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.)
This paper presents two methods for measuring dynamic antenna patterns of phased arrays in a compensated compact range. The first method uses the turntable of the compact range to counter steer the antenna beam. The dynamic pattern is created by measuring single points of the pattern over time. This method is successfully tested, and the measurement results show the effect of phase jumps during the steering process. The second method extends the range of application to fast steering phased arrays by decoupling the antenna scan angle and the azimuth angle of the turntable.
This contribution details a procedure to collect and process necessary data to describe the antenna patterns of PAAs using a planar near-field (NF) range. It is proposed that a complete characterization methodology involves not only capturing beam-steered antenna patterns, but also measuring embedded element patterns, exhaustive testing of the excitation hardware of the antenna under test (AUT), and performing a phased array calibration technique. Moreover, to demonstrate the feasibility of the proposed approach, the methodology is applied onto a 2x8 microstrip patch PAA, proving its utility and effectiveness. Finally, by means of the collected data, any array pattern could be predicted by post-processing, as proven by the great agreement found between a measured pattern and its computed predicted version.
An efficient technique, which allows the correction of known positioning errors in a near-field to far-field (NF-FF) transformation with planar wide-mesh scanning (PWMS), is here developed and experimentally assessed. The corresponding NF-FF transformation from correctly positioned samples allows a remarkable reduction of the acquisition time with respect to the classical plane-rectangular one, since the NF data needed by this last are accurately recovered from a reduced number of PWMS samples through a 2-D optimal sampling interpolation expansion, attained by modeling the antenna under test with a double bowl and applying the nonredun-dant sampling representations to the voltage detected by the scanning probe. When the PWMS samples are affected by known probe positioning error, the voltage values at the sampling points, set by the representation, are unknown and can be efficiently retrieved from the inaccurately positioned ones by means of a singular value decomposition based technique.
Traditional near-field to far-field transformation algorithms based on modal expansion are unable to deal with arbitrary measurement surfaces. To approach these problems, a matrix inversion method can be used to retrieve the spherical wave expansion (SWE) of the antenna under test (AUT) fields. Modeling the antenna with a set of multiple SWEs centered at arbitrary points over its surface offers a flexible approach for the solution of field transformation problems over arbitrary surfaces. The coefficients of each SWE are obtained using an iterative inversion approach where the matrix-vector products can be replaced by multilevel operators based on recursive aggregations and interpolations of the partial SWE fields, reducing the computational complexity from í µí± ¶(í µí² í µí¿) to í µí± ¶í µí² í µí¿ í µí°¥í µí°¨í µí° í µí². The proposed algorithm is tested using synthetic data and measurements showing good scalability and reduced transformation error.
Positioning in near-field antenna measurements is crucial and often an absolute position accuracy of ?\50 is required. This can be difficult to achieve in practice, e.g. for robotic arm measurement systems and/or high frequencies. Therefore, optical measurement devices are used to precisely measure the position and orientation. The information can be used to correct the position and orientation during the measurement or in the near-field to far-field transformation. The latter has the benefit that the measurement acquisition is typically faster because no additional correction movements are needed. Different methods for correction of non-ideal measurement positions in r, ? and f have been presented in the past. However, often not only the relative position but also the orientation between the antenna under test (AUT) and the probe coordinate system is not perfect. So far, correction and investigation of the related non-ideal probe orientations has been neglected due to the assumption that the probe receiving pattern is broad. In this paper, non-ideal probe orientations will be investigated and a spherical wave expansion procedure which corrects non-ideal probe orientations and positions will be presented. This is achieved by including an arbitrary probe pointing in the probe response calculation by additional Euler rotations of the probe receiving coefficients. The introduced pointwise higher-order probe correction scheme allows an exact spherical wave expansion of the radiated AUT field. The transformation is based on solving a system of linear equations and, thus, has a higher complexity compared to Fourier-based methods. However, it will be shown that most of the calculations can be precomputed during the acquisition and that solving the linear equation system can be accelerated by using iterative techniques such as the conjugate gradient method. The applicability of the proposed method is demonstrated by measurements where an intentional misalignment is introduced. Furthermore, the method can be used to include full probe correction in the translated spherical wave expansion algorithm. In conclusion, the proposed procedure is a beneficial extension of spherical wave expansion methods and can be applied in different measurement scenarios.
