Reviews of Electromagnetics, Vol.

II, 2023
Roadmap paper
DOI: 10.53792/RoE/2023/23003

Reviews of Electromagnetics
Roadmap paper

Antenna Measurement Challenges and

Opportunities
Guest Editors: Olav Breinbjerg1 , Manuel Sierra Castañer2 *

Abstract
This Roadmap overviews present challenges and opportunities for the development of antenna measurement
techniques and technologies to support the all-pervasive and ever-increasing demand for radio-frequency wireless
systems in modern society. The Roadmap comprises 19 inspiring contributions by 34 leading experts in antenna
measurements.

Key terms
Antenna Measurements; Wireless Systems; Challenges; Future Techniques and Technologies
1 ElMaReCo, Copenhagen, Denmark
2 ETSITelecomunicación, Universidad Politécnica de Madrid, Spain
*Corresponding editor: manuel.sierra@upm.es
Received: 21/05/2023, Accepted: 01/12/2023, Published: 12/12/2023

Modern society relies increasingly on well-functioning wireless perature; e.g., from almost zero to many hundreds of Kelvins.
systems - for communication, for sensing, or for energy transfer Also, antenna testing faces demands of increasing accuracy,
- and wireless systems rely significantly on well-functioning an- decreasing cost and time, need for characterization in produc-
tennas. Though computational tools improve continuously, the tion lines or in-situ operational conditions outside controlled
increasing complexity of modern antennas in terms of their func- measurement ranges. Finally, new wireless technologies call for
tionalities, materials, and structures, as well as always stricter determination of non-traditional antenna performance metrics
performance requirements, mean that experimental measure- – which often require substantial post-processing of the raw
ments remain of utmost importance for development, validation, measurement data.
and calibration of antennas. This Roadmap addresses state-of-the-art antenna measure-
Antenna measurement techniques and technologies face ment techniques and technologies and surveys solutions to the
numerous near-future challenges. Wireless systems from sub- many challenges. These solutions may well depend on the par-
marine communication to deep-space satellite radiometers now ticular sector of wireless systems, but they all rely on progress
span frequencies from less than 1 kHz to above 1 THz. Between in technical-scientific research and engineering across many
Internet-of-Things sensors to high-speed interconnects, band- disciplines; not least computational science where Artificial
width requirements range from tenths to tens of a percentage. Intelligence is currently of enormous interest. To this end, the
Radiation patterns assume almost any shape between 0 dBi Roadmap includes 19 contributions by 34 leading experts ad-
isotropic WiFi nodes and 100 dBi pencil-beam radio telescopes; dressing a multitude of developments in antenna measurement
and adaptive or reconfigurable antenna patterns may assume a theory, in measurement techniques and procedures, in measure-
multitude of different shapes for the same antenna. In addition, ment instrumentation and technology, in error mitigation and
antennas are increasingly integrated with front-end circuitry or uncertainty estimation, in post-processing of measurement data,
entire receiver/transmitter systems as well as embedded with and in measurement standards. In combination, these inspiring
the wireless device; or they are otherwise heavily dependent contributions document that antenna measurements constitute a
on the surrounding environment as for automotive applications vibrant and fast developing technical-scientific field that holds
and medical implants. From hearing aids to communication numerous opportunities for the individual antenna measurement
satellites, the size and weight of the antennas vary by several researcher or engineer.
orders of magnitude. Environmental conditions affecting the Acknowledgement. The Guest Editors want to acknowledge
antenna performance may span wide ranges of pressure and tem- Dr. Pablo Padilla in the editing and formatting of this Roadmap.

1
RoE Journal Antenna Measurement Challenges and Opportunities - 2/49

Novel Near-Field to Far-Field

Transformations
Stuart F. Gregson1 , Thomas F. Eibert2 *,
1 NextPhase Measurements, LA, CA, USA & Queen Mary
University of London, London, UK
2 Technical University of Munich, Munich, Germany

*Corresponding author: stuart.gregson@qmul.ac.uk

Introduction. Electromagnetic near-field (NF) far-field (FF)

