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 18-term NIST error model is a common tool for analyzing potential sources of error in antenna measurements. One of the error terms to be considered describes the phase errors occurring in a measurement system. However, this quantity plays a rather negligible role for conventional ranges, such as roll-over-azimuth positioning systems. In particular, the contribution caused by flexing cables is normally insignificant. This results from the fixed installation of the cables or the decoupling of the movement at important points using rotary joints. Current developments in the field of antenna measurement technology focus, among other things, on performing measurements using industrial robot arms. These are characterized by their high flexibility regarding the various measurement sequences, such as planar, cylindrical or spherical measurements. However, it is to be expected that the high freedom of positioning possibilities will introduce additional phase uncertainties, since the RF cables in the cable carrier chain of the robot arm itself are often not decoupled. Instead, a single cable is used for each signal path, which follows the movements of the robot. The robot-based measurement system at the Institute of High Frequency Technology at RWTH Aachen University has been designed for frequencies above 60 GHz, where phase stability is a challenging task. Depending on the setup, it may even be required to pass Intermediate Frequency (IF) signals on the same cable as the Local Oscillator (LO) signals. This results in different test cases for the phase deviations depending on the frequency range of the IF (279 MHz) and LO (typically 10 GHz to 18 GHz) signals. Additional factors such as the measurement path of the robot or the position of the linear axis must also be taken into account. Therefore, a thorough analysis of the phase uncertainties caused due to flexing cables is of outstanding importance for robot-based measurement systems.
There have been a number of numerical analyses of RF absorber presented in the literature. These analyses, however, tend to focus on the reflectivity of the material and not on the radar cross section (RCS) that it presents. Brumley studied the RCS of RF absorbers for the purpose of estimating the background RCS of anechoic ranges [1]. The study was done empirically; obtaining measurements of the RF absorber and looking at the RCS of different pyramids and wedges, with and without paint. Brumley presents some potential methods to improving the RCS signature of the range, thus reducing the background RCS of the site. In this paper, the suggestions presented by Brumley are revisited. Specifically, his recommendation for the twisted pyramid configuration which he was unable to measure due to the lack of absorber samples available for use in the test. In addition to the twisted pyramid, Brumley's approach of inserting smaller pyramids in the valleys of a larger pyramidal arrangement to reduce the edges parallel to the incoming wave are also presented. Different carbon loadings are modeled for the inserted pyramids. One is the standard loading of the inserted pyramid, and the other is the same loading as the main larger pyramidal arrangement such that all the absorber on the wall has the same material properties. Numerical studies are performed using time domain techniques as well as frequency domain techniques. The model is validated by comparing the RCS of a flat square plate with the theoretical solution. The results validate the data and the suggestions presented in [1] and present ways of improving some of the solutions by adjusting the material properties of the absorber.
Financial impacts often drive decisions to repurpose existing ranges instead of procuring new measurement facilities. These existing ranges have fixed geometries (height, width and length) that were set when the range was originally constructed and often are designed for a different purpose. The inability to set the geometry precludes the range designer from using the range geometry to improve measurement performance. Thus, the performance of the range is mostly dependent on the RF absorber and the range antenna directivity. In rectangular-shaped ranges for example, the lateral surfaces, side walls, ceiling and floor, are the critical surfaces to address in RF absorber arrangement. In this paper, numerical analyses of Chebyshev arrangements as well as dragon tail or tilted absorber are studied. This paper also analyzes the performance of Chebyshev absorber for normal incidence and for oblique incidence along with the proper arrangement of the Chebyshev period. While certainly these have been discussed previously in the literature, this paper consolidates the previous information and illustrates it with numerical examples to help the reader understand the best approach to use when repurposing a range.
