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 optimized approach to Near-Field sampling based on Singular Value Optimization allowing the definition of the optimal distribution of the number and measurement locations is here applied to the case of aperture antenna and the hemispherical scanning geometry. The numerical validation shows the performance of the approach and its robustness against the noise on the data. A numerical validation is presented.
Reliable in-flight Wi-Fi is an essential service in modern aviation, requiring efficient planning to ensure consistent coverage with minimal infrastructure. This paper presents a simulation-driven approach to optimize the placement of Wi-Fi access points (APs) inside an aircraft fuselage, aiming to achieve maximum signal coverage with a minimum number of APs while maintaining a signal strength threshold of -30 dBm. A detailed 3D model of a typical aircraft cabin is constructed, incorporating structural elements, seating layouts, and material-specific electromagnetic properties, including the presence of passengers. Signal propagation is simulated using Altair WinProp and Feko, which account for reflection, diffraction, and material attenuation. The optimization is carried out using Altair HyperStudy, leveraging the Global Response Surface Method (GRSM) to explore the large design space and converge on an effective solution. The results show that strategic placement of as few as three access points can provide over 75% cabin coverage, demonstrating the effectiveness of the combined simulation and optimization workflow. This methodology enables efficient Wi-Fi network planning in confined, complex environments such as aircraft cabins and can be extended to other transportation systems like trains and buses.
Anechoic ranges require constant temperature and humidity, proper lighting to be able to work inside the range and closed-circuit television (CCTV) cameras to monitor the system while the measurement is being done. In addition, anechoic chambers require fire detection and suppression. Traditionally these penetrations are minimized and placed in non-critical areas. But the true effect of them has not been fully investigated. In this paper, antenna measurements as simulated in an indoor far field range. The approach to model the measurement is like the one the author presented in [1] and [2]. Thus, a range antenna (or near- field probe) and an antenna under test (AUT) are placed in free space and the AUT is rotated at discrete angles as it was done in [1]. Then a second model includes CCTV cameras, HVAC vents, light fixtures and both air sampling tubes and fire suppression nozzles and placed around. The simulation with these disruptions is repeated at the given discrete angles. The model does not include the absorber on the range. The model assumes a perfect absorber and the results of the simulated antenna measurement are compared to an ideal case with no disruptions. The results, while being approximations, provide a worst-case error for those disruptions of the RF-absorber layout. The results can be used to estimate the potential uncertainty on the measurement caused by the different systems that must be part of the anechoic enclosure. The technique is applied here to indoor far field measurements, and for near-field systems. Results show that for your typical roll over azimuth positioner, the effects of the penetrations on the ceiling are very small with differences in the -35 to -40 dB levels.
The first mention of a Robot for near-field measurements of antennas appears is by Jeff Snow in [1]. This was a simple robotic arm to do planar measurements. About 7 years later, the use of off-the-shelf industrial robotic arms for doing antenna measurements is introduced [2]. Since then, industrial- robot-arm based antenna measurement systems have become increasingly popular due to their flexibility to measure over different surfaces allowing the system to do planar, spherical and cylindrical. The use of other methods to perform the transform, by numerically compute the currents on an arbitrary surface from the measured fields has helped in the growing popularity of robotic systems. This is related that the measurement surface does no longer have to be a canonical surface but can be any shape. However, the flexibility of the robots may be limited by the RF absorber coverage used in treating them. In this paper, the authors explore the potential scattering from the robotic arm in different positions and its effect on the probe illuminations. This is an area of research on the use of absorber that has not been explored until recently [3]. Numerical experiments are conducted to explore the effects of RF absorbers in the 300 MHz to 3 GHz range. Open ended waveguides (OEWG) as well as dual ridged horns (see Figure 1) are used as the probes. The results suggest that some areas of the arm need to be treated while others can be left bare. The analyses performed suggest that optimized treatment of robotic arms to maintain the flexibility of the technique while also reducing effects on the probe illuminations are possible.