Time and Direction of Arrival (TADOA) analysis of field probe data has been an accepted method for characterizing stray signals in an antenna measurement range for many years ([1], [2]). Recent uncertainty investigations at the OneRY range have shown a need for increased resolution to isolate and characterize energy in TADOA images so that resources can be carefully applied to reduce the uncertainty from these stray signals. This is accomplished by modeling the TADAO image as the solution to a Basis Pursuit (BP) l1 minimization problem. This paper outlines the model development and shows concrete examples from OneRY field probe data where BP allows for the identification of stray energy which was previously difficult to find. We also show how the BP optimization context can be using to remove contamination from the data through the inclusion of additional basis functions ([3]). I.J. Gupta, E.K. Walton, W.D. Burnside, “Time and Direction of Arrival Estimation of Stray Signals in a RCS/Antenna Range,” Proc. of 18th Annual Meeting of the Antenna Measurement Techniques Association (AMTA '96), Seattle WA, September 30-October 3, 1996, pp. 411-416. I.J. Gupta, T.D. Moore, “Time Domain Processing of Range Probe Data for Stray Signal Analysis,” Proc. of 21st Annual Meeting of the Antenna Measurement Techniques Association (AMTA '99), Monterey Bay CA, October 4-8, 1999, pp. 213-218. B.E. Fischer, I.J. LaHaie, M.H. Hawks, T. Conn, “On the use of Basis Pursuit and a Forward Operator Dictionary to Separate Specific Background Types from Target RCS Data,” Proc. of 36th Annual Meeting of the Antenna Measurement Techniques Association (AMTA '14), Tucson AZ, October 12-17, 2014, pp. 85-90.
The paper summarizes the performance of a new near-field to far-field (NF/FF) transform approach for a VHF vehicle mounted AUT test case, and compares the approach with the spherical measurement approach. The NF/FF transformation is based on the solution of an inverse problem in which the measured NF and predicted FF values are attributed to a set of equivalent electric and magnetic surface currents which lie on a convex arbitrary surface that is conformal to the antenna under test (AUT). The NF points are conformal to the AUT, reducing the number of samples and relaxing positioning requirements used in conventional spherical, NF/FF geometries. A pseudo inversion of the matrix representing the mapping of the equivalent sources into the near-field samples is obtained by using the singular value decomposition (SVD), which is used to form an approximation of the inverse of the matrix. This inverse, when multiplied by the NF measurement vector, solves for the efficiently radiating components of the current, which are used to compute the FF in a straightforward manner. Keywords—Antenna Near-Field to Far-Field Transformation, Electromagnetic Inverse Problems.
The principles of near-field antenna measurements and scanning in Cartesian and spherical coordinates are well established and documented in the literature, and in standards used on antenna ranges throughout government, industry, and academic applications. However the measurement methods used and the mathematics that are applied to compute the gain and radiation of the pattern of the test antenna from the near-field data assume typically that the antenna is operating in free space. This leaves several questions open when dealing with antennas operating over a lossy ground plane, such as the ocean damp soil, etc. In this paper, we shall discuss some of the motivation behind an examination of the physics and mathematics involved in performing a near-field antenna measurement over a seawater ground plane. Examples of past work in this are shall be discussed along with some of the challenges of performing far field antenna measurements in the presence of the air-sea interface. These discussions lead to some fundamental questions about how one defines gain in this environment and whether or not a near field approach could be beneficial. This will lead to some discussion of when and how the existing modal field expansions used in near-field measurements may need to be adjusted to account for the presence of the ground plane created by the ocean surface. An example of the limiting case of an antenna operating over a metallic ground plane will be discussed as a stepping stone to the more general problem of an antenna operating over a lossy ground plane.
