AMTA Paper Archive

This paper discusses the application of modern NF measurements and statistical analysis techniques to efficiently characterize the FF radiation pattern statistics of antennas and other EM emitters whose radiated EM fields vary erratically in a seemingly random manner. Such randomly-varying radiation has been encountered, for example, in measurements involving array antenna elements and reflector feed horn(s) containing active or passive devices that affect the relative phases and/or amplitudes of the pertinent RF signals in a non-deterministic manner [1-2]. In-Band (IB) as well as Out-Of-Band (OB) signals may be involved in some cases. Other possible randomly varying EM radiations include leakage from imperfectly-shielded equipment, connectors, cables, and waveguide runs [2- 4]       Previous work at GTRI [5-7] has shown that computations of key FF radiation pattern statistics  can be made based on NFFF transformations involving a) the sample average value of the complex electric field at each NF measurement point, b) the sample average value (a real number) of the standard deviation of the complex electric field at each NF measurement point, and c) the measured complex cross-covariance functions at all different NF measurement points. The key FF radiation pattern statistics of most interest are typically a) the statistical average FF radiation pattern, b) the standard deviation, c) the probability density function (p.d.f.), and d) the cumulative probability distribution (C.P.D.). Simulated data measurement protocols and the requisite statistical processing of the NF measured data will be presented and discussed in detail at the symposium.       The NF cross-covariance functions introduce a new level of complexity in NF measurements and analysis that is absent for “deterministic” EM field measurements because the cross covariance functions must be measured and processed for all different NF measurement points on the NF surface to compute valid Pattern FF statistics. However, pairs of linear or circular probe arrays can be used to great advantage to achieve tolerable NF measurement times for the cross covariance functions and the aforementioned NF statistical quantities, thereby enabling valid computations of the FF pattern statistics. The use of dual probe arrays will be presented and discussed in detail and compared with mechanical scanning of two “single” probes over two NF measurement surfaces. A technique for estimating the cross-covariance functions will be presented and compared with exact values.

MIL STD 461 is the Department of Defense standard that states the requirements for the control of electromagnetic interference (EMI) in subsystems and equipment used by the armed forces. The standard requires users to measure the unintentional radiated emissions from equipment by placing a measuring antenna at one meter distance from the equipment under test (EUT). The performance of the antenna at 1m distance must be known for the antenna to measure objects located at this close proximity. MIL STD 461 requires the antennas to be calibrated at 1 m distance using the Society of Automotive Engineers (SAE) Aerospace Recommended Practice (ARP) 958. This SAE ARP 958 document describes a standard calibration method where two identical antennas are used at 1m distance to obtain the gain at 1m for each antenna. In this paper the authors show using simulations that the SAE ARP 958 approach introduces errors as high at 2 dB to the measured gain and AF. To eliminate this problem the authors introduce a new method for calibrating EMC antennas for MIL STD 461. The Method is based on the well-known extrapolation range technique. The process is to obtain the polynomial curve that is used to get the far field gain in the extrapolation gain procedure, and to perform an interpolation to get the gain at 1 m. The results show that some data in the far field must be collected during the extrapolation scan. When the polynomial is calculated the antenna performance values at shorter distances will be free of near field coupling. Measured results for a typical antenna required for emissions testing per the MIL STD 461 match well with the numerical results for the computed gain at 1 m distance. Future work is required to study the use of this technique for other short test distances used in other electromagnetic compatibility standards, such as the 3 m test distance used by the CISPR 16 standard. Keywords: Antenna Calibrations, EMC Measurements, Extrapolation Range Techniques