One of the main disadvantages of Spherical Near-Field (SNF) measurements is their acquisition time. This is due to the need of sampling a whole sphere around the Antenna Under Test (AUT) to perform the Near-Field-to-Far-Field Transformation (NFFFT). A step of the NFFFT is to decompose the measured signal in each one of the spherical waves it consists of, thus retrieving the Spherical Mode Coefficients (SMCs) associated to the AUT. Under typical measurement conditions, the SMCs of most physical AUTs prove sparse, i.e., most of their terms are zero or neglectable. Using this assumption, the system of linear equations with the SMCs as variables can be solved with fewer equations, that is, fewer measurement samples. This is done by applying an l1-minimization solver, following classical methodology from the field of compressed sensing. However, the location of the measurement points that generate non-redundant equations is not trivial. In typical compressed-sensing applications, a random sampling matrix is taken. Since a random matrix is inefficient for the acquisition with mechanical roll-over-azimuth positioner systems, a recent approach is to take an equidistant distribution of points on elevation and to calculate their corresponding pair on azimuth that delivers the minimum coherence of the sampling matrix. However, the number of sampling points M required for a successful reconstruction depends on the sparsity level of the SMCs of the unknown AUT, making its choice critical and based on a pessimistic approach. A method for the adaptive choice of M is suggested. After the acquisition of a starting set of M_0 measurement points, chosen using phase transition diagrams, the SMCs are estimated online with few iterations of an l1-minimization algorithm. Afterwards, further points are acquired, and the SMCs are estimated again using them. Following the evolution and the decrease of the variation between estimates, it is possible to truncate the measurement at a point where a successful reconstruction is guaranteed. The method for the construction of a minimum-coherent sampling matrix for adaptive acquisition and the truncation criteria for a specific accuracy are discussed with a focus onimplementation, and supported with numerical experiments, performed with measurementdata.
The electrical alignment of the positioner of the Antenna Under Test (AUT) is an important issue to be accounted for before any antenna measurement can take place in a spherical near-field measurement facility. This is because the spherical transmission formula requires the AUT to be scanned on the surface of a sphere. Typically, the tower has been aligned by optical means but, usually, it is necessary to translate it along some axis to place the AUT in the center of the measurement sphere. Furthermore, mounting the AUT can alter the alignment due to its gravitational load. Therefore, alignment of the tower after mounting the AUT is a critical step, which is accomplished with the so-called flip tests [1]. These flip tests, which can detect only the axis intersection and zero-? errors of a roll over azimuth system, have been discussed in the past, for example [1-2], but not as extensively as their importance would require. Moreover, there were no analytical proofs provided for the error formulas given, which forces the interested researcher to derive them again in order to comprehend them and adapt them to his measurement facility. This paper starts with a concise and thorough presentation of how the flip-tests are performed in practice, as well as their theoretical justification. In the second part, the paper presents a novel idea regarding the interdependence of the alignment errors. It has been observed experimentally that two linear coupled equations can model their behavior. Consequently, they can be fully corrected simultaneously with only two flip-tests, without the need of correcting each one in small steps. Simulation tests were performed validating these results. Finally, the paper concludes with addressing a few miscellaneous issues that are inevitably risen by the nature of this procedure, such as the effect of the antenna gain, the positioning of the probe and the distance between the AUT and the probe.
Electrical material parameters such as permittivity and permeability are a prerequisite to analysis and to design of EM devices and systems.For the measurements of the EM material parameters, coaxial/waveguide methods and cavity method are used in low frequency range whereas free-space method is suitable in high frequency range. In free-space method, one of the non-destructive methods without prior machining of a MUT (Material Under Test), TRL (Thru-Reflect-Line) calibration method is used when the free-space measurement system has a linear slide to precisely adjust the separation distance between transmit (Tx) and receive (Rx) antennas, and GRL (Gated-Reflect-Line) calibration method in the case that the separation distance between the two antennas is fixed. As one of the well-known calibration methods, SOLR (Short-Open-Load-Reciprocal) calibration method assumes seven unknowns in the free-space material measurement configuration, i.e., the port-#1-side directivity, match, and tracking (E_DF, E_SF, E_RF) between the VNA test port #1 and the MUT, the port-#2-side directivity, match, and tracking (E_DR, E_SR, E_RR) between the VNA test port #2 and the MUT, and the quotient ?/…. This calibration, first performs two one-port calibrations at the port-#1-side and port-#2-side to determine (E_DF, E_SF, E_RF) and (E_DR, E_SR, E_RR) using three reflection standards, and then performs one transmission measurement using a reciprocal two-port standard (in this case, thru) for determining ?/…. Calibration of a free-space material measurement system by using SOLR method requires three free-space reflection standards. Recently, a planar offset short is proposed as a free-space reflection standard because its reflection property has the magnitude of unity and the phase proportional to the offset of the offset short. This paper proposes the SOLR method using three planar offset shorts with the respective offset of (0, ?/6, 2?/6) for calibrating a free-space measurement system. Effects due to the thickness of the MUT are compensated by a de-embedding process. The thinner the thickness is, the better results this method can get. The proposed method does not require a linear slide to precisely adjust the separation distance between Tx and Rx antennas of the measurement system. Theoretical details and measurement results will be presented at the symposium.