We investigate all-metal 3D printing as a viable option for millimeter wave applications. 3D printing is finding applications across many areas and may be a useful technology for antenna fabrication. The ability to rapidly fabricate custom antenna geometries may also help improve cub satellite prototyping and development time. However, the quality of an antenna produced using 3D printing must be considered if this technology can be relied upon. Here we investigate a corrugated feed horn that is fabricated using the powder bead fusion process for use in the PolarCube cube satellite radiometer. AlSi10Mg alloy is laser fused to build up the feed horn, including the corrugated structure on the inner surface of the horn. The intricate corrugations, and tilted waveguide feed transition of this horn made 3D printing a compelling and interesting process to explore. We will discuss the fabrication process and present measurement data at 118.7503 GHz. Gain extrapolation and far-field pattern results obtained with the NIST robotic antenna range CROMMA are presented. Far-field pattern data were obtained from a spherical near-field scan over the front hemisphere of the feed horn. The quasi-Gaussian HE11 hybrid mode supported by this antenna results in very low side lobe levels which poses challenges for obtaining good SNR at large zenith angle during spherical near field measurements. This was addressed through using a single alignment and electrical calibration while autonomously changing between extrapolation and near-field measurements using the robotic arm in CROMMA. The consistency in parameters between extrapolation and near-field measurements allowed the extrapolation data to be used in-situ as a diagnostic. Optimal near-field scan radius was determined by observing the reflection coefficient S11 during the extrapolation measurement. The feed horn-to-probe antenna separation for which |S11| was reduced to 0.1 dB peak-to-peak was taken as the optimal near-field scan radius for the highest measurement SNR. A comparison of these measurements to theoretical predictions is presented which provides an assessment of the performance of the feed horn.
One of the sources of the measurement errors in outdoor antenna test ranges, when testing from VHF through C-Band, is the ground reflected signal between probe antenna and the antenna under test (AUT). Those errors are due to antenna(s) relatively large beam width(s) at these frequencies, especially when AUT is placed on the large platform such as an aircraft. If reflected wave is not eliminated by the use of absorbers at the reflection point or redirection by the use of diffraction fences, then the range operates as a ground reflection range (GRR), where the reflected signal creates a lobbing pattern when the direct and reflected signals are overlaying in- and out-of-phase as a function of position and frequency, causing undesirable amplitude variations at the test point. Ground reflections may be a major cause of error for GRR measurements when testing large antennas or antennas mounted on large structures which require a large displacement of the AUT during the antenna pattern collection process. A concept of using vertically positioned two-element array probe antenna (source antenna) to suppress ground-reflected signals in GRR-s is presented in this article. Suppression is achieved by pointing first null of the probes gain pattern towards the reflection point on the ground. All analytical evaluations are based on geometrical optics approach. Comparison of the proposed approach to a traditional single-element probe (source) antenna approach, demonstrates a significant improvement in measurement accuracy. Estimates and verifications of analytical evaluations are based on Computational Electromagnetics (CEM) modeling tool such as WIPL-D code. Simulations are performed in the VHF frequency band (200 MHz).
Reconfigurable antennas provide the ability to electronically change the antenna’s performance, which allows the antenna’s band of operation and gain pattern to be rapidly adapted to meet system requirements. A cylindrical, conformal reconfigurable antenna is presented which tunes over a wide band and provides full 360° azimuth coverage. The antenna maintains a realized gain (with mismatch and loss) better than a dipole from 800 MHz to 3 GHz, using the antenna’s gain to compensate for losses in the antenna. The antenna is designed and characterized with the cylinder’s bottom over a finite ground plane (no other antenna ground planes are used). The antenna is constructed using a modular approach out of a series of identical boards which act as antenna pixels. Each pixel contain four RF switches (one for each side of the board) along with contacts for control and ground wires. By fragmenting the reconfigurable antenna into individual pixel boards, one can construct elements of arbitrary size and shape with the primary physical constraint being how densely the electronics can be fabricated. By providing flexibility to scale in size, the antenna implementation can be optimized for more gain or for a smaller footprint. Two scaled versions of the same architecture have been constructed out of the same pixels to demonstrate the flexibility of the approach. In this paper we present data demonstrating more than 2 dBi gain from 1.2 GHz to 2.5 GHz band with beamwidths as narrow as 60°. Beam patterns are presented for GSM-900, GMS-1900, and WiFi frequencies. Finally, we will show the antenna element’s ability to maintain gain in a specific direction while forming a null over a series of offset angles.
