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.)
Traditional rectangular waveguide measurements are operated in the frequency regime of the dominant TE10 mode. The general guideline for determining the dominant mode frequency regime is to operate 25% above the TE10 mode cutoff to avoid high dispersion and 5% below the TE20 cutoff to avoid higher-order mode excitation. The X-band waveguide for example, with cross-sectional dimensions 0.9 inches by 0.4 inches, has a TE10 and TE20 cutoff frequency of 6.56 and 13.12 GHz, respectively. Using the above guideline, the approximate bandwidth of operation is 8.2-12.4 GHz. In addition, coax-to-waveguide adapters must be employed in order to connect the network analyzer coaxial cables to the rectangular waveguide sections. In modern (i.e., commercially off the shelf - COTS) microwave coax-to-waveguide adapters, tuning stubs are employed to minimize voltage standing wave ratio and thus maximize energy coupling into the waveguide sections. Unfortunately, these tuning stubs are placed in asymmetric patterns that can cause coupling into the TE20 mode, which is the very reason why one must operate at a frequency of at least 5% below this mode to safely avoid higher-order mode contamination. The goal here is to show that, by designing symmetric coax-to-waveguide adapters, excitation of the TE20 mode can be avoided for operational frequencies above the TE20 cutoff. Thus, the frequency of operation may be extended to the TE11 mode (the next higher-order mode that can exist) having a cutoff frequency of 16.16 GHz. Consequently, the operational frequency band is enhanced from 8.2-12.4 GHz to 8.2-15.4 GHz, representing a 70% improvement in operational bandwidth. A comparison of newly-designed symmetric and COTS asymmetric coax-to-waveguide adapters for material characterization measurements will be provided and advantages/limitations will be discussed.
Current antenna measurement techniques are based on the underlying idea that echoes generated by nearby structures should be avoided. Indeed, the absence of echoes allows a precise measurement of the line-of-sight radiation of the antenna under test (AUT), via mechanical rotation to span some or all spatial directions until the radiation pattern is formed. In this paper, this idea is challenged by introducing an alternative test approach that generates controlled echoes and use them as a useful source of information. Preliminary results are presented and it is shown how frequency diversity can be fruitfully used to retrieve the free space radiation pattern. A special care is given to the conditioning of the mathematical problem. Accordingly, it is shown how the different parameters involved in the set-up influence the feasibility of the technique. The proposed technique is expected to lead to a faster characterization of the AUT, as the need for mechanical rotation is cut down.
Reverberation chambers (RCs) are important measurement tools, and thus it is often required to simulate their behaviour numerically. However, due to their special characteristics, especially for high Q factors, they are often considered too challenging for application of standard numerical software. In particular, a recent publication [1] listed the perceived state-of-the-art in integral equation modelling of RCs, and identified numerous unsolved problems. The present paper illustrates that using Higher-Order (HO) basis functions in the integral equation discretization can allow the numerical analysis of relatively large RCs to be performed with limited computer resources. Applying a dedicated HO Multi-level Fast Multipole Method scheme allows even larger problems to be solved. After a discussion and brief review of existing methods for RC modelling, we will turn to a description of the key features of HO basis functions and their related MLFMM implementation, focusing on how they allow surpassing some of the challenges faced by lower-order discretizations. Then, several RC test-cases are analyzed, drawing comparisons to other results from the relevant litterature. The conclusion is that with use of HO basis functions and a thorough MLFMM implementation, some of the challenges identified in [1] can be overcome. [1] H. Zhao, “MLFMM-Accelerated Integral-Equation Modeling of Reverberation Chambers,” IEEE Antennas and Propagation Magazine, vol. 55, no. 5, pp. 299–307, Oct. 2013.