While the use of antennas in automobiles was earlier confined to AM and FM radio, most vehicles that we see today employ antennas additionally for satellite navigation, remote keyless entry etc. More number of antennas are likely to be needed in the future in vehicles for the purposes of internet and video browsing, collision avoidance radar systems, and for communication either between vehicles or between vehicles and infrastructure. Over the years, many planar monopole antennas that utilized patch slotting techniques have provided the desired multiband operation. But the choice of shape, size and position of slots in most of these works has been experimental and no generalized approach has been adopted. Fractal antennas, more particularly, those using the self-similar Sierpinski gasket geometries facilitate log periodic multiband behaviour. Although, the planar monopole configuration of this antenna geometry can be more useful candidate for most automotive applications, it has been found that the impedance mismatch at design frequency is a common problem in these antennas. We demonstrate a simple solution to this problem by considering an example of a third iterated gasket antenna of height 100 mm that has a geometric scale factor of 0.5. When the substrate of this antenna is loaded with a combination of SRR and strip wire on its rear side, a negative time delay gets added to the wave travelling inside the medium which results into the miniaturization of antenna. The CST MWS simulations clearly indicate this effect through the reduction of initial frequency resonances of 0.464 GHz, 1.568 GHz, 3.336 GHz and 6.2 GHz to 0.456 GHz, 1.312 GHz, 2.576 GHz and 5.032 GHz. The implementation of SRR-strip wire combination as a NFRP element in the antenna structure provided the impedance match at the reduced set of frequencies which is reflected in terms of improved return loss The loaded antenna structure provided gains of 2.44 dB, 4.74 dB, 5.7 dB and 6.2 dB respectively at the corresponding multiband frequencies. Based on such performance, the DNG-MTM loaded planar gasket monopole antennas of different heights can be expected to replace other planar monopole antennas for multiband automotive applications.
The planar wide-mesh scanning (PWMS) methodology is based on Non-redundant scanning schemes allowing faster measurements than classical Nyquist-compliant acquisitions based on denser, regular, equally spaced Near Field (NF) sampling. The methodology has no accuracy loss and has been validated at different bands and with different antennas [1]. The effectiveness of the PWMS technique has always been proven in error-free (or quasi-error-free) scenarios, assuming that possible errors introduced by the technique itself are independent of the typical source of measurement uncertainty. In this paper, we investigate for the first time the sensibility of the method wrt one of this error source included in the 18-terms lists [2], considered by the measurement community as an exhaustive list of the NF errors: X and Y probe positioning errors. Such errors are unknown and random and are associated to the mechanical vibrations and/or backlash of the system. The investigation has been done considering actual measurements of a multi-beam reflector antenna with approximately 35 dBi gain (MVG SR40 fed by two MVG SH5000 dual ridge horn). The AUT has been measured in planar geometry emulated by a 6-axis Staubli robot. The test was performed at 22-33 GHz. A set of measurements has been performed introducing a uniformly distributed random error in the range [0-1] mm, corresponding to ?/10 at 30 GHz. Errors are considered unknown. In the paper it will be shown that both in the classical and PWMS approaches the main beam is basically not affected by the introduced errors. The sidelobes are instead affected by such errors especially in the pattern cut where the beam is tilted. Such error levels obtained with the classical approach are comparable to those obtained with the PWMS approach, meaning that the latter is stable and against such type of perturbations.