This paper presents research validating the forward scattered wave in bistatic radar geometries. In contrast to monostatic radar, where transmitter and receiver are collocated, bistatic radar incorporates receivers positioned in different locations from the transmitter. In the case of forward scattering, the receiver is positioned in the far field behind the reflecting target in line of sight of the transmitter. The paper describes forward scattering physics and presents forward scatter radar cross section measurements conducted at the Naval Air Warfare Center Weapons Division (NAWCWD) Radar Reflectivity Laboratory (RRL), which are validated with computational electromagnetic predictions.
This paper presents the results of a recent study into the use of the generalized vector addition theorem for antenna position translation for the purposes of orthogonalizing and then extracting the deleterious effects of parasitically coupled scatterers in spherical antenna measurements directly from the computed spherical mode expansion without the need to transform to the asymptotic true far-field in order to implement the all-important antenna translation. The generalized vector addition theorem can be successfully utilized to perform antenna position translations in any direction, and for displacements that can be smaller or larger than the maximum radial extent of the antenna. The revised algorithm represents a notable development as it is rigorous and general incorporating both reactive and propagating components, thereby making the processing applicable to a wider range of problems than has previously been the case. Results are presented that validate and illustrate the effectiveness of the new algorithm.
This paper proposes a novel approach based on a Multi-Probe Near-Field Scanner system utilizing bipolar and phased array antenna technology to measure antenna arrays with highly reliable precision. The innovative system incorporates a phased array antenna with full amplitude and phase control, enabling complete manipulation of the polarization state. This advanced capability allows the system to accurately measure and characterize antenna arrays while addressing and correcting polarization distortions caused by polar motion. By using the multi-probe configuration, the system significantly enhances measurement efficiency by capturing near-field data simultaneously across multiple probes. The integration of bipolar technology [1] ensures robust signal processing, while the phased array design enables the electronic rotation of the polarization of the probe array antenna. This feature is critical for mitigating errors introduced by polarization mismatches or distortions, ensuring high-fidelity measurements even in complex scenarios. The proposed system demonstrates superiority over traditional single-probe scanners by reducing measurement time, providing comprehensive polarization control and size reduction of the controlled environment. The paper discusses the system’s design, implementation, and performance.
The achievable signal-to-noise ratio (SNR) at the output of an antenna measurement system receiver is typically a parameter of interest. It is common practice to decrease the intermediate frequency bandwidth (IFBW) of the receiver to improve the SNR at the output of the receiver. This paper demonstrates that the achieved SNR improvement via IFBW reduction does not always match the expected SNR improvement when pink noise (1/f noise) is included in a model of an antenna measurement system. We present a simplified antenna measurement system model containing white and pink noise. We present equations for computing time average power, SNR at the input to the receiver, and SNR at the output of the receiver. We then present numerical results for achieved SNR when ignoring and when including pink noise. Finally, we present a few recommendations for reducing the 1/f noise power in the transmit chain of an antenna measurement system.
A new method is developed for accurate transmission measurements through a surface using a far-field Gaussian weighting approach for use in anechoic chambers. A trivial approach to measuring transmission characteristics would be to mount a sample-under-test between two antennas and simply measure the boresight transmission using a network analyzer. This approach is enhanced by instead measuring the transmitted fields over a hemispherical field-of-view and then weighting the measured far-fields to synthesize a Gaussian illumination of the sample. This approach can be utilized to effectively illuminate the sample-under-test with an arbitrary excitation with a high degree of customization in post-processing (e.g., beam polarization, direction, waist size, and location). The approach is validated by characterizing transmission through a copper clad substrate and a frequency selective surface (FSS) from 4 to 40 GHz. Additionally, a sample of Rogers 5870 is measured from 8 to 40 GHz. For each sample-under-test, S21 measurements are compared for the boresight and Gaussian weighted methods, showing greater agreement with theoretical or simulated values for the Gaussian case.