The results of integrated antenna measurements are often severely distorted by reflections from the measurement environment. In order to feed passiveintegrated antennas wafer probes have to be used. Wafer probes are not only electrically large, but are also located in the immediate environment of the antenna undertest (AUT) and reflect part of the radiated signals. This causes significant distortions and erroneous results in radiation pattern, directivity, and gain measurements.Custom wafer probes have been used to reduce reflections for meaningful measurement results, but these special probes are difficult to fabricate and expensive. If the antenna is measured within an active system that generates the transmit signal, wafer probes are not required to feed the AUT, but bond wires, circuit elementsclose to the antenna, and parasitic radiation of surface waves also add distortions, which still limit the achievable accuracy of the measurements. In this paper modal filtering is used to mitigate the influence of these unwanted distortions in post-processing for both standard wafer probe and active antennameasurements. In the first part of the paper the performance of the post-processing technique is assessed for standard probe measurements at 160 GHz by comparing the post-processed results to a measurement of the same antenna using a custom made wafer probe that was designed for minimum reflections. In the second part modal filtering is used to reduce unwanted reflections for an active antenna measurement at 160 GHz. When the active circuitry that generates thetransmit frequency is integrated on the same chip as the AUT, the phase of the transmit signal is unknown. As the phase information is required for the post processing,a static external probe antenna is used as a reference to eliminate the phase drift of the measured signal. It is shown that modal filtering can be applied to integrated antenna measurements above 100 GHz and that reflections from wafer probes, bond wires, and the PCB canbe reduced significantly for passive and active antenna measurements, respectively.
Millimeter wave vehicular radar operating in the 77 GHz band for automatic emergency breaking (AEB) applications in detecting vehicles, pedestrians, and bicyclists, test data has shown that the radar cross section (RCS) of a target decreases significantly with distance at short range distances typically measured by automotive radar systems, where the reliable detection is most critical. Some attribute this reduction to a reducing illumination spot size from the antenna beam pattern. Another theory points to the spherical phase front due to measurement in the Fresnel region of the target, when the distance for the far-field zone is not met. The illumination of the target depends on the antenna patterns of the radar, whereas the Fresnel region effects depend on the target geometry and size. Due to fluctuations in measured data for RCS as a function of range in the near-field, upper and lower bounds for the target RCS versus range have been determined empirically as a method for describing the expected RCS of target. So far, the range-dependent RCS bounds used in AEB test protocols have been determined empirically. The study discussed in this paper aims to study the underlying physics that produces range-dependent RCS in near field and provide analytical model of such behavior. The resultant analytical model can then be used to objectively determine the RCS upper and lower bounds according to the radar system parameters such as antenna patterns and height. A comparison of the analytically predicted model and empirical near-field RCS as a function of range data will be presented for pedestrian, bicyclist, and vehicle targets.