Broad band, dual polarized probes are becoming increasingly popular options for use in near-field antenna measurements. These probes allow one to reduce cost and setup time by replacing several narrowband probes like open-ended waveguides (OEWG) with a single device covering multiple waveguide bands. These probes are also ideal for production environments, where chamber throughput should be maximized. Unfortunately, these broadband probes have some disadvantages that must be quantified and corrected for in order to make them viable for high accuracy near-field measurements. Most of these broadband probes do not have low cross polarization levels across their full operating bandwidths and may also have undesirable artifacts in the main component of their patterns at some frequencies. Both of these factors will result in measurement errors when used as probes. Furthermore, the use of a dual port RF switch adds an additional level of uncertainty in the form of port-to-port channel balance errors that must be accounted for. This paper will describe procedures to calibrate the pattern and polarization properties of broad band, dual polarized probes with an emphasis on a newly developed polarization correction algorithm. A simple procedure to measure and correct for amplitude and phase imbalance entering the two ports of the near-field probe will also be presented. Measured results of the three calibration procedures (pattern, polarization, channel balance) will be presented for a dual-polarized, broad band quad-ridged horn antenna. Once calibrated, this probe was used to measure a standard gain horn (SGH) and will be compared to baseline measurements acquired using a good polarization standard open-ended waveguide (OEWG). Results with and without the various calibration algorithms will illustrate the advantage to using all three routines to yield high accuracy far-field pattern data.
The typical antenna measurement system setup working above 20 GHz makes use of frequency multipliers and harmonic mixers, usually working in standard waveguide bands, and thus several parts need to be procured and interchanged to cover several frequency bands. In this paper, we investigate an alternative solution which makes use of a standard wideband VNA without external frequency conversion units. The operational capability of the Planar Near-Field (PNF) Antenna Measurement Facility at the Technical University of Denmark was recently extended to 60 GHz employing an Agilent E8361A VNA (up to 67 GHz). The upgrade involved procurement of very few additional components: two cables operational up to 65 GHz and an open-ended waveguide probe for tests in U-band (40-60 GHz). The first tests have shown good performance of the PNF setup: 50-60 dB dynamic range and small thermal drift in magnitude and phase, 0.06 dB and 6 degrees peak-to-peak deviations over 4 hours. A PNF measurement of a 25 dBi Standard Gain Horn was carried out and the results were compared to those from the DTU-ESA Spherical Near-Field Facility with a good agreement in the validity region. Uncertainty investigations regarding cable flexing effects at 60 GHz have shown that these introduce an uncertainty of about 0.02 dB (1 sigma) around the main beam region indicating a very good performance of the PNF setup.
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. 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 mast (polystyrene or Plexiglas) mounted on a rotating positioning system (vertical axis). The combination of the two rotation capabilities allows full 3D near field monostatic RCS characterization. This paper investigates the influence of the material of the mast supporting the target under test. Across several measurement steps, we compare different RCS measurement results of canonical targets in order to eliminate the unwanted RCS measurement contribution due to the mast. The aim is to find out the mast which disturbs the least the RCS of the target under test but still compatible with the measurement facility ARCHE 3D. All these measurements are also compared to Near Field and Far Field calculations taking into account the material of the supporting mast.