transformations (NFFFTs) compute the FF radiation pattern
of a device under test (DUT) from a sufficiently large number
of NF observations, with the DUT being measured either in
transmit or in receive mode due to reciprocity. In particular,
when the influence of a non-trivial measurement probe, i.e.,
not an infinitesimal dipole, shall be corrected, NFFFTs need
to be set up in the form of a linear inverse problem, where the
underlying forward operator gives a relation between a set of
degrees of freedom (DOFs) representing the radiated fields of
the DUT, and the measurement signals as observed at the output
Figure 1: Principle of NFFFTs.
of the measurement probes, see also Fig. 1 for an illustration.
Remarkably, the early NFFFTs were implemented numerically
as extremely efficient, fast Fourier transform (FFT) based ac- FFT acceleration [5], with the associated restriction of regular
celerated direct inversion methods utilizing the orthogonality discretizations. Subsequently, however, fully three-dimensional,
of modal field expansions in cylindrical, spherical, or Carte- i.e., non-canonical, approaches with arbitrary (triangular) dis-
sian (planar) coordinate systems [1, 2, 3]. Such methods are cretizations and full probe correction capabilities have emerged
moreover very robust, they can achieve excellent numerical and become established tools [6, 7, 8]. Here, computational ef-
accuracy, and they allow for accurate measurement probe cor- ficiency is in particular obtained by utilizing the concepts of the
rection, providing the orthogonality of the modal expansion is multilevel fast multipole method (MLFMM) [3, 7]. Such meth-
not destroyed by the probe or the measurement arrangement. ods do not only allow the visualization of the obtained sources,
As such, these NFFFTs require regularly-spaced sample loca- thereby providing a bridge to non-invasive diagnostic methods
tions and uniformly oriented probes, where the probes must also and subsequent radiation studies within various radiation en-
have certain symmetries as in the spherical case. In view of the vironments, but they also enable the suppression of spurious
availability of such extremely powerful NFFFTs, this perhaps in parasitic radiation within the measurement environment.
combination with sophisticated interpolatory schemes [4], was
arguably the state of the art until comparatively recently with Future Challenges and Developments. Standard NF mea-
most NF antenna measurement facilities being designed in order surement approaches, and the corresponding canonical NFFFTs
to comply with the requirements of the NFFFTs. The fact that will continue to provide excellent measurement results in the
modal expansions of a certain order are able to accurately repre- future, but we will also see a large variety of measurement se-
sent the radiation fields of a canonical volume of a certain size tups, not all of which are specifically designed for high-quality
containing the DUT (a cylinder with a given radius and height, antenna measurements. Of particular note are industrial multi-
the intersection of a sphere and a cylinder with given radii, axis robots, which have recently been used for measurements
or a box with given side lengths, corresponding to cylindrical, within arbitrary environments. Future NFFFTs must be able
spherical, or planar measurements, respectively) underpinned to perform well with the measurement data of all these mea-
the reliability and tremendous success of the approach. surement approaches. As such, we need NFFFTS which work
Greater flexibility in terms of DUT representation and ac- well in echoic or anechoic environments, possibly also in the
quisition type is provided by inverse equivalent source methods, extreme NF, and which are able to extract as much informa-
which work with a spatial source representation, e.g., in the tion as possible about the DUT from as little measured data as
form of a discrete set of surface current densities defined on a possible. We need NFFFTs, which can work with the standard
meshed surface surrounding (or sometimes just in front of) the operational signals of radiating devices. However, we also need
DUT. Such methods discretize the forward operator from a dis- NFFFTs which can work without providing a reference for the
crete set of sources to the field observations into a linear system phase or the magnitude or for both, that work with complex
of equations, and then solve the discrete problem in the form waveforms and ideally; we would like to have NFFFTs which
of an appropriate mean square norm minimizing normal equa- do not need any information about the probing antenna, i.e.,
tion, or alternatively as some kind of pseudo inverse solution which determine the DUT and the probe properties simultane-
[3]. Initially, these methods attempted to harness the power of ously. Moreover, we would like to have NFFFTs, which can