Modern cars are equipped with a large number of antennas which are strongly integrated with the car. A full characterization of the radiating properties of the entire vehicle is thus typically required. In order to characterize the radiating properties of the installed antennas, large measurement systems accommodating the full vehicle are required. As in standard antenna measurements, a full spherical near field (NF) scanning around the car is desirable in order to perform an accurate NF/FF transformation. However, due to size and weight of the Device Under Test (DUT) and/or economic factors a full spherical scan is often unfeasible. For this reason, truncated spherical scanners (such as hemispherical) are typically involved. A classic solution is to combine hemispherical scanning with a metallic ground plane which is assumed to be a Perfect Electric Conductor (PEC) in the NF/FF transformation. However, the PEC ground-plane is less representative of realistic automotive environments such as asphalt that is strongly dielectric. A further drawback is the strong scattering from the large metallic ground-plane which highly compromises the NF measurements at low frequencies. In many situations, it is thus desirable to perform the NF measurements in a condition similar to free-space by using absorber materials on the floor. It is well-known that standard NF/FF transformations applied to partial spherical acquisitions generates the so called truncation errors. Such errors are stronger at lower frequencies due to the lower number of spherical modes for fixed DUT size. Moreover, typical antennas for automotive applications are generally low directive thus, the impact of the truncation on the measured pattern is often non-negligible. In such cases advanced post-processing techniques must be involved to mitigate the effect of the truncation errors. In this paper two truncation error mitigation techniques will be compared when applied to automotive measurements performed in free-space conditions. The first technique is an iterative process which at each iteration applies a modal filtering based on the size of the DUT. The second technique is based on the computation of the equivalent currents of the DUT over an equivalent surface which acts as spatial filter. Both techniques give excellent mitigation performance with different computational effort. The good agreement between two different techniques effectively defining the lower bound for what can be successfully mitigated by post processing techniques.
Among the near-field - far-field (NF-FF) transformations, that with spherical scan [1] is the most appealing due to its feature to allow the whole radiation pattern reconstruction of the antenna under test (AUT). To get a considerable measurement time saving, spherical NF-FF transformations for AUTs with one or two predominant dimensions, requiring a minimum number of NF data, have been developed in [2], by using the nonredundant sampling representations of the electromagnetic (EM) fields [3] and adopting a prolate or oblate ellipsoid to shape the AUT. Another effective possibility to save the measurement time is to make faster the scan by collecting the NF data through continuous and synchronized movements of the probe and AUT. To this end, NF-FF transformations with spherical spiral scan have been recently proposed. They rely on the nonredundant representations and use optimal sampling interpolation (OSI) formulae [3] to effectively recover the NF data needed by the traditional spherical NF-FF transformation [1] from the acquired ones. The nonredundant sampling representation on the sphere from spiral samples and the related OSI expansion have been developed in [4-6] by adopting a spherical AUT model and choosing the spiral pitch equal to the sample spacing needed to interpolate along a meridian. Then, NF-FF transformations with spherical spiral scan for long or quasi-planar AUTs [7] have been obtained by applying the unified theory of spiral scans for non-volumetric AUTs [8]. Unfortunately, due to practical constraints, it is not always possible to mount the AUT in such a way that it is centered on the scanning sphere centre. In this case, the number of NF data required by the NF-FF transformation [1] and the related measurement time can remarkably increase, due to the corresponding grow of the minimum sphere radius. Aim of this work is the development of a fast and accurate nonredundant NF-FF transformation with spherical spiral scan suitable for quasi-planar antennas, which requires practically the