NF-FF transformations have proven to be a convenient tool to accurately reconstruct the antenna pattern from NF measurements. In this framework, a very hot issue is the reduction of the time required to perform the measurements. To obtain a remarkable reduction of this time, nonredundant (NR) NF-FF transformations with planar spiral scannings have been developed in [1], by applying the NR representations of electromagnetic fields [2]. Optimal sampling interpolation (OSI) formulas have been used to efficiently reconstruct the massive NF data for the classical plane-rectangular (PR) NF-FF transformation from the NR spiral samples. The drastic measurement time-saving is due to the reduced number of needed NF samples acquired on fly, by adopting continuous and synchronized motions of the linear positioner of the probe and of the turntable of the AUT. However, such a time-saving is obtained at the expense of a nonuniform step of the spiral. Therefore, the linear positioner velocity is not constant, but must vary according to a not trivial law to trace the spiral, and this implies a complex control of the linear positioner. This work aims to develop an effective NF-FF transformation with planar spiral scanning for volumetric AUTs, wherein the spiral step is uniform and, hence, the linear positioner velocity becomes constant. To this end, the AUT is considered as enclosed in a sphere, the spiral is chosen in such a way that its step coincides with the sampling spacing needed to interpolate along a radial line according to the spatial band-limitation properties, and the NR representation along such a spiral is determined. Then, an OSI algorithm is developed to recover the NF data needed by the PR NF-FF transformation from the spiral samples. Numerical simulations assessing the accuracy of the developed NF-FF transformation will be shown.
Spherical near-field systems installed in shielded anechoic chambers are typically involved in modern automotive antenna measurements [1-3]. Such systems are often truncated at or close to the horizon to host the vehicle under test while limiting the size/cost of the chamber. The vehicle is usually placed on a metallic floor [4] or on a floor covered by absorbers [5]. The latter solution is intended to emulate a free space environment and is a key factor to perform accurate measurements down to 70 MHz. The availability of the free-space response also enables easy emulation of the car's behaviour over realistic grounds [6-7] while such emulations are more complex when a conductive ground is considered [8]. Conductive ground measurements also suffer from a strong interaction between the conductive floor and the measurement system and only in a limited number of situations such types of floor are a good approximation of realistic grounds (such as asphalts). However, the main advantage of conductive floor systems is the ease of accommodation of the vehicle under test which is simply parked in the center of the system. In absorber-based systems, instead, more time is generally needed to remove/place the absorber around the vehicle. Moreover, at low frequencies (70-400 MHz), large and bulky absorbers are normally used to ensure good reflectivity levels and the vehicle needs to be raised to avoid shadowing effect of absorbers. In this paper we investigate whether the measurement setup phase in absorber-based systems can be simplified by using smaller absorbers at low frequencies and/or not using them at all but considering conductive floors. The loss of accuracy in such scenarios will be studied considering a scaled vehicle and an implemented scaled automotive system where it is possible to access the full-spherical, real free-space scenario which is used as reference. The analysis is carried out considering (scaled) frequencies relevant to automotive applications in the 84-1500 MHz range. Two types of scaled absorbers, of different size and reflectivity, are considered to emulate the behaviour of the realistic full-scale 48-inch and 18-inch height absorbers. Measurements over metallic floor are included also in the analysis.