2
RoE Journal Antenna Measurement Challenges and Opportunities - 3/49

incorporate as much information as possible about the DUT harder than for the case with phase and magnitude informa-
for the benefit of working well with a reduced number of ob- tion. In principle, magnitude-only NFFFTs are available in the
servation samples, whilst still being sufficiently sensitive to be form of optimization methods, where the major problem is not
able to reliably detect deficiencies within the DUT itself. If the NFFFT, but rather the question of how to collect enough
we think about such developments, we must, however, always information for the unique solution of the optimization prob-
keep in mind that the measurement community is spoiled by lem [15]. Therefore, these problems are perhaps more closely
the tremendous computational power of the existing canonical connected with the measurement approaches than the NFFFTs,
NFFFTs. Therefore, attractive computational efficiency and with non-linear optimization based RF measurement techniques
sensitivity is also a key requirement for any new NFFFT. also proving fruitful [16].
(a) NFFFTs and Computational Electromagnetics. Enor- (e) Probe Correction and Error Correction Methods. Probe
mous potential for the development of future NFFFTs will correction is a most important pre-requisite for accurate NFFFTs.
evolve from their unification with computational electromagnet- In general, the FF results can only be as good as the probe cor-
ics (CE). Many of the currently available NFFFTs have already rection, where the influence of the probe is certainly different
been inspired by CE methods, e.g., by fast integral methods for different measurement configurations and, e.g., is also de-
such as the MLFMM. In future, we will see sophisticated mod- pendent on the measurement distance. An attractive NFFFT
els of the DUT or the measurement environment integrated into would be an NFFFT, which does not need any a-priori probe in-
NFFFTs, as, e.g., seen in [9], or we will model the measurement formation at all, i.e., one which is able to retrieve the necessary
setup in the form of a digital twin in order to better understand probe information from the observation data itself [17]. Such
and analyze the error behaviour of the measurement and field methods may be seen in a similar light as phase-less methods.
transformation process [3]. Here, the necessary information for solving the NFFFT problem
(b) Echo Resilient Methods. A key capability for modern is not automatically available within a standard measurement,
NF measurements situated within arbitrary environments is the and we have to bring more information into the NFFFT process.
availability of methods for the suppression of parasitic echoes Once sufficient information for the solution of the problem is
[3]. Time gating and other measurement hardware supported available, similar algorithms as in the case of the phase-less
methods will continue to play an important role. However, methods can probably retrieve the desired information from
we also need improved methods which are integrated into the the measured data. Such thoughts can also be carried over to
NFFFT itself, as already started in [10]. We need accurate methods for the intrinsic correction of errors, e.g., probe posi-
spatial localization, combined with spectral localization, and tioning or orientation errors. With a sufficiently large amount of
also combined with temporal localization. Furthermore, the observation data available, via high-speed acquisition systems,
concepts of virtual probe arrays and of highly-oversampled this now appears more plausible than ever before.
measurements will of course further extend these aims. (f) Artificial Intelligence and Machine Learning. Artificial
(c) Reduced Sampling Methods. Reduced sampling methods Intelligence (AI) and Machine Learning (ML) methods will
have been attracting notable attention for several years now have great potential for further improving NFFFTs. They can
with sparse compressed sensing methods in particular yielding help us in the identification and subsequent extraction of un-
many new ideas [11, 12]. Equivalent source based NFFFTs can known echoes, in the appropriate placement of measurement
easily process the sparse observation data, however, improved samples in reduced sampling methods, and of course also in the
concepts concerning the placement of the sources and methods placement of the equivalent sources, as well as for the purpose
for appropriately including the available information about the of efficient and accurate phase recovery. In principle, AI can
DUT into the NFFFT will further enhance these methods to learn AUT types, probe representations, and echo signal repre-
achieve new levels of performance [13]. Here, not only statis- sentations during the course of many measurements, and it will
tical approaches such as those based on sparsity assumptions inevitably help us in the interpretation of both measurement,
may be of use, but also methods, which utilize the peculiarities and transformation results.
of the underlying radiation operator itself, may prove beneficial (g) Improved Computational Efficiency & Sensitivity. Excel-
[14]. lent numerical efficiencies are a pre-requisite for the acceptance
(d) Phase-less Methods. Interest in phase-less near-field mea- and adoption of any new NFFFTs. Therefore, a significant
surements dates back many decades and stems from noting that amount of effort and ingenuity will need to be invested in the
the availability of phase-less NFFFTs would be very attractive continuous improvement of computation speeds of NFFFTs.
for the simplification of many measurement approaches. The Improved equivalent source representations can, e.g., lead to im-
rapid increase in the interest of using un-tethered, uninhabited proved efficiencies [18], or we may see novel preconditioning
air vehicles (UAVs), i.e., drones, higher frequency applications techniques [19], where even multiple frequency solutions may
employing industrial multi-axis robots, and 5G/6G communi- benefit from each other. The aforementioned reduced sampling
cation system testing utilizing complex waveforms have only methods may lead to further speed-ups, or we may pre-compute
served to further increase the need for phase-recovery. How- parts of a particularly complicated solution algorithm in a smart
ever, really reliable, truly general purpose approaches which way. Finally, although many existing transformation algorithms
work completely in the absence of phase information are prob- are implemented utilizing parallel processing techniques, there
ably hard to achieve for microwave frequencies and below. In is still great scope for further enhancement with these algo-
particular, it is often forgotten that the accuracy requirements rithms being ported to highly-parallel computation platforms
for magnitude only measurements are commonly considerably such as graphics processing units (GPUs), etc.

3
RoE Journal Antenna Measurement Challenges and Opportunities - 4/49

Reduction of Near-Field for several different geometrical models of the source and for
the three canonical scanning geometries [20, 21]. An approach

Antenna Measurement Time relying on the optimization of the singular value behavior of
the discretized radiation operator (Singular Value Optimization
- SVO) has been introduced, applied to the case of aperture
through Non-Regular antennas and to the three canonical scanning geometries again
[22, 23]. A thinned equiangular or igloo sampling scheme has
Scanning been also introduced in [24, 25]. Compressed Sensing (CS) has
also found application for spherical near-field measurements
Amedeo Capozzoli1 *, Francesco D’Agostino2 to reduce the number of sampling points thanks to a sparse
1 Dipartimento di Ingegneria Elettrica e delle Tecnologie
representation of the measured field [11, 12, 26]. For cases
dell’Informazione, Università di Napoli Federico II, Napoli, when the phase is missing, the sampling problem is even more
Italy critical since the lack of phase information must be compensated
2 Dipartimento di Ingegneria Industriale, Università di Salerno,
making, as a result, the number of acquisitions larger than that
Fisciano, Italy needed for the complex case, significantly affecting the scanning
*Corresponding author: a.capozzoli@unina.it
time [27].