same number of NF data both in the centered and offset mountings of the AUT. To this end, an offset mounted quasi-planar AUT is modeled as contained in a oblate ellipsoid, and an effective representation of the probe voltage over the scanning sphere, using a minimum number of samples collected on a proper spiral wrapping it, is developed by applying the unified theory of spiral scans for non-volumetric AUTs [8] in the spherical coordinate system having the origin coincident with the AUT centre at distance from the scanning sphere one. The related OSI expansion allows to accurately reconstruct the NF data required for the NF-FF transformation. [1] J. Hald, J.E. Hansen, F. Jensen, F.H. Larsen, Spherical near-field antenna measurements, J.E. Hansen, (ed.), London, Peter Peregrinus, 1998. [2] O.M. Bucci, C. Gennarelli, G. Riccio, C. Savarese, “Data reduction in the NF–FF transformation technique with spherical scanning,” Jour. Electr. Waves Appl., vol. 15, pp. 755-775, June 2001. [3] O.M. Bucci, C. Gennarelli, C. Savarese, “Representation of electromagnetic fields over arbitrary surfaces by a finite and nonredundant number of samples,” IEEE Trans. Antennas Prop., vol. 46, pp. 351-359, March 1998. [4] O.M. Bucci, F. D’Agostino, C. Gennarelli, G. Riccio, C. Savarese, “NF–FF transformation with spherical spiral scanning,” IEEE Antennas Wireless Prop. Lett., vol. 2, pp. 263-266, 2003. [5] J F. D’Agostino, F. Ferrara, J.A. Fordham, C. Gennarelli, R. Guerriero, M. Migliozzi, “An experimental validation of the near-field - far-field transformation with spherical spiral scan,” IEEE Antennas Prop. Magaz., vol. 55, pp. 228-235, Aug. 2013. [6] F. D’Agostino, C. Gennarelli, G. Riccio, C. Savarese, “Theoretical foundations of near-field–far-field transformations with spiral scannings,” Prog. in Electr. Res., vol. 61, pp. 193-214, 2006. [7] R. Cicchetti, F. D’Agostino, F. Ferrara, C. Gennarelli, R. Guerriero, M. Migliozzi, “Near-field to far-field transformation techniques with spiral scannings: a comprehensive review,” Int. Jour. Antennas Prop., vol. 2014, ID 143084, 11 pages, 2014. [8] F. D’Agostino, F. Ferrara, C. Gennarelli, R. Guerriero, M. Migliozzi, “The unified theory of near–field–far–field transformations with spiral scannings for nonspherical antennas,” Prog. in Electr. Res. B, vol. 14, pp. 449-477, 2009.
The portable antenna measurement system PAMS was developed for arbitrary and irregular near-field scanning. The system utilizes a crane for positioning of the near-field probe. Inherent positioning inaccuracies of the crane mechanics are handled with precise knowledge of the probe location and a new transformation algorithm. The probe position and orientation is tracked by a laser while the near-field is being sampled. Far-field patterns are obtained by applying modern multi-level fast multipole techniques. The measurement process includes full probe pattern correction of both polarizations and takes into account channel imbalances. Because the system is designed for measuring large antennas the RF setup utilizes fiber optic links for all signals from the ground instrumentation up to the gondola, at which the probe is mounted. This paper presents results of the Ka-band test campaign in the scope of an ESA/ESTEC project. First, the new versatile approach of characterizing antennas in the near-field without precise positioning mechanics is briefly summarized. The setup inside the anechoic chamber at Airbus Ottobrunn, Germany is shown. Test object was a linearly polarized parabolic antenna with 33dBi gain at 33GHz. The near-fields were scanned on a plane with irregular variations of over a wavelength in wave propagation. Allowing these phase variations in combination with a non-equidistant grid gives more degree of freedom in scanning with less demanding mechanics at the cost of more complex data processing. The setup and the way of on-the-fly scanning are explained with respect to the crane speed and the receiver measurement time. Far-fields contours are compared to compact range measurements for both polarizations to verify the test results. The methodology of gain determination is also described under the uncommon near-field constraint of coarse positioning accuracy. Finally, the error level assessment is outlined on the basis of the classic 18-term near-field budgets. The assessment differs in the way the impact of the field transformation on the far-field pattern is evaluated. Evaluation is done by testing the sensitivity of the transformation with a combination of measured and synthetic data.