We present an antenna measurement methodology requiring a radically lower number of field samples than the standard Nyquist-based theory maintaining a comparable accuracy. simulations and partial knowledge of the geometry of the Antenna Under Test are combined to build a set of numerically defined expansion functions: the method uses basic knowledge of the antenna and the assumption that scattering from large surfaces can be predicted accurately by numerical tools; areas of the antenna such as feeding structures are treated as unknown and represented by equivalent electric and magnetic currents on a conformal surface. In this way, the complexity, and thus the number of unknowns, is dramatically reduced wrt the full problem for most antennas. The basis functions representing the full antenna are used to interpolate a radically reduced set of measured samples to a fine regular grid of Near Field (NF) samples in standard geometries. Regular NF to Far Field (FF) transformation techniques are then employed to determine the FF. The sampling reduction is evaluated compared to a regular sampling on standard Nyquist-complaint grids. The method can be employed in standard sampling ranges. In [1] asymptotic simulation tools were used to build the numerical basis. In this paper, methods based on Surface Integral Equations (SIEs) are used to compute currents and fields. The currents induced on the antenna structure by each elementary source are computed and used to evaluate the radiated field. Both electric and magnetic elementary sources are placed around the antenna and the SIE problems use a fast algorithm to evaluate matrix-vector products. The methodology is validated with planar and spherical acquisitions on a reflector antenna (MVG SR40) fed by a dual ridge horn SH4000 and in a multi-feed configurations (using several SH5000) at 18 and 30 GHz. Patterns obtained with down-sampled fast approach are compared to standard measurements. Down-sampling factors up to 8 are achieved maintaining very high correlation levels with standard techniques.
Characterizing the performance of automobile-mounted antennas has been an ongoing and evolving challenge for the antenna measurement community. Today, the automotive test environment poses unique challenges with its diversity and complexity of wireless on-board systems and the large electrical size of the test article. The evolution of cellular technologies over the past decade means that the basic mobile handset has now become a smartphone with significantly increased capability; this exact same trend has been mirrored by the automotive industry where we have witnessed the basic car radio and cassette player evolve into a multi-function infotainment unit. Modern vehicles include a multitude of wireless technologies, including cellular (2G, 3G, LTE), Bluetooth, WiFi, Global Navigation Satellite System (GNSS), collision avoidance radar, and more. Testing the complete vehicle is currently the only method available that certifies the correct mode of operation for each technology (including co-existence and interference) and also assures the manufacturer that the various sub-systems are performing as expected in the presence of all other sub-systems and the vehicle itself. While modern vehicles now function like large mobile devices, the conventional Over-the-Air (OTA) measurement systems and techniques available for small form factor devices (e.g. mobile phones) are ill-suited to testing such large devices. In this paper, we will highlight some of the unique challenges encountered in the automotive test environment. We will start by looking into existing methods of measuring radiation patterns of automobile-mounted antennas and providing a qualitative assessment of the various techniques with a focus on near-field solutions. A brief description of OTA testing will follow, coupled with an in-depth look into how techniques that are proven for handset type OTA measurements are being translated to automotive measurements. This section will provide a breakdown of key OTA test metrics, the measurement hardware typically required and key assumptions about the device under test. Finally, some performance tradeoffs and challenges associated with designing a multi-purpose antenna/OTA measurement system will be described.
Anechoic chamber performance is largely dependent on the chamber layout and effectiveness of the RF absorbing material. Over time, absorbers begin to degrade and need to be replaced. Also, new health and safety standards, combined with more modern building fire code regulations, can necessitate updates to a chamber's absorber layout. These factors led to the complete restoration of the Canadian Space Agency's largest anechoic chamber, which is located at the David Florida Laboratory in Ottawa, Canada. Ideally, the internal walls and ceiling of an anechoic chamber should be free of intrusions to facilitate the installation of RF absorbers. Unfortunately, because of the chamber's coupled structure to the building, this was not possible and an extensive sprinkler system and large air circulation vents were installed within the chamber. In this paper, their respective impacts on reflectivity performance is studied and novel mitigation techniques are introduced. Based on practical considerations, these techniques were used to conceal the infrastructure of the fire suppression system and HVAC ductwork inside the anechoic chamber; initial measurements appear to indicate their validity. The techniques and lessons learned from this exercise may be applied by others considering a similar endeavor.