Spherical spiral scanning involves coordinating the motion of two simultaneous axes to accomplish near-field antenna measurements along a line on a sphere that does not cross itself. The line would ideally start near a pole and trace a path along the sphere to the other pole. An RF probe is moved along this path in order to collect RF measurements at predefined locations. The data collected from these measurements is used along with a near-field to far-field transformation algorithm to determine the radiated far-field antenna pattern.  The method for transforming data collected along spherical spiral scan has been previously presented [1]. Later laboratory measurement studies have shown the validity of the spherical spiral scanning technique [2]. Here, the authors present a review of the spherical scanning technique and present recent advances and the applicability of the method to testing antennas mounted on automobiles. The method has the advantages of a reduction of the overall number of data points required in order to meet a minimum sampling requirement determined using non-redundant sampling techniques. This reduction in the number of data points and the advantage of moving two axes simultaneously result in a significant reduction in the time required to collect a set of measured data. Keywords: Spherical Near-Field, Telematics, Automotive References: [1] F. D’Agostino, C. Gennarelli, G. Riccio, and C. Savarese, “Theoretical Foundations of near-field to far-field transformations with spiral scannings,” Prog. In Electromagn. Res., vol. PIER 61, pp 193-214, 2006. [2] F. D’Agostino, F. Ferrara, J. Fordham, C. Gennarelli, R. Guerriero, and M. Migliozzi, “An Experimental Validation of the Near-Field to Far-Field Transformation with Spherical Spiral Scan,” Proc. Of the Antenna Measurement Techniques Association, 2012.

We previously demonstrated that 60 GHz planar near-field antenna measurements without external frequency conversion can provide far-field radiation patterns in good agreement with spherical near-field antenna measurements in spite of the cable flexing and thermal drift effects [P.I.Popa, S.Pivnenko, J.M.Nielsen, O.Breinbjerg, ”60 GHz Antenna Measurement Setup Using a VNA without External Frequency Conversion ”36thAnnual Meeting and Symposium of the Antenna Measurements Techniques Association, 12-17 October, 2014]. In this work we extend the validation of this 60 GHz planar near-field set-up to antenna diagnostics and perform a detailed systematic study of the extreme near-field of a standard gain horn at 60 GHz from planar and spherical near-field measurement data.  The magnitude and phase of all three rectangular components of the electric and the magnetic aperture fields are calculated, as is the main component of the Poynting vector showing the power flow over the aperture. While the magnitude of the co-polar electric field may seem the obvious object for antenna diagnostics, we demonstrate that there is much additional information in those additional quantities that combine to give the full picture of the aperture field. The usefulness of the complete information is illustrated with an example where the horn aperture is disturbed by a fault.  We compare the results of the planar and spherical near-field measurements to each other and to simulation results.

As millimeter-wave applications become more widely available technologies, there is a demand to know material properties for design and application purposes.  However, many mass produced materials are either not specified at these frequencies or the price materials can be costly. Therefore the easiest method for characterization is by measurement. Traditional methods of this measurement type involve the reflectivity of a fabric sample placed on a flat metallic reference plate. However, this method has some major difficulties at these high frequencies. For example, the surface of the reference plate must be very flat and smooth and must be carefully oriented such that their surface is precisely facing the transmitting and receive and antennas. Furthermore the electrically large size of the reference plate of this setup makes it difficult to measure in far-field and anechoic range time is expensive.  Resistive and conductive fabrics have applications such as shielding, anti-static, and radio wave absorption. Radio wave absorption and radar cross section engineering is currently of high interest to the automotive industry for testing newly emerging automotive radar systems. Such fabric measurement has already been utilized to accurately characterize artificial skin for radar mannequins to recreate the backscattering of human targets at 77 GHz. This paper presents a new and convenient method for measuring the reflective properties of conductive and resistive materials at millimeter wave frequencies by wrapping fabrics around a metallic reference cylinder. This new approach to fabric characterization method is able to obtain higher accuracy and repeatability despite the difficulties of measuring at high frequency.