Introduction. Antenna measurements from near-field data Emerging challenges. Notwithstanding the significant ad-
acquisitions represent, nowadays, an established methodology vancements and results achieved in the lastest decades on meth-
to collect the field information about an Antenna Under Test ods to reduce the measurement time which allowed to introduce
(AUT) to characterize its radiative behaviour as well as to diag- non-regular sampling techniques that can be now assumed es-
nose its working conditions. Typically, the Near-Field samples sentially established, different challenges remain to be tackled:
are collected on a surface with a conventional geometry (planar,
cylindrical, or spherical), limiting the acquisitions to a portion 3. Further improving previous points #1 and #2.
thereof in the first two cases, as well as in the case of spherical 4. Definition of unique sampling grids for carrying out the
systems, when allowed by the radiative features of the AUT. characterization/diagnostics at multiple frequencies.
Standard guidelines have been formulated, leading to conven-
tional sampling strategies over canonically defined acquisition 5. Definition of the minimum region (size and shape) to be
regions. However, a sub-optimal extent of the sampling region sampled.
and the un-necessarily large number of sampling locations can
make the acquisition time uselessly dramatically high. The guid- 6. Use of strategies and hardware for the movement (multi-
ing light towards a reduction of the measurement time is the use axis/robots) and control of the probe in connection to the
of non-regular sampling strategies, aiming at gathering just the definition of optimal scanning paths/motion laws account-
indispensable amount of information to the targeted purpose. ing for the features of the available equipment.
Up to now, the following key points returning significant results 7. Development of measurement strategies in the very near-
have been faced: field of the radiator.
1. Definition of appropriate representations of the source 8. Development of techniques with reduced resolution or
accommodating all the available a priori information. multiple resolution.
2. Determination of the number of samples in the measure- 9. Development of techniques for partial characterizations.
ment region, and their spatial distribution which allows
gathering the information needed by the antenna charac- Future developments to satisfy these challenges. The re-
terization/diagnostics. search activity of the community in very recent years is attempt-
ing to give answers to the issues raised above. Developments
Concerning the above points, which are strictly inter-related, are expected in the next future to delineate even more effective
to reduce the scanning time for the characterization/diagnostics solutions. In particular, concerning point #3, further refinements
by properly defining the measurement locations, it becomes on the source representation are being introduced to improve
crucial to accommodate all the available a priori information the performance of the sampling techniques, possibly exploiting
on the radiator. In particular, in characterization problems, the a unique “optimal” sampling lattice for all the frequencies (see
exploited information should rely on mild assumptions typically point #4) involved by the characterization/diagnostics. This
given on the geometry of the source. For diagnostics purposes, becomes critical for phaseless acquisitions. Regarding point #5,
the a priori assumptions should not involve only the information the possibility of defining the optimal region to be scanned with-
of real interest enabling the identification of the faults. Much out impairing the results according to a prefixed tolerance (trun-
work has appeared on the definition of the sampling number cation error) is even more crucial, particularly when referring to
and locations leading to non-regular sampling strategies with scanning surfaces with unconventional shapes. Furthermore, the
diverse performance. Different solutions have been proposed use of unconventional shapes, made possible today by the use of
with different rationales. A non-redundant sampling exploiting multi-axis robot arms moving the probe, opens new avenues in
the concept of local bandwidth of the field has been developed near-field sampling. Indeed (see point #6), probes installed on

4
RoE Journal Antenna Measurement Challenges and Opportunities - 5/49

one or more robot arms enabling more convenient scanning ar-

eas thanks to the offered degrees of freedom in positioning and
orientation of the sampling grids are giving a new perspective
to near-field acquisitions [28, 29]. Also optimized controllers
are being devised, profiting of the non-regular sample positions
to access a significant increase in the average probe speed be-
tween two consecutive points even in continuous acquisition
schemes [30]. As far as point #7 is considered, the acquisition
in the very near field, with a non-significant perturbation of
the working conditions of the AUT, could permit a significant
reduction of the size of the scanned area with beneficial effects
on the overall acquisition time. Obviously, the sampling in the
very near field represents an open problem, which should take
into account for contribution from both visible and invisible
domains, but should profit as much as possible of non-regularity
to reduce the number of needed sampling locations. At the
same time, non-invasive probes are required in very near field
acquisitions, with a reduced mutual coupling with the AUT.
Concerning point #8, typically, traditional systems attempt to
retrieve the radiation behaviour from near-field samples with
full resolution. In some cases, a fast snapshot can be useful,
particularly at preliminary characterization steps. In some cases,
a multi-resolution characterization could be useful, returning
full-resolution only wherever needed, but rougher details where
sufficient. Obviously, not requiring the full resolution could
reduce the amount of measured data and speed up the acquisi-
tions. Moreover, in other cases (see point #9), full-resolution
characterizations, but at prescribed cuts of the radiated field,
can be very appealing, particularly when this comes with a dra-
matic reduction of the needed field samples [31, 32]. Finally,
parallel probes, already exploited by standard, regular sampling
techniques, could introduce a further improvement in speeding
up non-regular acquisitions.
Conclusion. Field sampling in the near-field exploiting non-
regular distributions of the samples has paved the way to several
significant improvements in antenna characterization/diagnostics
with amazing outcomes. Significant reductions of the overall
acquisition time have been obtained, with strong relapse in the
applications, but solutions to some relevant open problems are
expected to provide a further practical impact.
Acknowledgment. The Authors would like to thank Prof.
Claudio Curcio and Prof. Angelo Liseno for the discussions
and the contribution in preparing the manuscript.