This paper presents the results of a recent study concerning the computational electromagnetic simulation of a spherical near-field (SNF) antenna test system. The new plane-wave scattering matrix approach [1, 2] allows many of the commonly encountered components within the range uncertainty budget, including range reflections, to be included within the model [3]. This paper presents the results of simulations that verify the utility of the spherical mathematical absorber reflection suppression (S-MARS) technique [3, 4] for the identification and subsequent extraction of artifacts resulting from range reflections. Although past verifications have been obtained using experimental techniques this paper, for the first time, corroborates these findings using purely computational methods. The use of MARS is particularly relevant in applications that inherently include scatterers within the test environment. Such cases include instances where a SNF test system is installed within an existing compact antenna test range (CATR) as is the configuration at the recently upgraded Queen Mary University of London (QMUL) Antenna Laboratory [5, 6]. Thus, this study focuses on this installation with results of CEM simulations and actual range measurements being presented. The method enables a quantitative measure of the levels of suppression offered by the MARS system. References A.C. Newell, S.F. Gregson, “Estimating the Effect of Higher Order Modes in Spherical Near-Field Probe Correction”, Antenna Measurement Techniques Association (AMTA) 34th Annual Meeting & Symposium, Bellevue, Washington October, 2012. A.C. Newell, S.F. Gregson, “Computational Electromagnetic Modelling Of Spherical Near-Field Antenna Test Systems Using Plane Wave Spectrum Scatting Matrix Approach”, Antenna Measurement Techniques Association (AMTA) 36th Annual Meeting & Symposium, Tucson, Arizona, October, 2014. C.G. Parini, S.F. Gregson, J. McCormick, D. Janse van Rensburg “Theory and Practice of Modern Antenna Range Measurements”, IET Press, 2014, ISBN 978-1-84919-560-7. G.E. Hindman, A.C. Newell, “Reflection Suppression in a large spherical near-field range”, Antenna Measurement Techniques Association (AMTA) 27th Annual Meeting & Symposium, Newport, RI, October. 2005. A.D. Olver, C.G. Parini, “Millimetre-wave Compact Antenna Test Range”, JINA Nice, November 1992. C.G. Parini, R. Dubrovka, S.F. Gregson, "CATR Quiet Zone Modelling and the Prediction of 'Measured' Radiation Pattern Errors: Comparison using a Variety of Electromagnetic Simulation Methods" Antenna Measurement Techniques Association (AMTA) 37th Annual Meeting & Symposium, Long Beach California, October 2015.
The accurate alignment of antennas and field probes is a critical aspect of modern antenna metrology systems, particularly in the millimeter-wave region of the spectrum.Commercial off-the-shelf robotic arms provide a sufficient level of positional accuracy for many industrial applications.The Antenna Metrology Project in the Communications Technology Laboratory at the National Institute of Standards and Technology has shown that path-corrected commercial robotic arms, both in hardware and software analysis, can be used to achieve sufficient positioning and alignment accuracies (positioning error ~ /50) for antenna characterization measurements such as gain extrapolation and near-field pattern out to 183 GHz [1]. Position correction is achieved using a laser tracker with a 6 degree of freedom sensor attached to the robot end effector.The end effector’s actual position, measured using the laser tracker, is compared to its commanded position and a path correction is iteratively applied to the robot until the desired level of accuracy is achieved in the frequency range of interest.At lower frequency ranges (< 40 GHz), sufficient positional accuracy can be achieved, without path correction, using a using a calibrated kinematic model of the robot alone [2].This kinematic model is based on knowledge of the link frame transformations between adjacent links and captures deviations due to gravitational loading on the joints and small mechanical offsets between the joints.Additionally, the calibration procedure locates the robot’s base frame in the coordinate system of the robot’s end effector.Each link frame is described by four physical quantities, known as Denavit-Hartenberg (DH) parameters [3]. We performed calibration measurements of our CROMMA system’s DH parameters over a working volume of ~1 m3.We then use the laser tracker to compare the robot’s positional accuracy over this working volume with and without the calibrated kinematic model applied.The path errors for the calibrated case set an upper frequency limit for uncorrected antenna characterization measurements. [1]D. R. Novotny, J.A. Gordon, J.R. Guerrieri, “Antenna Alignment and Positional Validation of a mm Wave Antenna System Using 6D Coordinate Metrology, ” Proceedings of the Antenna Measurements Techniques Association, pp 247-252, 2014 [2]R.Swanson, G. Balandran, S. Sandwith, “50-micron Hole Position Drilling Using Laser Tracker Controlled Robots, ” Journal of the CMSC, Vol 9, No 1, Spring 2014 [3].J.J. Craig, “Introduction to Robotics: Mechanics and Control, 3rd ed.,” New Jersey, Prentice Hall, 2004, pp. 62-69