The problem of modelling a radiator or a scatterer using an equivalent radiator is of interest in a large number of applications as, for example, antenna synthesis [1], electromagnetic compatibility [2] and the design of echo generators [3]. Such a modelling problem requires determining the shape and the dimensions of a radiating surface capable to generate, in a certain region of space, an electromagnetic field close to that produced by the radiator/scatterer of interest. The purpose of this contribution is that of proposing, for a fixed equivalent radiator's shape, an approach for the solution of such dimensioning issue. The solution is new and, at the best of our knowledge, the dealt with problem has not been yet addressed throughout the literature. The proposed approach relies on the use of the Singular Value Decomposition (SVDs) of the operators linking the radiator/scatterer to the field on the region of interest, say D, and the equivalent radiating panel to the field on D, again. The singular functions of such operators corresponding to the most significant singular values represent the spaces to which the fields radiated by the primary radiator/scatterer and that radiated by the equivalent one essentially belong to, respectively. The approach consists into determining the dimensions of the equivalent radiator minimizing the error by which the field radiated on D by the equivalent radiator approximates the primary radiated/scattered one. The error is expressed as a hermitian, definite positive quadratic form so that the problem amounts to the maximization of its minimum eigenvalue. Numerical results will be presented for an equivalent, planar radiator of rectangular shape.
Spherical Near-Field (SNF) systems using a swing arm gantry configuration have been the go to solution for automotive measurement systems. Recent advances in the automotive industry have warranted a need for SNF systems with high mechanical positioning accuracy supporting measurements up to 40 GHz and beyond. This paper presents the design and implementation of a new swing arm gantry positioner having an 8-meter radius and a radial axis to support high frequency SNF measurements. We first define the relation of the gantry axis to the global coordinate system and discuss primary sources of errors. Next, a robust mechanical design is presented including design considerations and implementation. We then present errors measured using a tracking laser interferometer for probe position through the range of gantry axis travel. Static corrections for probe positioning errors are implemented in the control system using the radial axis. The resultant residual error for the swing arm gantry is then shown to have the accuracy required for high frequency SNF measurements.
The Austin RCS Benchmark Suite has recently been introduced to enable quantitative and objective comparison of computational systems for solving electromagnetic scattering problems, particularly, those relevant to aerospace applications. In the last year, five sets of problems were added to it: dielectric almonds (problem set III-B), mixed material almonds (III-C, III-D), perfectly electrically conducting (PEC) aircraft models (IV-A), and dielectric aircraft models (IV-B). For each problem set, a range of lengths and frequencies of interests are identified, interesting features are highlighted, and datasets containing reference results (from measurements, analytical methods, or numerical methods) are shared online. Although data from several radar cross section (RCS) measurement campaigns of non-metallic targets are available in the literature, these lack the information necessary to precisely model the materials, target geometries, and measurement setups, to quantify uncertainties in the data, and to identify appropriate directions for improving computational methods' performance. This limits their utility for benchmarking computational systems. This article presents an expansion of the Suite to include problems with more complex materials and reference results from a measurement campaign that attempted to ameliorate the deficiencies of existing datasets. Specifically, a set of thin-plate problems are added to the Austin RCS Benchmark Suite to increase material diversity. These consist of problem sets II-B: thin perfect-electrically conducting (PEC) plates, II-C: thin dielectric plates, II-D: thin magnetic radar absorbing material (MagRAM) plates, and II-E: thin MagRAM-coated PEC plates. Reference RCS data that enables validation, RCS measurement and material property uncertainty quantification, and benchmarking are also provided by conducting a simulation-supported measurement campaign in a compact range. To facilitate reproducibility, a popular low-loss dielectric material and a commercially available MagRAM were chosen for these problems: The dielectric material for problem set II-C is PolymaxTM polylactic acid (PLA). For problem sets II-D/E, the ARC Technologies' DD-13490 material is used. Thin plates were manufactured and their RCS were measured at Lockheed Martin's Rye Canyon Facility. The monostatic RCS measurement results and supporting simulation results are available online. Performance data for simulations as well as RCS measurement results with accompanying uncertainty will be presented for problem sets II-B/C/D/E at the conference.