5
RoE Journal Antenna Measurement Challenges and Opportunities - 6/49

Post Processing for Antenna nals in the radiated field [37], antenna placement investigations
[38], and performance analyses of 5G devices [39].

Diagnostics and Spurious niqueDifferent approaches of using the equivalent current tech-
for solving the inverse problem of finding the radiating
currents from measured fields have been developed in the past
Signals Cancellation two decades. The first type of methods are based on modal trans-
formations, for example plane wave to plane wave (PW-PW)
Andreas Ericsson 1 *, Giuseppe Vecchi 2 transformations (microwave holography) or spherical wave to
1 TICRA, Copenhagen, Denmark
plane wave (SW-PW) transformations [40]. These methods are
2 Department of Electronics and Telecommunications, Po- fast and well suited for electrically large problems, but only
litecnico di Torino, Italy reconstruct the currents on a planar surface. While PW-PW
*Corresponding author: ae@ticra.com transformations provide a spatial resolution limited to half a
wavelength, SW-PW transformations can provide higher spatial
This contribution provides a brief overview of recent devel- resolution than half wavelength with noise free measurement
opments related to post processing of measured antenna ra- data, but in practice this is difficult to be achieved with a typical
diation data for diagnostics applications and cancellation of 60 dB signal to noise ratio (S/N).
spurious signals. Traditional post processing methods such as The second type of methods is based on representing the
back-projection have been known and used for decades, but inverse problem as integral equations that are solved by some
limiting factors such as low resolution makes these methods kind of Method of Moments (MoM) based implementation
inadequate for some applications. Here we review a Method of [35, 41, 42, 43] These methods can operate on measured data
Moments (MoM) based method that can instead be applied that sampled on regular, irregular as well as truncated surfaces, and
offers superior resolution and 3D reconstruction, but is more reconstruct equivalent currents on 3D surfaces of general shapes.
computationally demanding. New accelerated MoM current The spatial resolution of the reconstructed currents is in general
reconstruction methods that are presented in this work open up high and may be better than half a wavelength, even in the pres-
the possibility for performing antenna diagnostics of electrically ence of noise, indicating that the 3D reconstruction methods are
large and complicated antennas, reducing the prototype cycles superior to traditional microwave holography, especially for ar-
and increasing the quality of the antenna prototype. ray antennas [44]. Moreover, 3D reconstruction allows filtering
of undesired radiation and scattering, which is not possible with
Introduction. The accuracy of antenna measurement test ranges microwave holography. Two examples of commercial software
has improved significantly in the past decades. In addition to products based on this type of current reconstruction methods
many hardware improvements, also advances in the computa- are DIATOOL1 by TICRA and Insight2 by MVG. Drawbacks
tional methods used to process the acquired measured data have with the 3D reconstruction methods are that their computational
greatly enhanced performances. Despite all the mentioned ad- requirements are generally high, and their baseline versions
vances, it may be the case that the measured radiation pattern of scale poorly with frequency and the electrical size of the prob-
an antenna under test (AUT) differs from the designed/expected lem. However, a number of new developments have been made
results, and that the cause of this discrepancy can not be easily in recent years to meet the challenge of applying this approach
identified. In the past, a trial and error procedure has commonly to large antennas; in this work, alternative methods are presented
been utilized. Today, larger and more complicated antenna de- for accelerating the computational solutions to the equivalent
signs are being utilized, such as passive/active array antennas, current reconstruction problem.
large deployable antennas and complex science instruments;
antennas are often mounted on platforms such as cars, satellites Theory. In order to solve the inverse source reconstruction
and ships, where the antenna surrounding has a non-negligible problem numerically, using measured field in amplitude and
effect on the antenna performance. In these cases, the trial and phase as input, the reconstruction surface and the unknown
error antenna prototyping approach can be very costly and time currents are discretized as a linear system of equations
consuming.
Post processing techniques of the measured antenna radia- Ax = b (1)
tion are indispensable tools for digging deeper and finding the
source of discrepancies in the radiated field, and thus shorten where A is a matrix representing the radiation from the unknown
the prototyping cycle of antenna designs. A specific post pro- currents x on the reconstruction surface S that generate the
cessing technique that has been very successful for this purpose measured fields in b. For applications with diagnostics purposes,
is the equivalent current reconstruction technique (also referred the source reconstruction equations should be augmented with
to as source reconstruction). It consists of computing equivalent Love’s condition of zero fields inside a surface enclosing the
currents with a known location that radiate a given complex sources. This condition ensures that the found currents provide
vector field [8]. The equivalent currents are typically computed a unique solution that represents the actual tangent fields (i.e.
at a surface in front of, or enclosing, the AUT. Once these equiv- physical currents) on the structure. The mathematical problem
alent currents have been found they can be used for a number of 1 DIATOOL Software, website: https://www. ticra.com/software/diatool/.
applications such as antenna diagnostics [33, 34, 35], near-field 2 Insight Software, website: https://www. mvg-
to far-field transformation (e.g. [36]) filtering of spurious sig- world.com/en/products/antenna-measurement/ software/insight.