The synthesis of a specific field distribution in a certain volume with a given set of sources is an issue which arises in acoustics as well as in electromagnetics. Field Synthesis is of increasing interest for over the air (OTA) testing of multiple input multiple output (MIMO) based communication devices as arbitrary multipath communication channels can be simulated synthesizing the corresponding field distribution around the device under test (DUT). Plane-wave Field Synthesis methods have already been applied to improve the quality and extents of the quiet zone region of compact antenna test ranges (CATR). Furthermore, by synthesizing a plane wave field in a test region for an antenna under test (AUT), using an array of probe antennas in its near-field region, near-field far-field transformations (NFFFT) can be performed. Since there exists a variety of important applications for electromagnetic Field Synthesis, a Field Synthesis approach with high flexibility and low computational complexity is presented in this contribution. Usually, depending on the application, a single moving probe antenna or an array of probe antennas is used to synthesize a desired field distribution in the test zone volume where the DUT will be placed. The challenge is to determine appropriate excitation signals for the individual probe antennas. For that purpose an equation system is iteratively solved which arises from the boundary condition for the tangential field components on the surface of the test volume. As a consequence of the uniqueness theorem, equality of the desired and synthesized tangential field components induces that the desired and synthesized field distribution are identical in the source free test volume. Field testing on the surface of the test volume is performed by vector testing functions defined on a triangular mesh of the test zone surface enabling field synthesis in arbitrarily shaped test volumes. For accelerated evaluation of the coupling between probe antennas and vector testing functions, principles of the fast multipole method (FMM) are adopted. The implied plane wave expansions enables to incorporate the radiation characteristic of the probe antenna sources just by directly employing its plane wave spectrum representation which is nothing else but its far-field pattern. Additionally, the multilevel approach minimizes the number of translation operations between source and receiver boxes organized in a hierarchical oct-tree. Altogether the approach is applicable to arbitrarily shaped test volumes and arbitrarily arranged probe antennas and still shows a linearithmic complexity. In this contribution, detailed insight in the Field Synthesis method is given. Results for synthesized field distributions for arbitrarily shaped test volumes are presented. Finally the application of plane-wave Field Synthesis to NFFFT is shown for synthetic as well as for real near-field antenna measurement data.
The US NEXRAD weather surveillance Doppler radar (WSR-88D) was recently upgraded to polarimetric capability. This upgrade permits identification of precipitation characteristics and type, thus providing the potential to significantly enhance the accuracy of radar estimated rainfall, or water equivalent in the case of frozen hydrometeors. However, optimal benefits are only achieved if errors induced by the radar hardware are properly accounted for through calibration. Hardware calibration is a critical element in delivering accurate meteorological information to the forecast and warning community. The calibration process must precisely measure the gain of the antenna, the Polarimetric bias of the antenna, and the overall gain and bias of the receive path. The absolute power measurement must be accurate to within 1 dB and the bias between the Polarimetric channels must be known to within 0.1 dB. These requirements drive a need for precise measurement of antenna characteristics. Engineers and scientists with the NEXRAD program employ solar scanning techniques to ascertain the absolute gain and bias of the 8.53 m parabolic center fed reflector antenna enclosed within a radome. They are also implementing use of daily serendipitous interference strobes from the sun to monitor system calibration. The sun is also used to adjust antenna gain and pedestal pointing accuracy. This paper reviews the methods in place and under development and identifies some of the challenges in achieving the necessary calibration accuracies.
After decades of international launches and varying space expeditions the low Earth orbit (LEO) has become littered with man-made objects and debris. With over 22,000 objects larger than a softball, and hundreds of thousands in smaller size existing, remediation efforts must take place to ensure the continuation of both collision free space flight and orbits. While smaller objects are difficult to track, and would consume more resources, the larger bodied debris offer a means to collect greater volumes of orbital debris clutter with less operations. In an effort to assist in the architectural design of microwave remote sensors, for the detection, tracking and identification of the large tumbling bodies, apriori knowledge of their relevant electromagnetic scattering parameters is essential. This paper work focus on the scattering phenomenology from possible large bodied orbital debris, such as rocket bodies, whose geometries are publically available. The results will strengthen existing data sets, Radar architectures, required signal processing, and even guidance navigation and control (GNC) routines that would be supported by resultant sensor information. Data products developed from commercially available electromagnetic simulation software will be presented, and the induced phenomenological scattering differences from the geometric variations between the possible targets will also be discussed.