A set of new 3:1 Dual-Polarized Antennas has been developed for use in near-field ranges as the probe or range antenna and for use as a Compact Antenna Test Range (CATR) feed [1]. Key development parameters of the antenna are: a wideband impedance match to the coaxial feed line, E and H-plane 1 dB beam widths in excess of 30 degrees, -30 dB on axis cross-polarization, minimum polarization tilt and a phase center that varies over a small region near the aperture. To accomplish these design parameters, a family of range antennas has been developed and previously introduced [1]. Two versions of the antenna have been manufactured and tested for performance. A 2-6 GHz version has been developed using traditional machining techniques and a 6-18 GHz version has been produced using additive manufacturing (3D printing) techniques. In this paper, we focus on the performance of the 6-18 GHz antenna produced using additive manufacturing. The measured performance of the antenna will be presented and compared to previous simulation. The advantages of additive manufacturing for this type antenna will be discussed. Finally, the applicability of the antenna as a CATR feed and its use in near-field scanning will be discussed.
Tapered Chambers are best suited for antenna pattern measurements at low frequencies. The advantage of such chambers over rectangular shaped chambers would be achieving a desired performance level in terms of field uniformity and ripple at the quiet zone due to the low reflectivity of the chamber. To achieve such performance using a rectangular shaped chamber could lead to design of larger rooms and associated significant cost. Hence, this paper tries to analyze the characteristics of the tapered chambers using the novel Fourier analysis characterization method. The Fourier analysis method was applied on the transverse and longitudinal planar scan data at the quiet zone of a tapered chamber. The analysis yield the performance of the chamber at different frequencies depicting the plane wave behavior at the low frequencies and breakpoint of the plane wave behavior with increase in frequency. It also shows the images or reflection hotspots formed at the throat of the tapered section at the higher frequencies. In addition, the longitudinal scan analysis portray the reflections from the back wall of the chamber. In conclusion, the known concepts and ideologies of the tapered chamber design are reexamined from a different perspective based on the analysis results.
There is a growing need for wearable RF modules integrated into clothing for medical sensing applications. Also, there is a concurrent interest for RF devices to collect energy and power body-worn devices, such as biosensors used to monitor vital signs. To do so, an approach is to integrate into garments, near-field wireless power transfer and harvesting RF circuits that can collect ambient RF energy from nearby Wi-Fi sources to charge biosensors. In this regard, resonant RF harvesting structures for near-field power transfer have been explored before and demonstrated close to 100% of power transfer efficiency (PTE) at close distances. However, this impressive efficiency can only be realized when there is perfect polarization and special alignment between transmitter and receiver. Herewith, we provide an approach that mitigates the efficiency challenges due to misalignment. With the above in mind, we propose a new class of resonators referred to as corrugated X resonators that are resilient to lateral, angular, and diagonal misalignments. The resonators operate at 500 MHz. Using these resonators, their PTE was found to range from 40% to 60% across 1 to 10-cm distances for broadside direction. In case of only lateral misalignment in the direction perpendicular to the feedline, the PTE is 70% across the same distances. Also, a PTE of up to 85% was achieved when the misalignment was only in the direction of the feedline. At the meeting, we will present the design and performance of the developed low-profile resonators RF surfaces. The evaluation is done when integrated into wearable textile surfaces under near-field illumination for RF harvesting.
In the 5G communication, antenna array has been widely used for high-speed wireless communication. For reliable antenna array system, the failure diagnosis of antenna array is one of the most important problems that has been studied for a long time. The back-projection method using near-field measurements is a one of the failure diagnosis technique based on the plane-wave expansion. However, when antenna elements are densely placed, it is difficult to estimate the excitation coefficients of the antenna elements with the back-projection method, because the obtained images from the conventional back-projection method has only a resolution of one wavelength. In addition, since there is usually a trade-off between measurement accuracy and measurement time. Therefore, it is difficult to satisfy the both requirements of accuracy and short measurement time. We have reported the element failure detection algorithm using a 2-layer shallow neural network with planar near-field measurement last year. In this report, the element failure detection of antenna array is performed with a minimum number of measurement points while maintaining enough accuracy by learning the relationship between excitation coefficients of antenna array and the electric far-field distribution by a shallow neural network. In the case of 64-elements short dipole antenna arrays, the estimation error of excitation coefficients of antenna array less than 1% are achieved by our trained neural network with a minimum number of far-field measurements with 50 dB SNR. The detailed algorithm and simulation results will be reported in the full-paper and the presentation.