6
RoE Journal Antenna Measurement Challenges and Opportunities - 7/49

to solve can thus be formulated as

min ∥Ax − b∥2 (2)

x

s.t. Lx = 0 (3)

where L is the matrix representation of Love’s condition. The

problem in (2)–(3) is solved iteratively to find the currents x.

Recent numerical improvements for source reconstruction.

(a) Calderon projections for fast antenna diagnostics: An
alternative solution procedure to the problem in (2)–(3) was
initially indicated in [8], and recently pursued in [45]. Instead of
solving the data equation in (2) by including (3) in each iteration,
the unique current condition enforced by (3) is represented by Figure 3: Diagnostics of a Planar Wave Generator (PWG) large
a projection operator T that is applied a single time, after the array: the AUT in the measurement setup and Ring n.2
solution to (2) has been found. We refer to this final step as a excitation verification.
Calderón mapping3 ; a possible way to effect that is to restate
(2)–(3) simply as a preconditioned Least Squares problem
(b) Fast multilevel Low-Rank solver: The original numeri-
min ∥AT z − b∥2 , x = Tz (4) cal problem (2)–(3) closely resembles the MoM problem for a
z
penetrable body; hence, as mentioned above, it can be acceler-
An implementation tailored for electrically large problems ated by fast factorization schemes with O(N log N) complexity
has been introduced in [46] and applied to large reflector anten- per iteration. The main, relevant difference to standard MoM
nas for space applications in [47]. This implementation leads problems is that the essential testing condition happens on the
to a complexity scaling at most as O(N log N) per iteration, measurement surface, significantly farther away from the re-
where N is the number of unknowns, with drastic improve- construction surface; on it testing points due to measurements
ment with respect to standard approaches. Furthermore, the are typically much more spaced that in MoM problems (close
method employs higher order basis functions on higher order to λ /2 in the ideal sampling case): this renders the standard
quadrilateral mesh elements, with a significant reduction in the application of fast factorizations less efficient. This can be obvi-
number of unknowns required for a given problem with respect ated by an ad-hoc algebraic factorization [48, 49, 50] that has
to Rao-Wilton-Glisson (RWG) basis functions. been tested on a large and complex structure as in Fig. 3; the
An example application of the Calderón method is presented improvement is above a factor of 35 in memory and time with
in Fig. 2, where equivalent electric and magnetic currents have respect to standard implementation [51].
been computed on a conformal reconstruction surface enclosing Conclusion. We have reviewed baseline and advanced equiv-
a satellite with three reflector antennas, where one antenna is alent currents methods, that have witnessed a constant improve-
radiating at 10 GHz. This equivalent current reconstruction ment in application, theory and implementation - as well as new
problem consists of roughly 3.2 million higher-order unknowns challenges.
(corresponding to about 15 million RWG unknowns) and only
require a RAM allocation of 60 GB. Acknowledgment. The Authors would like to acknowledge
collaborators Oscar Borries, Erik Jørgensen, Martin Haulund
Gæde and Peter Meincke at TICRA; and Giorgio Giordanengo
and Marco Righero at Links Foundation, Turin, Italy.

Figure 2: Reflector antennas on a satellite platform enclosed by

a conformal reconstruction surface (left) and reconstructed
equivalent electric current density (right).

3 In [8] this approach was not indicated by this name, but as ”field boundary

integral identities”

7
RoE Journal Antenna Measurement Challenges and Opportunities - 8/49

Gain Budget Error Term Uncertainty (dB)