During the last years, new algorithms, based on time filtering, spatial or modal filtering, have been designed for echo reduction techniques applied to antenna measurements. These algorithms have been used for different applications where the effect of the echoes is important, as far field system, VHF or UHF applications, automotive systems, small antennas, etc. The authors, in previous papers, have analysed the effect of different algorithms: time filtering (fft, non uniform dft or matrix pencial), modal filtering based on Spherical modes (MV-Echo) and spatial filtering based on Integral Equations (Insight) and holographic techniques (fft and dft) to cancel the effect of the reflections. This comparison has been applied to the measurements of a dipole antenna (SD1900) using a StarLab system. It is observed that each of the algorithms is better for different situations, depending on the source of the echo. For instance, time filtering techniques are good for reflections coming from different distances with respect the direct ray, but not so good for close reflections. In addition hey need a large frequency band to work properly. Spatial algorithms can correct the effect of positioners or other structures close to the antenna under test, but they are better for planar near field acquisitions and worse for classical single probe spherical near field where the antenna is rotated and probe is fixed (e.g. roll-over-azimuths systems). Moreover, they require extra information of the AUT geometry. This paper presents first a comparison of each algorithm and then, a combination of time and spatial techniques based on uniform or non-uniform DFT to take advantage of the benefits of each algorithm for different origins of the reflections.

Unlike most antenna performance parameters (directivity, beamwidth, and efficiency, e.g.), phase center is not strictly defined and warrants further clarification when used.  Put simply, the phase center is the point at which antenna radiation seems to emanate and is determined as the center of a spherical surface of constant phase in the far field.  For practical antennas, however, such a point is fictional and can only be established by minimizing the phase variation on a portion of the spherical surface over a smaller angle of interest, generally where the radiation intensity is greatest (e.g. the 3dB beamwidth).  Most commonly, the phase center is defined for a two dimensional planar cut parallel to the direction of propagation, for example the E or H plane of a horn. Knowledge of the phase center is particularly critical in the feeds of reflectors or lenses, where it is required to be located at the focal point of the reflecting or refracting structure to maximize aperture efficiency.  Due to its electro-mechanical properties the horn antenna has often been used as the feed for the above mentioned configurations.  For wideband applications, the stabilization of the phase center over the entire frequency band poses a significant challenge since this point generally tends from the mouth to the throat of a horn as frequency is increased.  The design discussed in this paper involves a feed horn operated in conjunction with a Lunenburg Lens for increased directivity and gain over 18-45 GHz bandwidth.  A design overview is discussed with the primary focus on phase stabilization considerations.  Methods for determining the phase center of the design are also discussed and compared.  These include analytical solutions using the aperture current approximation, simulations using method of moments and finite element method from FEKO and HFSS, respectively, as well as measurements taken in the anechoic chamber at the University of Colorado Boulder.

Compensated Compact Range Facilities are the state-of-the-art RF test facilities for spacecraft payload modules and/or antennas. The outstanding features of the compact range technique are the (a) real-time testing capability, (b) easy to use far-field measurement technique, (c) extremely high frequency capability, (d) end-to-end payload testing at multiple test zones due to scanning features, and last but not least the (e) considerable low cross-polar contribution over the full frequency band between 1 - 200 GHz which is one of the important parameters for telecommunication antenna testing.  Upcoming spacecraft antennas with single feed per beam configuration and broadband transponder requirements (up to 500 MHz) need rapid test environments for antenna and payload (end-to-end) measurement campaigns. For the desired wide frequency spectrum the Ka-Band and even higher bands (U, and V) are of interest for the next generation of telecommunication spacecraft antennas. Compensated Compact Ranges provide an excellent test environment for such scenarios. Recent developments for the range feeds up to 200 GHz, a new heavy load and highly accurate specimen positioner design, and the easy enlargeable reflector system within the existing chamber complete the picture of a state-of-the-art test facility for present and future spacecraft testing. The paper will explain the advantages of the selected system design and preferred technology with its resulting features to optimally cover the future requests focusing to new developments in the high frequency range. For typical spacecraft antenna scenarios a comparison between Compact Range and Near-Field facilities will demonstrate the applicability in the frequency range from 1 to 200 GHz. Beside the developed test set-up for the required measurement parameters, typical measurement times and achievable performance with its related error budget will be depicted.