At NIST, we have developed a precision, wide-band, mmWave modulated-signal source with traceability to primary standards. We are now extending the traceability path for this modulated-signal source into free space to be used for verifying over-the-air measurements in 5G, wireless receivers. However, to obtain a traceable modulated signal in free space, the full scattering matrix of the radiating antenna must be measured. We have extended the extrapolation methods used at NIST, based on the work of Newell, et al. [1]. The extrapolation measurement provides a very accurate, far-field, on-axis, scattering matrix between two antennas. When combined with scattering-matrix measurements made with permutations of pairs of three antennas, far-field scattering, and, thus, gain, is obtained for each antenna. This allows an accurate extrapolation of the antenna’s near-field pattern. We have incorporated the extrapolation fitting algorithms into a Monte Carlo uncertainty engine called the NIST Microwave Uncertainty Framework (MUF) [2]. The MUF provides a framework to cascade scattering matrices from various elements, while propagating uncertainties and maintaining any associated correlations. By incorporating the extrapolation measurements, and the three-antenna method into the MUF, we may provide traceability of all measurement associated with the gain, including the scattering parameters. In this process, we studied several aspects of the gain determination. In this work, we show simulations determining the efficacy of filtering to reduce the effect of multiple reflection on the extrapolation fits. We also show comparisons of using only amplitude (as is traditionally done) to using the full complex data to determine gain. Finally, we compare uncertainties associated with choices in the number of expansion terms, systematic alignment errors, uncertainties in vector network analyzer calibrations and measurements, and phase error introduced by cable movement. With these error mechanisms and their respective correlations, we illustrate the NIST MUF analysis of the antenna scattering-matrix with data at 118 GHz. [1] A. C. Newell, R. C. Baird, and P. Wacker “Accurate Measurement of Antenna Gain and Polarization at reduced distances by an extrapolation technique” IEEE Transactions on Antennas and Propagation. Vol. 21, No 4, July 1973 pp. 418-431. [2] D. F. Williams, NIST Microwave Uncertainty Framework, Beta Version. NIST, Boulder, CO, USA, Jun. 2014. [Online]. Available: http://www.nist.gov/pml/electromagnetics/related-software.cfm
Measurement of the radiation properties of low gain antenna operating at VHF frequencies is well known to be a challenging task. Such antennas are sometimes tested in outdoor Far Field (FF) ranges which are unfortunately subject to errors caused by the electromagnetic pollution and scattering from the environment. Near Field (NF) measurements performed in shielded anechoic chambers are thus preferable to outdoor ranges. However, also in such cases, the accuracy of the results may be compromised by the poor reflectivity of the absorbing material which might be not large enough wrt the VHF wavelength. Other source of errors may be caused by the truncation of the scanning area which generates ripple on the FF pattern after NF/FF transformation. Spherical multi-probe systems developed by MVG are optimal measurement solution for low directive Device Under Test (DUT). Such systems allow to perform a quasi-full spherical acquisition combining a rotation of the DUT along azimuth, with a fast electronically scanned multi-probe vertical arch. The DUT can be accommodated on masts made of polyester material which allows to minimize the interaction with the DUT. Measurements of low directive device above 400 MHz performed with such type of systems have been demonstrated to be accurate and extremely fast in previous publications. In this paper, measurements of a low directivity antenna, performed at VHF frequencies in a MVG spherical multi-probe system, will be presented. The antenna in this study is an array element, part of a larger array, which has been developed for space-born AIS applications. Gain and pattern accuracy of the measurement will be demonstrated by comparison with full wave simulation of the tested antenna.