This paper presents a novel approach to radome measurement. Traditional radome measurements in anechoic chambers provide information on radome losses and beam deflection but cannot fully capture the RF environment the system will be operating in once deployed, which includes scattering sources and diffraction from the radome. It is also very difficult to characterize radome effects using similar approaches once the radome has been installed. This research investigates the use of a USRP-based field probing system to characterize the RF field of view for a radome located at the OneRY Range at Wright-Patterson Air Force Base (WPAFB) in Dayton, Ohio. Using a custom-built planar positioner along with two channels of the USRP, relative amplitude and phase measurements are taken inside the radome and processed into angle and time of arrival data, which are presented as angle vs distance plots. These show the direction and distance of scattering sources in the field of view of the aperture. For this radome, two scattering sources are identified along the path from the installation to a secondary transmit site. Diffraction from the radome wall is also detected and characterized. This method represents a new way of visualizing scattering and diffraction in radome-based installations. This method also provides a way to characterize existing radome installations that have not been measured in anechoic chambers, or in situations where the RF hardware has been changed to cover frequencies not previously characterized. The processing algorithm is presented along with plots comparing indoor and radome collects. Results from a 2D electromagnetics simulation are also presented to further validate diffraction measurements.
In recent times, we have become familiar with the use of commercial software for designing our antennas and microwave devices. This is very common since it is easy to find high-performance desktop computers at affordable prices in our daily lives. The use of general-purpose commercial software is widespread because it allows for the simulation of any arbitrary configuration. However, many of us have experienced, given the ease of using commercial software, trying to simulate electrically large electromagnetic devices which take days or, in some cases, cannot be completed at all. While it is true that we now have very powerful simulation tools, by making a few simple assumptions, we can significantly reduce computational time without sacrificing accuracy. In this talk, I will introduce a simple ray-tracing technique that can be used, in combination with physical optics, to calculate the radiation pattern of antennas, as well as directivity, gain, mutual coupling, and even early-time response in complex configurations. The results are not only faster than those produced by conventional commercial software, but also more accurate, as they avoid many of the numerical errors that typically arise when computing electrically large structures.
The use of open-ended coaxial probes for dielectric characterization of breast tissue requires precise coupling between the probe and the skin surface, avoiding air gaps and excessive pressures. Since dielectric properties are very sensitive to these conditions, improper adaptation can compromise the precision and repeatability of the measurements, limiting their effectiveness in clinical applications such as tissue grading or pathology monitoring. This article evaluates the geometric performance of four open-ended coaxial probe tip configurations (flat, conical, beveled, and semispherical) with the objective of evaluating their fit to biological tissues. A custom-built measurement system operating between 1 and 8 GHz was employed to computationally simulate and analyze the S11 response for each probe geometry, assessing both precision and effective penetration depth. The results show that the conical and beveled probes provide shallow penetration depths (<2 mm), making them suitable for in vivo or ex vivo procedures, such as biopsies. The semispherical probe achieves deeper penetration (approximately 2 mm deeper than the flat probe), which is beneficial in surgical settings or before more complex imaging techniques, such as mammography. In terms of precision, the semispherical and conical probes demonstrated the best performance for breast tissue and tumor characterization, with an effectiveness rate exceeding 80%. These findings allow for the assignment of complementary functions to each probe type, improving their integration into clinical settings for rapid and reliable tissue diagnosis and assessment.
Monolithic radomes for aircraft nosecones are designed for maximum transmission, minimum sidelobe increase, and minimum beam deflection at the desired radar frequency. In some cases, a radome may have a position dependent thickness to account for radar incidence angle. These radomes are manually tuned with an iterative process that uses radio frequency test range measurements to guide placement of tuning tape on to the radome. Generally, the radomes are thinner than desired, and the tuning adds an appropriate number of tape layers to center the optimum transmission window in frequency. This trial-and-error method is time consuming and range measurements are costly. Instead, this paper discusses an alternative method that is faster and less resource intensive. A microwave spot probe is raster scanned around the radome surface, and an algorithm inverts the electrical thickness at each measurement location from the measured reflection amplitude and phase. This thickness map is compared to the desired thickness and the appropriate number of tuning tape layers are calculated. This paper demonstrates the spot-probe method with a canonical wedge-shaped radome, where one side of the radome was purposely too thin. In addition, a lens-based far- field range is used to measure transmission efficiency and apparent beam deflection error of the wedge radome before and after it is tuned with fiberglass tape.