Highly accurate antenna measurements can require precise alignment and positioning of the probe antenna to the antenna under test. The positioning of the antenna during acquisition can involve the movement of several simultaneous axes of motion. This places a global positioning accuracy requirement on the positioning system. To achieve precision in global positioning and alignment, an understanding of dominant error factors such as load induced deflection/resonance, thermal deflection, positioning error sources and mechanical alignment tolerances is essential. This paper focuses on how global accuracy and stability were achieved, addressing these factors, on a recently delivered large far field antenna measurement system.  The system involved eight axes of positioning with the ability to position 950 lbs antenna under test 19.5 ft above the chamber floor achieving 0.007 inch and 0.005 degrees positioning accuracy relative to the global range coordinate system. Stability of the probe antenna after motion was within 0.001 inch. Key Words: Global Position Accuracy, Far Field, Position Stability, Simultaneous Motion, Position Error Correction, High Accuracy, Precise Motion

A multi-probe array (MPA) spherical near-field antenna measurement system, comprised of COTS equipment, has been developed for testing UHF antennas mounted in an aircraft rotodome. The spherical probe radius is 5 meters, which accommodates a 24 ft. diameter rotodome. The probe array, arranged in a circular arc about the test zone center, provides rapid time multiplexed samples of dual polarized spherical theta angle measurements. These measurements are collected at incremental steps of spherical phi angles, provided by a floor azimuth turntable.  The rotodome is mounted on the azimuth turntable, and is rotated 360 degrees during a data collection. During one azimuth rotation, completed in a few minutes, a full set of 3D, dual polarized, multi-frequency near-field pattern data is collected. The data is transformed to full 3D far-field patterns in another few minutes, providing a complete rotodome test time within 15 minutes. The entire system is contained within a room 42’ x 42’ x 25’. This paper will describe the test requirements, physical requirements of the DUT, size constraints of the facility, and measurement speed goals. Alternate solutions and range geometries will be discussed, along with why the MPA solution is best given the requirements and size constraints. The system will be described in detail, including discussion of the room design, RF instrumentation, multi-probe array, positioning equipment, and controllers. Measurement results will be presented for test antennas of known pattern characteristics, along with other performance metrics, such as test times.

We have successfully designed and developed an innovative “CLose-range Antenna Scanner System” (or CLASS) suitable for measuring the far-field radiation pattern of installed antennae at short distances. The system consists of three key components: (1) a uniquely designed lens horn antenna that generates plane waves in close proximity, (2) a mechanical x-y scanner to scan the antenna-under-test, and (3) a customized stitching software to compute the far-field antenna pattern from the measured field information. The developed system has a scan area of 4.6 x 4.6 m, with resolutions of ±0.1mm in both the x and y traverse directions. The scanner structure is designed in a scalable fashion to cater for measurement of antenna installed at various locations (e.g. front and sides) on a platform. The system is capable of measurement from 1 to 18 GHz and generates far-field radiation pattern with a gain accuracy of ±1 dB.