Within RF measurement systems, engineers commonly wish to know how much phase ripple will be present in a signal based on a given signal-to-noise ratio (SNR). In a past AMTA paper (Measurement Considerations for Antenna Pattern Accuracy, AMTA 1997), John Swanstrom presented an equation which demonstrated how the bound on the phase error could be calculated from the peak SNR value. However, it can be shown that the Swanstrom bound is broken when the signal has a peak SNR value of less than approximately 15 dB. This paper introduces a new equation that bounds the maximum phase error of a signal based on the signal’s peak SNR value. The derivation of this new bound is presented, and comparisons are made between the old Swanstrom bound and the new bound. In addition, the inverse relationship (i.e., calculating the SNR value of a signal from phase-only measurements) is investigated. In the past, analytical equations for this relationship have been presented by authors such as Robert Dybdal (Coherent RF Error Statistics in IEEE Trans. on Microwave Theory and Techniques) and Jim P.Y. Lee (I/Q Demodulation of Radar Signals with Calibration and Filtering in a Defense Research Establishment Ottawa publication). The analytical equations for calculating the SNR value using phase-only measurements are reviewed and discussed, and a brand new numerical relationship based on a polynomial curve fitting technique is proposed.
This paper introduces a new spherical near-field sampling grid, which we shall refer to as the equally-spaced sampling grid, where the sample points are spread over the surface of the measurement sphere using an almost-uniform triangular tessellation of the unit sphere. By spacing the points evenly over the surface of the sphere, the grid uses 50% the number of sample points as the classical grid. In addition, it can be shown that the equally-spaced grid requires the use of fewer points than required by the commonly cited “Nyquist criteria” at certain values of ka. Since the spherical modes required for the SNF-to-FF transform cannot be obtained from the equally-spaced grid through traditional FFT methods, the solution is demonstrated by other means. The details behind a brute-force matrix inversion method and a brief discussion on the condition number of the equally-spaced sampling grid will be included. Some comparisons will be shown between results obtained using the brute-force matrix inversion method and results obtained using the more computationally-efficient conjugate gradient method (as proposed by Wittmann, Alpert, and Francis).
Electrically small antennas present tremendous design challenges. Plagued with a small radiation resistance and high quality factor (thus narrow bandwidth), these types of antennas are difficult to accurately measure. For use in HF communication applications, the problems associated with the entire development cycle become even more pronounced. This paper focuses on the development of two such electrically small HF antennas for a vehicular platform, specifically the Amphibious Assault Vehicle (AAV). The primary design objective is to develop antennas that operate over the entire near-vertical incidence (NVIS) band (2 – 10 MHz) with a minimum of 3kHz bandwidth. Additional design objectives are low profile, broadside directive pattern, and high power handling capability. The inverted L antenna and the half loop antenna were selected as probable candidates for this application. At 2 MHz, the antenna – vehicle system fits within the envelope ka < 0.2, where k is the free space wave number and a is the radius of a sphere completely enclosing the radiator. The full scale antenna design and performance were evaluated using method of moments and finite element method codes FEKO and HFSS respectfully. It is observed that the presence of the real ground plane poses a serious challenge for well established modeling techniques and considerable care must be exercised to obtain credible design data. For measurement validation and characterization of the antenna/vehicle interaction, a set of scaled antenna and vehicle prototypes were developed. Rapid prototyping and 3D printing were employed to build a scaled model (1:50 scale) of the complete antenna – vehicle system. The step-by-step process from the computational model to the measurement validation is discussed along with the description of the adopted fabrication techniques. In the concluding section of the paper, the measured results from the scaled model are presented alongside the simulated results. The good agreement between these results paves the way towards the successful use of such scaled model testing for more complicated antenna designs in the future.
Historically, the inverse synthetic aperture radar (ISAR) reflectivity assumption has been used in the implementation of Image-Based Near Field-to-Far Field Transformations (IB-NFFFT) to estimate monostatic far field radar cross-sections (RCS) from monostatic near field radar measurements. The ISAR assumption states that all target scattering occurs at the location of the incident field excitations, i.e., the target is composed entirely of non-interacting localized scatters. Certain non-localized scattering phenomenon cannot be effectively handled by the IB-NFFFT approach with the ISAR assumption. Here we have used the adaptive Gaussian representation, which is a joint time-frequency decomposition technique, to coherently decompose near field measured data into two subsets of scattering features: one subset of localized scatterers and the other of non-localized scatterers. The localized scattering features are processed through the IB-NFFFT as typical, which includes compensating for the R4 fall-off present in the near field measured data. The non-localized scattering features, more appropriately scaled, are then coherently added back in to the post-NFFFT localized scattering phase history. Although this does not properly transform the non-localized scattering features into the far field, it does avoid the over-estimation error associated with improperly compensating distributed non-localized scattering features by a R4 power fall off based strictly on downrange position.