Antenna Measurement 1. Probe Relative Pattern 0.00
Uncertainty 2. Probe Polarisation Radio
3. Calibrated Probe Gain
0.00
0.15
4. Probe Alignment 0.00
Anna Granich1 *, Tian Hong Loh2
1 Institute 5. Normalisation Constant 0.16
of High Frequency Technology, RWTH Aachen
6. AUT Impedance Mismatch 0.06
University, Aachen, Germany
2 Electromagnetic 7. AUT Alignment N/A
& Electrochemical Technologies Depart-
ment, National Physical Laboratory, Teddington, United
8. Data Point Spacing 0.02
Kingdom 9. Data Truncation N/A
*Corresponding author: granich@ihf.rwth-aachen.de 10. Sphere Radius Errors 0.00
11. Sphere Theta/Phi Errors 0.01
12. Higher Order Coupling 0.03
Introduction. Antennas are an inherent part of any wireless
13. Receiver Amplitude Non-Linearity 0.00
transmission. Especially their efficient use of energy and ul-
timate utilization of the electromagnetic spectrum are essen- 14. System Phase Errors 0.03
tial requirements for various applications such as remote sens- 15. Receiver Dynamic Range 0.00
ing, radar, next generation wireless systems with high-speed 16. Room Scattering 0.01
and high link reliability. A raft of emerging wireless tech- 17. Cable Leakage 0.00
nologies such as ultra-massive multiple-input-multiple-output 18. Repeatability and Random Errors 0.00
(UM-MIMO), Orbital Angular Momentum (OAM), etc. is in- Gain Total Uncertainty (RSS) 0.23
creasingly arising in the arena of modern devices and system
design [52]. Furthermore, technology advancements together Table 1: Gain uncertainty budget example [54].
with spectrum scarcity has driven the exploration of new radio
frequency (RF) spectrum suitable for upcoming wireless tech-
nologies in the millimetre wave (mm-wave), terahertz (THz) efforts redesigning without insight of the actual cause, which
and even optical frequency bands. can limit the uptake of these technologies. Antenna radiation
efficiency is an important attribute of antennas as it has a sig-
All types of omni-directional and directional antenna mea-
nificant effect on the performance, reliability, and efficiency of
surements have their associated uncertainties, which are un-
wireless communications systems. The recently revised IEEE
avoidable due to limitations of the measurement instrument
Standard 149 [55] include the measurement techniques such
capabilities. Their measurement results can only be relied upon
as pattern integration, Wheeler cap and reverberation chamber
with confidence when all the uncertainties associated with them
methods. In the following, we discuss the current measurement
are fully understood and known. Therefore, it is important to
challenges and future developments required for improving
evaluate their measurement uncertainty to quantify the reliabil-
antenna measurement uncertainty.
ity of the measurement result. The traditional 18-term sources
of uncertainty errors that have a direct impact on the accuracy of Current and Future Challenges. With the industrial exploita-
the far-field measurement of an antenna under test (AUT) mea- tion and adoption of complex new radio (NR) signals, energy
sured in the near-field has been given in [53]. This uncertainty efficient devices and large-scale multi-antenna beamforming
analysis can be extended to far-field (see example in Table 1 technologies at different RF bands in emerging wireless sys-
[54]) and compact range measurements (see example in Chapter tems, several worldwide industries, research communities and
13 in IEEE Standard 149 [55]). standard bodies are now facing new measurement challenges
In practice, the antenna calibration process involves the on efficient and accurate verification of NR products that meet
measurement of the ratios of powers and as such a power ra- desired performance parameters for fulfilling the diverse techni-
tio measurement system is required with sufficient dynamic cal requirements set by ITU-R [58], especially, for high-volume
range and suitable linearity to achieve the required uncertainties. beam-reconfigurable issues [59].
For determining the gain of an antenna, the most accurate of Typical antenna measurands are antenna factor, gain, axial
the many measurement techniques is the three-antenna extrap- ratio, efficiency, radiation pattern, etc., but as wireless tech-
olation technique [56]. By using three antennas to make the nologies evolve and modern antennas are becoming highly
measurement, one does not require an a priori knowledge of the integrated into wireless systems, there is a shift towards an
gain of any of the antennas used. When measuring the antenna over-the-air (OTA) radiated testing approach (due to the lack of
radiation pattern, the antenna-to-range interface (including an- antenna connectors) measuring some other integrated power pat-
tenna feed, positioner, etc.) can have a large effect, for example tern metrics such as total radiated power (TRP), total isotropic
for omni-directional antennas without balun or adequate match, sensitivity (TIS), effective isotropic radiated power (EIRP) and
as they are prone to unwanted common mode currents on the effective isotropic sensitivity (EIS). The 3rd generation partner-
coaxial cable used to feed the antenna during the measurement. ship project (3GPP) defines three OTA test methods, namely, the
They radiate and interfere with the radiation of the antenna, Direct Far-Field (DFF), Indirect Far-Field (IFF) and Near-Field
which can cause differences between the measured pattern and to Far-Field Transform (NFTF) in TR 38.810 [60]. However,
the expected pattern [57]. The designer can spend fruitless other potential candidates, such as mid-field [61] and reverbera-