For today's sophisticated antenna applications, the accurate knowledge of 3D radiation patterns is increasingly important. To measure the antennas under far-field conditions over a broad frequency band is hereby hardly impossible. By near-field to far-field transformation, one can overcome the difficulties of limited measurement distances. In common spherical near-field antenna measurement software, the transformation based on spherical mode expansion is typically implemented. These software tools only provide to correct the influence of first order azimuthal probe modes. The influence of the probe’s higher order modes though increases with shorter measurement distances. To measure a broad frequency range in one measurement set-up and to save time, dual ridged horns are popular candidates since they operate over a wide frequency range. The drawback is that they are probes of higher order. In this contribution, we will present an investigation on near-field measurements which are transformed into the far-field deploying the transformation technique based on spherical modes which is extended by a higher order probe correction capability. The resulting diagrams comparing first and higher order probe correction show that a correction is important in particular for the cross polarization In addition, the near-field data is transformed with an algorithm which employs a representation by equivalent currents. In this method, a higher order probe correction based just on the probe’s far-field pattern is integrated. The equivalent currents supported by an arbitrary Huygens surface allows to reconstruct the current densities close to the actual shape of the AUT which is mandatory for precise antenna diagnostics. Another issue needs to be accounted for regarding limited measurement distances and spherical modal expansion. While representing the AUT and the probe in spherical modes the radii of the spheres grow the more modes are included which depends on the sizes of the TX and the RX antennas. It has to be ensured that both spheres do not interfere.  All measurements were carried out in the anechoic chamber of our laboratory in which measurements starting at 1 GHz are practicable according to the dimension of the chamber and of the absorbers. Due to our restricted measurement distance of 0.57 m, all the above mentioned rules need to be considered. In conclusion, small anechoic chambers are also capable of delivering precise antenna measurements over a broad frequency range due to algorithms capable of higher order probe correction.

The spherical multipole based near-field far-field transformation is one of the most widespread algorithms for field transformation due to its very low computation time achieved by employing the fast Fourier transform (FFT) and imposing the utilization of first order probe antennas which obtain regularly distributed near-field samples on a spherical surface. Thus, huge efforts in highly accurate scanner system and antenna design are invested to fulfill the transformation algorithm requirements. In comparison, the recently developed inverse source reconstruction methods are very undemanding as they allow to use arbitrary probe antennas and arbitrarily shaped measurement surfaces as long as the probe’s relative position and orientation with respect to the device under test (DUT) is accurately known. Furthermore, the diagnostics capabilities of the algorithms give insight into the radiation mechanisms of the antenna. Although multilevel fast multipole boosted inverse source reconstruction algorithms such as the fast irregular antenna field transformation algorithm (FIAFTA) provide an excellent linearithmic complexity, their computation time is still higher than the one of the spherical transformation. The flexibility to process near-field samples on an irregular grid is yet only of interest for some challenging measurement scenarios where it is easier to determine the exact position and orientation of the probe than to accurately position it at certain grid points. Moreover, most antenna measurement facilities are already equipped with positioner systems for spherical scans. Therefore, a spherical multipole based transformation with higher order probe correction capability is proposed to perform a fast near-field far-field transformation. Once the far-fields and thus, the plane wave representation of the antenna has been obtained, a hierarchical plane wave representation is utilized to efficiently determine the equivalent sources of the antenna. For best sources localization and diagnostic features, equivalent surface currents on a Huygens’ surface enclosing the antenna are used. Their organization in a hierarchical octree is the key to a fast transformation from the antenna far-field to its equivalent sources. In this way, the blend of the spherical multipole based transformation and the hierarchical plane wave based field representation allows to profit from the benefits of both transformation approaches.

The source reconstruction or equivalents source method provides an accurate near-field representation of any radiating device in terms of equivalent electric and magnetic currents. The equivalent currents can be determined from measured near or far field data through a post-processing step involving the solution of an integral equation. The currents constitutes an accurate 3D electromagnetic model, maintaining near and far field properties of the measured device. A newly created link, enable the export of the model to a number of commercial computational electromagnetic (CEM) solvers in the form of a near-field Huygens box. Of special interest to the EMC community, equivalent current representation of measured devices are directly applicable in diagnostics/hot-spot finding and in the determination of radiated emission at any distance. The Huygens box, derived from measurements, is applicable in the simulation of emission in different scenarios when the device is in vicinity of different objects such as shielding, cables etc. This papers shows examples of diagnostics and emission analysis of a representative printed circuit board (PCB) based on commercially available near field measurement systems, post-processing and CEM tools.