The approach of non-redundant near-field sampling has been available for many years. A general and automated approach that yields the expected time reduction for an arbitrary antenna volume, however, has been elusive. One of the more practical approaches is the “PNF wide-mesh” sampling, where the probe grid is separable in x and y, and this approach is the one explored in this paper. A fundamental step in non-redundant sampling is to identify a volume that fully contains the AUT. Constraints imposed by theory have typically led this volume to be rotationally symmetric about a z-oriented line, and often also require that the volume be more spherical (less oblate) than a volume circumscribing the AUT. That larger volume generally results in more acquisition time than would a conformal volume, but allows those samples to be readily interpolated to the conventional half-wavelength PNF grid. This paper examines the impacts of relaxing those constraints in order to further reduce the required sampling time for a box-shaped AUT. It then looks for ways to reduce or remove those impacts. The implementation of this algorithm involved a minor reformulation, specific to the PNF (or linear-axis) geometry, of the underlying non-redundant sampling theory. That reformulation is briefly described herein. A new family of tunable AUT-volume edge treatments similar to the existing “double-bowl” is also described. The paper will show minor reductions in predicted acquisition time compared to non-redundant sampling with a circular double-bowl volume. Each non-redundant approach typically offers a 40-60% reduction with a rectangular AUT volume compared to a full conventional scan. A more notable advantage of the new approach is a significant reduction in pre-acquisition activity defining the several parameters that govern the non-redundant acquisition and processing.
A new sparse sampling and compressive sensing based reconstruction and near-field imaging technique is introduced for the measurement of electrically large production test and diagnostics of nose-mounted commercial radomes. Simulated results are presented, where it is demonstrated that far- field results with an equivalent multi-path level of better than - 60dB can be obtained from circa 10% of the points required by a classical Nyquist equiangular spherical near-field acquisition scheme for the case of an electrically large, i.e. full size, commercial airliner nose-mounted radome enclosing an x-band weather radar. Furthermore, a new method for the rapid noninvasive nondestructive imaging and identification of defects within these radomes is presented that provides significantly clearer fault detection at a far earlier stage within the radome measurement campaign than has previously been possible.
We present an overview of a recently developed technique for performing antenna gain measurements with gain extrapolation that uses significantly fewer data points and at shorter distances than traditional gain extrapolation. This enhanced technique purposely incorporates third-order mutual coupling between antennas, which can be thought of as a useful homodyne signal, rather than an unwanted degradation of the antenna-to-antenna coupling signal as has been the historically accepted viewpoint. From Wacker’s fundamental extrapolation equations, we give the development of the third-order signal, which underpins this technique. From the third-order signal, the framing of gain extrapolation can be approached as a measure of interference fringes, as opposed to a by-rote curve fitting problem, and thus provides ways of specifying the number of required data points and measurement distances so as to reduce both significantly from the traditional gain extrapolation approach. The truncation order of the full signal expansion, as it relates to the conditioning of the problem, is presented in light of the behavior of the design matrix that defines the gain extrapolation scenario and the orders of scattering, thus leading to fewer required samples. Along with considerations of the matrix conditioning, guidelines are presented from the third-order signal and interference fringes for sampling criteria and sampling accuracy criteria. These guidelines aid in choices of measurement system accuracy and precision requirements based on known values of the operating frequency, wavelength, and antenna dimensions. Bounds for gain uncertainty based on these sampling criteria are also given. Results comparing NIST reference antenna measurements made with the traditional gain extrapolation and enhanced gain extrapolation technique are presented. It is shown that the enhanced technique can produce gain values in agreement to within uncertainties of the traditional technique for the reference antennas.
The wall-reflectivity technique is a proven method to validate the reflectivity reduction of large metallic walls covered with electromagnetic absorbing materials. While for a theoretical perspective it is sufficient to perform only two quasi-monostatic measurements, one for a known reference target and one of the wall, it has been shown that system dynamic range as well as clutter reduction can be achieved by performing spatial averaging using a linear slide, taking several measurements along the perpendicular direction against the wall. In this work the effect of improvements of the accuracy and positioning repeatability of the linear track is investigated. Measurement results are shown indicating the improved capabilities of the measurement setup.
