of the front and / or rear fins, and the presence or not of mechanical defects, … Results are analyzed and compared in order to study the 3D radar imaging technique performances.
The major disadvantage of Spherical Near-Field (SNF) measurements is their long acquisition time. To calculate the Antenna Under Test's (AUT) far-field radiation characteristics , a sphere containing the AUT must be sampled. Classically, equiangular sampling is chosen, being the resulting sphere heavily oversampled. Since the Spherical Mode Coefficients (SMCs) are usually sparse, an approach to reduce the measurement time of SNF measurements is to undersample the sphere and to reconstruct the SMCs using compressed-sensing techniques. Using a sampling matrix with a minimum mutual coherence for the given bases of the SMCs increases the probability of recovery. The SMCs are defined in the basis of the spherical harmonics or Wigner D-functions, which limits the geometries in which this technique can be applied. In this work, the application of pointwise probe correction for the description of non-spherical surfaces in the Wigner-D basis expansion is suggested. The chosen sampling points are radially projected onto the measurement surface and the new distance to each point is calculated. New equivalent probe response coefficients are calculated per measurement point according to their distance to the AUT. To compensate for different orientations other than the probe pointing to the AUT's minimum sphere's center, the probe's SMCs are rotated to reflect the real orientation of the probe at each point prior to the calculation of the probe response coefficients. Although more computationally demanding than classical probe correction, this technique allows measurements with different, potentially faster geometries and enables the application of compressed sensing to other, non-spherical conventional scanning systems.
As the wireless industry continues the move to 5G, the development and subsequent testing of mmWave radios for both base stations and user equipment still face numerous hurdles. The need to test most conformance and performance metrics through the antenna array at these frequencies poses significant challenges and has resulted in excessively large measurement uncertainty estimates to the point where the resulting metrics themselves may be useless. A large contribution to this measurement uncertainty is the impact of the over-the-air (OTA) test range used, driving the industry towards expensive compact range reflector systems in order to overcome the path loss considerations associated with direct far-field measurements. However, this approach necessitates the use of a combined axis measurement system, which implies the need for considerable support structure to hold the device under test and manipulate it in two orthogonal axes. This paper explores some of the limitations and considerations involved in the use of traditional "RF transparent" support materials for mmWave device testing.
Three measurement procedures and associated post-processing for the fast characterization of antennas are presented. First, an approach for the fast diagnosis of antenna under test (AUT), ie. the identification of potential defaults with respect to an ideal antenna, is described. The technique leverages the knowledge of the ideal (expected) radiation pattern and uses a sparse recovery algorithm to locate the few potential defaults. Second, a scheme is proposed to interpolate the near field radiated by the AUT. It exploits the low complexity of the electromagnetic field and does not resort to any knowledge on the AUT. Third, an approach to speed up the measurement of the AUT far field radiation pattern is detailed. The only input is the maximum dimension of the AUT. The technique relies on the sparse expansion of antenna radiation patterns on spherical harmonic basis. For each of the three examples, experimental results will be shown for various complex radiating structures in different frequency bands.
An extended long object usually gives rise to a bright reflection (a glint) when viewed near its surface normal. To take advantage of this phenomenon and as a new concept, a discrete Fourier transform (DFT) on the RCS measurements, taken within a small angular range of broadside, would yield a spectrum of incident wave distribution along that object; provided that the scattering is uniform per unit length, such as from a long cylinder [1, 2]. In this report, we examine the DFT spectra obtained from three horizontal long objects of different lengths (each of 60, 20, and 8 feet). Aside from the end effects, the DFT spectra looked similar and promising as an alternative to the conventional field probes by translating a sphere across a horizontal path. Keywords: RCS measurements, compact range, field probes, extended long objects 1. The Boeing 9-77 compact range The Boeing 9-77 indoor compact range was constructed in 1988 based on the largest Harris model 1640. Figure 1 is a schematic view of the chamber, which is of the Cassigranian configuration with dual-reflectors. The relative position of the main reflector and the upper turntable (UTT) are as shown. The inside dimensions of the chamber are 216-ft long, by 80-ft high, and 110-ft wide. For convenience, we define a set of Cartesian coordinates (x: pointing out of the paper, y: pointing up, z: pointing down-range), with the origin at the center of the quiet zone (QZ). The QZ was designed as an ellipsoidal volume of length 50-ft along z, height 28-ft along y, and width 40-ft along x. The back wall is located at z = 75 ft, whereas the center of the roll-edged main reflector (tilted at 25 o from vertical) is at z =-110 ft. It is estimated that the design approach controls the energy by focusing 98% of it inside the QZ for target measurements. The residual field spreading out from the main reflector was attenuated by various absorbers arranged in arrays and covering the chamber walls.-, Tel. (425) 392-0175 2. Anechoic chamber In order to provide a quiet environment for RCS measurements, the inside surfaces of an anechoic chamber are typically shielded by various pyramidal and wedged-shaped absorbers, which afford good attenuation at near-normal incidence for frequencies higher than ~2 GHz. At low frequencies and oblique angles [3], however, Figure 1. A schematic view of the Boeing 9-77 compact range with dimensions as noted. insufficient attenuation of the radar waves by the absorbers may give rise to appreciable backgrounds. Figure 2 shows a panorama view inside the compact range, as viewed from the lower rear toward the main reflector and the UTT. With the exception of the UTT, all other absorbers are non-moving or stationary. A ring of lights on the floor shows the rim around the lower turntable (LTT), prior to the installation of absorbers. In order to minimize the target-wall interactions, the surfaces facing the QZ from the ceiling, floor, and two sidewalls are covered with the Rantec EHP-26 type of special pyramidal absorbers.