If a planar near-field (PNF) scanner is large and there is insufficient temperature regulation in the chamber to keep ordinary thermal expansion/contraction from causing unacceptable position errors, then consideration must be given to compensation techniques that can adjust for the changes. Thermal expansion/contraction will affect almost everything in the chamber including the floor, the scanner structure, the encoder or position tapes, the AUT support, and the mount for any extra instrument(s) used to measure and correct for position error. Since the temperature will generally cycle several times during a lengthy acquisition, error-correction solutions must account for the dynamic nature of the temperature effects. This paper describes a new automated tracking-laser compensation subsystem that has been designed and developed for very large horizontal PNF systems. The subsystem is active during the acquisition to account for both static and dynamic errors and compensates for those errors in all three dimensions. The compensation involves both open-loop corrections for repeatable errors with high spatial frequency and closed-loop corrections for dynamic errors with low spatial frequency. To close the loop, laser data are measured at a user-defined interval between scans and each scan that follows the laser measurements is fully compensated. The laser measurements are fully automated with no user interaction required during the acquisition. The challenges, goals, and assumptions for this development are listed, high-level implementation considerations are provided, and resulting measured data are presented.
Inter-comparisons and validations of antenna measurement ranges are useful tools allowing the detection of various problems in the measurement procedures, thus leading to improvements of the measurement accuracy and facilitating better understanding of the measurement techniques. The maximum value from a validation campaign is achieved when a dedicated Validation Standard (VAST) antenna specifically designed for this purpose is available. The widespread use of the known VAST-12 (12 GHz) antenna, developed by the Technical University of Denmark (DTU) in 1987 and operated by the DTU-ESA Facility since 1992, demonstrates the long-term value of the dedicated VAST antennas [1]. The driving requirements of VAST antennas are their mechanical stability with respect to any orientation of the antenna in the gravity field and thermal stability over a given operational temperature range. The mechanical design shall ensure extremely stable electrical characteristics with variations typically an order of magnitude smaller than the measurement uncertainty. At the same time, it must withstand high g-loads under frequent transportations and it shall also support convenient handling of the VAST antenna (practical electrical and mechanical interfaces, low mass, attachment points for lifting, etc.). Today, there is a well identified need for increased operational frequencies to get access to large bandwidth for broadband communication. Upcoming satellite communication services utilize up/down link at K/Ka-bands, while the use of Q/V bands is contemplated for the feeder links in the coming years. In response to this need, a millimeter-wave VAST (mm-VAST) antenna is currently under development in a collaboration between the DTU and TICRA under contract from the European Space Agency. In this paper, the electrical and mechanical requirements of the DTU-ESA mm-VAST antenna are discussed and presented. Potential antenna types fulfilling the electrical requirements are briefly reviewed and the baseline design is described. The emphasis is given to definition of the requirements for the mechanical and thermal stability of the antenna, which satisfy the stringent stability requirement for the mm-VAST electrical characteristics. [1] S. Pivnenko et al., “Comparison of Antenna Measurement Facilities with the DTU-ESA 12 GHz Validation Standard Antenna within the EU Antenna Centre of Excellence”, IEEE Trans. Antennas Propagat., 2009, vol. 57, no. 7. pp. 1863-1878.
In this work, we propose a new method for measuring the gain of a circularly polarized antenna at millimeter wave (mm-wave) and terahertz (THz) frequencies. Traditional methods of CP-gain measurement require measurement of linearly polarized gain in two orthogonal planes. However, for mm-Wave frequencies, this method is not practical since rotation of antenna under test can cause errors in measured phase. To avoid this, we present a novel phase-less method. For validation, circularly polarized slot array antenna under test is measured at 106 GHz. The antenna design and fabrications process is also presented.