This paper presents the design and simulation-based evaluation of a high-resolution millimeter-wave (mmWave) MIMO (Multiple-Input Multiple-Output) antenna array system for non-invasive medical diagnostics. The system is specifically optimized for applications such as early-stage tumor detection and soft tissue anomaly mapping, where high spatial resolution and tissue penetration are crucial. A 4×4 MIMO antenna array operating in the 28–40 GHz frequency band is proposed, leveraging the inherent advantages of mmWave frequencies— namely, shorter wavelengths for finer imaging resolution and wide bandwidth for enhanced contrast.The MIMO antenna array is designed using Rogers RT5880 substrate with a dielectric constant of 2.2 and a thickness of 0.787 mm to ensure minimal dielectric loss and mechanical stability. High-fidelity electromagnetic simulations were conducted using ANSYS HFSS to validate the antenna design. The resulting 3D radiation patterns confirm the beam directivity and uniform power distribution across all elements. The array was then integrated into a synthetic aperture radar (SAR)-based imaging model in MATLAB, where point- spread function (PSF) analysis revealed a lateral resolution of 3.2 mm and an axial resolution of 2.5 mm at 35 GHz. Imaging simulations on a multilayer human tissue-equivalent phantom model—comprising skin, fat, and muscle layers—demonstrated the system’s ability to resolve dielectric contrasts simulating benign and malignant tissue anomalies. The proposed MIMO antenna array enables real-time, contactless, and radiation-free imaging, positioning it as a cost-effective alternative to traditional imaging modalities such as X-ray or MRI for preliminary screening and continuous monitoring. The fully simulated results validate the concept’s feasibility and effectiveness for non-invasive medical diagnostics, particularly in point-of-care settings.
One of the most important key performance indicators of Reconfigurable Intelligent Surfaces (RISs) represents the RIS reflection or bistatic RIS Radar Cross Section (RCS) pattern, which needs to be evaluated under Far-Field (FF) conditions. Since a bistatic setup that fulfills FF conditions is mechanically complex, expensive and results in a large setup footprint, the measurement of RIS reflection patterns within a monostatic setup is proposed. Subsequently, the monostatic pattern results are transformed into bistatic patterns using a Monostatic to Bistatic Equivalence Theorem (MBET). This approach reduces the total measurement time tremendously. However, the state- of-the art MBET proposed for RIS measurements, is limited to 1-D pattern application, where the illumination-/transmit- and the probing-/measurement- antennas are located in the same plane. As RIS can also be configured to reflect off-the-plane, the evaluation of 2-D scenarios is crucial. With the state-of-the-art 1-D MBET such scenarios cannot be investigated. To close this gap, this paper derives a novel 2-D MBET well suited for RIS measurements. The novel MBET is validated with an analytical metal plate reflection model, by comparing the resulting 2-D MBET transformed monostatic pattern with the bistatic reference pattern. This procedure is repeated for measurement and simulation data of two different RIS prototypes designed for the mmWave frequency range. The deviation of the resulting patterns after applying the novel derived MBET from the bistatic reference patterns is analyzed based on the pattern difference metric. This difference metric evaluated in the main cuts exhibits a worst case mean value of –24dB which proves the suitability of this novel MBET for 2-D RIS measurements.
In this work, a novel two-steps technique to correct 3D probe positioning errors affecting the non-redundant (NR) spherical near-field/far-field transformation (NF/FFT) technique for quasi-planar antennas, modeled with a double bowl, is presented. The developed approach benefits from the synergy of two complementary correction steps. In the first one, the so-called NR correction compensates for the phase errors due to the deviations of the sampling points from the nominal scanning sphere. This is achieved through an ad hoc phase correction scheme, whereby an optimal phase factor is extracted from the acquired samples, accounting for their actual (i.e., erroneous) position. In the second step, an iterative approach is exploited to manage the residual 2-D errors affecting the data attained at the end of the previous step, allowing for an accurate and efficient restoration of the correctly located samples. Ultimately, the large number of input data for the classical spherical NF/FFT, uniformly spaced on the scanning spherical grid, are obtained via an optimal sampling interpolation formula. Numerical results showing the effectiveness of the proposed method in compensating even severe 3-D positioning errors are reported.