Abstract Antenna far field phase pattern is important for some applications. It can be directly obtained in pattern measurement by far field test range (FFTR) or compact range (CR). However, it is found that the antenna far field phase pattern measured by current near field test range (NFTR) is not correct. For a uniform phase feeding plane array, its far field phase pattern should be near constant in 3dB beam width.  However, the antenna phase pattern measured by current NFTR looks square curve vs angle. This paper found out that the root cause of the error is due to different reference planes. Both the amplitude pattern and the phase pattern obtained by current NFTR, in fact, refer to the probe scanner plane, not the antenna plane. This shifting of the reference plane has no effect on amplitude pattern, but has effect on phase pattern. After that, a correction method is proposed. One example is used for the root cause finding and correction technique explanation. According to this paper, if one wants to get phase pattern using NFTR, it is necessary to measure the distance between AUT and probe aperture accurately so as to correct it accurately after measurement and obtain accurate phase pattern.

Among the near-field – far-field (NF–FF) transformation techniques, the one employing the plane-polar scanning has attracted a considerable attention [1]. In this framework, efficient sampling repre­sentations over a plane from a nonredundant number of plane-polar samples, which stays finite also for an unbounded scanning plane, have been developed, by applying the nonredundant sampling representa­tions of the EM fields [2] and assuming the antenna under test (AUT) as enclosed in an oblate ellipsoid [3] or in a double bowl [4], namely, a surface formed by two circular bowls with the same aperture diameter but eventually different lateral bends. These effective representations make possible to accu­rately recover the NF data required by the plane-rectangular NF–FF transformation [5] from a nonredun­dant number of NF data acquired through the plane-polar scanning. A remarkable reduction of the number of the needed NF data and, as a consequence, of the measurement time is so obtainable. However, due to an imprecise control of the positioning systems and their finite resolution, it may be impossible to exactly locate the probe at the points fixed by the sampling representation, even though their position can be accurately read by optical devices. Therefore, it is very important to develop an effective algorithm for an accurate and stable reconstruction of the NF data needed by the NF–FF transformation from the acquired irregularly spaced ones. A viable and convenient strategy [6] is to retrieve the uniform samples from the nonuniform ones and then reconstruct the required NF data via an accurate and stable optimal sampling interpolation (OSI) expansion. In this framework, two different approaches have been proposed. The former is based on an iterative technique, which converges only if there is a biunique correspondence associating at each uniform sampling point the nearest nonuniform one, and has been applied in [6] to the uniform samples reconstruction in the case of cylindrical and spherical surfaces. The latter, based on the singular value decomposition method, does not exhibit this constraint and has been applied to the nonredundant plane-polar [7] scanning technique based on the oblate ellipsoidal modelling. However, it can be conveniently used only when the uniform samples recovery can be split in two inde­pendent one-dimensional problems. The goal of this work is to develop these two techniques for compensating known probe position­ing errors in the case of the nonredundant plane-polar scanning technique using the double bowl modelling [4]. Experimental tests will be performed at the UNISA Antenna Characterization Lab in order to assess their effectiveness. [1] Y. Rahmat-Samii, V. Galindo Israel, and R. Mittra, “A plane-polar approach for far-field construction from near-field measurements,” IEEE Trans. Antennas Prop., vol. AP-28, pp. 216-230, 1980. [2] 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, 1998. [3] O.M. Bucci, F. D’Agostino, C. Gennarelli, G. Riccio, and C. Savarese, “NF–FF transformation with plane-polar scanning: ellipsoidal modelling of the antenna,” Automatika, vol. 41, pp. 159-164, 2000. [4] O.M. Bucci, C. Gennarelli, G. Riccio, and C. Savarese, “Near-field–far-field transformation from nonredundant plane-polar data: effective modellings of the source,” IEE Proc. Microw. Antennas Prop., vol. 145, pp. 33-38, 1998. [5] E.B. Joy, W.M. Leach, Jr., G. P. Rodrigue and D.T. Paris, “Application of probe-compensated near-field measurements,” IEEE Trans. Antennas Prop., vol. AP-26, pp. 379-389, May 1978. [6] O.M. Bucci, C. Gennarelli, G. Riccio, C. Savarese, “Electromagnetic fields interpolation from nonuniform samples over spherical and cylindrical surfaces,” IEE Proc. Microw. Antennas Prop., vol. 141, pp. 77-84, 1994. [7] F. Ferrara, C. Gennarelli, G. Riccio, C. Savarese, “Far field reconstruction from nonuniform plane-polar data: a SVD based approach,” Electromagnetics, vol. 23, pp. 417-429, July 2003