One design of a smart plasma antenna is to surround a plasma or metal antenna by a plasma blanket in which the plasma density can be varied. In regions where the plasma frequency is much less than the antenna frequency, the antenna radiation passes through as if a window exists in the plasma blanket. In regions where the plasma frequency is high the plasma behaves like a perfect reflector with a reactive skin depth. Hence by opening and closing a sequence of these plasma windows this design can be computerized to electronically steer or direct the antenna beam into any and all directions. The plasma windowing design is one approach to the smart plasma antenna design. The beamwidth can vary from an omnidirectional radiation pattern with all the plasma windows open or a very directional radiation pattern when only one plasma window is open. The advantages of the plasma blanket windowing design are: 1. Beam steering of one omnidirection antenna with the plasma physics of plasma windowing. 2. A reconfiguable directivity. 3. The beamwidth can vary from an omnidirectional radiation pattern with all the plasma windows open to a directional radiation pattern with less than all the plasma windows open.
Satellite plasma antennas benefit from the lower thermal noise at the frequencies they operate. Ground-based satellite antennas point at space where the thermal noise is about 5K. A low thermal noise, high data rate satellite plasma antenna system is possible with low noise plasma feeds and a low noise receiver. Satellite plasma antennas can operate in the reflective or refractive mode. Satellite plasma antennas can be flat or conformal and effectively parabolic. Electromagnetic waves reflecting off of a bank of plasma tubes get phase shifted as a function of the plasma density in the tube. This becomes an effective phase array except that the phase shifts are determined by the plasma density. If the plasma density in the tubes is computer controlled, the reflected beam can be steered or focused even when the bank of tubes is flat or conformal. In the refractive mode, the refraction of electromagnetic waves depends upon the density of the plasma. In the refractive mode, steering and focusing can be computer controlled even when the bank of tubes is flat. For two-dimensional steering and/or focusing, two banks of plasma tubes are needed. Receivers can be put behind satellite plasma antennas operating in the refractive mode.
Plasma frequency selective surfaces (FSS) use plasma instead of metal for the FSS elements. In conventional metal FSS, each layer has to be modeled using numerical methods and the layers are stacked in such a way to create the desired filtering for antenna radomes. Genetic algorithms are sometimes used to determine the stacking needed for the desired filtering. This can be complicated, numerically expensive process. The plasma frequency selective surfaces can be tuned to a desired filtering by varying any or all of the density, size, shape, and spacing of the plasma elements for reconfigurable filtering in antenna radomes. As the density of the plasma is increased, the plasma skin depth becomes smaller until the elements behave as metallic elements and we create filtering similar to FSS with metallic elements. Up until the metallic mode for the plasma, our theory and experiments showed that the plasma FSS had a continuous change in filtering. A model for a plasma FSS is developed by modeling the plasma elements as half wavelength and full wavelength dipole elements in a periodic array on a dielectric substrate. The theoretical model with numerical predictions predicted results in good agreement with our experiments on the plasma FSS.
We present a microwave near-field imaging method using Fractional Fourier Transform (FrFT). Since the FrFT of finite chirp signal including a Fresnel integral, the scattering field of target in Fresnel zone can be described in the FrFT form. It is a valuable expression of scattering, because it unified the expression of scattering field in three different range along with distance, Rayleigh zone, Fresnel zone, and Fraunhofer zone. Then, one can get the target image from the Fresnel scattering field using the inverse FrFT (IFrFT) technology. The fast algorithm of FrFT or IFrFT has the same algorithm complexity as Fast Fourier Transform (FFT), which made a fast near-field imaging method compare to conventional method like Back Projection (BP), Time Domain Correlation (TDC), etc. Two near-field imaging results, one simply target and one complex target, will be presented to prove the method.