End-of-the-line over-the-air (OTA) testing of fully assembled wireless devices is one of the most important tests done in factories. It is designed to detect defective devices to avoid them being shipped out to the end customers. There are many requirements in designing over-the-air test systems for factory testing, including small factory real estate, measurement repeatability, and fast test time. These requirements prompt to challenges in OTA test system designs. Few existing widely-used test systems exist: near-field coupling systems where the test antenna is located very near the device’s antenna under test, small TEM cells, and shielded enclosures with one or several test antennas. Each technology has advantages and disadvantages, such as system size, defect detection capabilities/limitations, and performance measurement correlation to that from a far-field method. However, they all lack in dealing with improving test time with devices having technologies working with multiple simultaneous antennas/streams. For example, the current test time for a 2-antenna device (MIMO or received diversity capable devices) is doubled because each antenna chain is tested sequentially. Furthermore, possible coupling effect between antennas is not typically tested. The newly proposed OTA test system is an adaptive system with an array of test antenna elements inside a shielded enclosure. It takes advantage of the multi-path environment inside the enclosure to adapt itself and create a static channel environment with the specified requirement needs. For example, to improve test time for a 2-antenna device, the system groups the antenna elements of the system into two arrays to create two signal streams creating a 2x2-matrix channel with the cross-coupled matrix values minimized (e.g. minimization of the matrix condition number). This created static channel environment with optimized isolation between the two direct signal paths enables testing of the two antenna streams concurrently with minimized perturbation between the streams, hence reducing test time by almost half. The system will reconfigure the antenna elements for each test channel. This proposed new method of an adaptive over-the-air test system opens up to new ways of testing fully-assembled wireless devices in factories and also enables testing of certain performance qualities that current existing OTA test systems cannot perform.

The single-offset compact antenna test range (CATR) is a widely deployed technique for broadband characterization of electrically large antennas at reduced range lengths [1]. The nature of the curvature and position of the offset parabolic reflector as well as the edge geometry ensures that the resulting collimated field is comprised of a pseudo transverse electric and magnetic (TEM) wave. Thus, by projecting an image of the feed at infinity, the CATR synthesizes the type of wave-front that would be incident on the antenna under test (AUT) if it were located very much further away from the feed than is actually the case with the coupling of the plane-wave into the aperture of the AUT creating the classical measured “far-field” radiation pattern. The accuracy of a pattern measured using a CATR is primarily determined by the phase and amplitude quality of the pseudo plane-wave with this being restricted by two main factors: amplitude taper (which is imposed by the pattern of the feed), and reflector edge diffraction, which usually manifests as a high spatial frequency ripple in the pseudo plane wave [2]. It has therefore become customary to specify CATR performance in terms of amplitude taper, and amplitude & phase ripple of this wave over a volume of space, termed the quiet-zone (QZ). Unfortunately, in most cases it is not directly apparent how a given QZ performance specification will manifest itself on the resulting antenna pattern measurement. However, with the advent of powerful digital computers and highly-accurate computational electromagnetic (CEM) models, it has now become possible to extend the CATR electromagnetic (EM) simulation to encompass the complete CATR AUT pattern measurement process thereby permitting quantifiable accuracies to be easily determined prior to actual measurement. As the accuracy of these models is paramount to both the design of the CATR and the subsequent determination of the uncertainty budget, this paper presents a quantitative accuracy evaluation of five different CEM simulations. We report results using methods of CATR modelling including: geometrical-optics with geometrical theory of diffraction [3], plane-wave spectrum [4], Kirchhoff-Huygens [4] and current element [3], before presenting results of their use in the antenna pattern measurement prediction for given CATR-AUT combinations. REFERENCES [1]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. [2]M. Philippakis, C.G. Parini, “Compact Antenna Range Performance Evaluation Uging Simulated Pattern Measurements”, IEE Proc. Microw. Antennas Propag., Vol. 143, No. 3, June 1996, pp. 200-206. [3]G.L. James, “Geometrical Theory of Diffraction for Electromagnetic Waves”, 3rd Edition, IET Press, 2007, ISBN 978-0-86341-062-8. [4]S.F. Gregson, J. McCormick, C.G. Parini, “Principles of Planar Near-Field Antenna Measurements”, IET Press, 2007.

The National Institute of Standards and Technology (NIST) has developed computationally efficient algorithms for probe location and polarization compensation in near- to far-field transformations for use when measurements are not made on the standard canonical grids. A major application of such methods is at higher frequencies, where it is difficult or impractical to locate a probe to required tolerances for the standard transforms. Our algorithms require knowledge of the actual position of the probe at the measurement points. This information can be furnished by state-of-the-art optical tracking devices. Probe position information is routinely obtained by the NIST CROMMA (Configurable Robotic MilliMeter-wave Antenna) Facility. Even at lower frequencies, probe-location compensation techniques allow in principle, the use of less precise and therefore, less expensive scanning hardware. Our approach also provides the flexibility to process data intentionally collected on nonstandard grids (plane-polar, spiral, etc.) or with mixed geometries (such as a cylinder with a hemispherical or planar end cap).   We present simulations and actual probe position compensation results at 183 GHz. The possibility of compensating for known variations in the probe pointing is considered.

It is a common practice when computing radiation patterns from non-uniformly sampled planar fields to interpolate the samples into a regular grid [1], although it might cause inaccuracies due to the interpolation process. The non-uniform fast Fourier transform (NUFFT) has been applied to process near field measurements in non-uniform planar grids with arbitrary precision [2], and also to analyze aperiodic arrays [3]. However, samples are usually treated as punctual sources. In this contribution, an efficient and accurate method to calculate the far field radiated by non-uniformly sampled planar fields which comply the Nyquist theorem using the non-uniform fast Fourier transform (NUFFT) is shown. The method takes into account the amplitude of the unit cell radiation pattern, which allows to compute more accurately the copolar and crosspolar components of the far field with regard to the array factor [3], which models the samples as punctual sources. For measured fields it is assumed that post-processing has been done, for instance, taking into account probe corrections. Because the NUFFT is precision-dependent, a discussion of how its accuracy can affect the computed radiated fields will be carried out. Numerical examples will be provided to show the accuracy and performance of the NUFFT with regard to the FFT and direct evaluation of the far fields. Finally, a study of computing times comparing the FFT, NUFFT and direct evaluation will be presented. References [1] Y. Rahmat-Samii, L. I. Williams, and R. G. Yaccarino, “The UCLA bi-polar planar-near-field antenna-measurement and diagnostics range,” IEEE Antennas Propag. Mag., vol. 37, no. 6, pp. 16–35, Dec. 1995. [2] R. C. Wittmann, B. K. Alpert, and M. H. Francis, “Near-field antenna measurements using nonideal measurement locations,” IEEE Trans. Antennas Propag., vol. 46, no. 5, pp. 716–722, May 1998. [3] A. Capozzoli, C. Curcio, G. D'Elia, and A. Liseno, “Fast phase-only synthesis of conformal reflectarrays,” IET Microw. Antennas Propag., vol. 4, no. 12, Dec. 2010.