8
RoE Journal Antenna Measurement Challenges and Opportunities - 9/49

tion chamber methods [62] are missing. Moreover, measuring

the performance of the devices or systems with adaptive anten-
nas and characterising the propagation channel creates further
challenges at mm-wave bands, where innovative test methods
and novel measurement equipment are required throughout re-
search and development stages.
While on the one hand the accuracy of the measurement
directly depends on the named 18 error term models [53], on
the other hand the properties of the measurement laboratory
itself as well as the measurement method play an important role.
Figure 4: DTU-ESA 12 GHz VAST (left) and DTU-ESA
Therefore, the error description of an antenna in a particular
mm-wave VAST (right) antennas.
measurement laboratory remains incomplete as long as it has
not been validated against other measurements performed in
other measurement facilities [63]. Validation and accreditation measurements, due to their portability and the need to measure
of different facilities can be achieved with inter-comparisons electrically large antennas in situ. Despite the higher degrees
and careful evaluation of error budgets and different measure- of freedom UASs can offer, it is important to note that addi-
ment standards. A crucial factor here is a robust reference tional uncertainty may incur due to the precise positioning in
antenna that consistently performs under different measurement environmental conditions like wind and gust [69].
environment conditions. An example of the requirements and
design for such robust reference antenna is given in [64] for the Future developments to satisfy these challenges. With
DTU-ESA mm-wave Validation Standard (VAST) antenna. the use of new ranges like the robotic antenna test range the
To fulfil the needs of new technologies, new measurement uncertainty budget is envisaged to change. For every new test
campaigns are constantly being launched. For example, there range and testing method the well-known uncertainty terms
are several on-going inter-comparison measurement campaign [53] have to be revisited, revised and maybe supplemented
activities carried out under the measurement working group of or replaced. The same applies to new testing methods like
the European Association on Antennas and Propagation (Eu- OTA and the measurements of active antenna systems where
rAAP), IEEE technical committee on antenna measurements, other parameters (e.g., TRP, EIRP, etc.) are measured or the
etc. focusing on antennas with different characteristics like uncertainties of the antenna systems itself have to be considered.
both narrow and wide-band antennas within different frequency New methods like the test-zone field compensation tech-
ranges, omni-directional and highly directional antennas. Also, nique [70] can be used and established to improve measure-
an international comparison for antenna gain measurement of ment accuracy and enable precise antenna measurements even
two Ku-band standard gain horn (SGH) antennas that involved with non-ideal reflectivity of the measurement environments.
12 national measurement institutes (NMIs) has been carried Whereas the evaluation of a complete uncertainty budget us-
out [65] whereby a variety of measurement techniques such ing measurements can be extremely time consuming, with in-
as extrapolation, far field, gain transfer and reciprocity were creasingly powerful computation and enhanced simulation algo-
used. A similar international comparison for antenna pattern rithms, simulation tools can be used to support and accelerate
measurement that involved eight universities and industry was measurements. This would also allow uncertainty budgets and
also performed in 2004 and 2005 with the DTU-ESA 12 GHz measurement environment properties to be determined more
VAST antenna at 12 GHz [66] using different measurement quickly and accurately in the future.
techniques and antenna facilities, including spherical near-field, Acknowledgment. The work of T. H. Loh was supported by
compact range, planar near-field and far-field facilities. In 2019- the 2021-2025 National Measurement System Programme of
2022 a measurement campaign with DTU-ESA mmVAST for the UK government’s Department for Science, Innovation and
19.76 GHz, 37.80 GHz and 48.16 GHz was conducted with Technology, under Science Theme Reference EMT23 of that
13 participating facilities [67]. Pictures of the two DTU-ESA Programme. The authors want to thank Prof. Dirk Heberling
devices are given in Fig. 4. More details on facility compar- for the technical discussion and support.
isons are also given in the Section entitled “EurAAP and IEEE
Standardization and Facility Comparison”.
New ranges with higher flexibilities, like robotic antenna
test ranges are of upcoming importance within the past few
years. Due to the higher mechanical flexibility, the measure-
ment uncertainties differ from the known uncertainty terms
from well-established ranges [68]. In addition, non-classical or
irregular measurement geometries outside of spheres, cylinders
and planes can be applied, which also require further consid-
eration. On the other hand, such higher flexibility also allows
for more degrees of freedom in alignment correction. Besides
the robotic test ranges the use of Unmanned Aerial Systems
(UASs), so called drones, is an emerging technology in antenna

9
RoE Journal Antenna Measurement Challenges and Opportunities - 10/49

Large Satellite Antenna

Measurements at Airbus
Christian Hunscher*
Airbus Defence & Space GmbH, Taufkirchen, Germany
*Corresponding author: christian.hunscher@airbus.com

Large deployable satellite antennas and respective structures

can hardly be characterized w.r.t. their RF performance using
state-of-the-art antenna measurement systems such as compact
ranges or cylindrical / vertical planar near-field scanners, which
are usually implemented in the integration and test areas of the
spacecraft manufacturers. This is mainly due to the size of the
reflecting surfaces, but also by its gravity sensitive elements like
meshes or deployment booms, which require complicate gravity Figure 5: PAMS Gondola with calibrated test probe and
compensation devices to bring them into their representative removed bottom cover during Ka-Band Test.
in-orbit configuration during on-ground testing. Consequently
in some cases those antennas have even not been fully RF-
measured before being placed into orbit. To overcome this limi- Following the successful system qualification itself the PAMS
tation Airbus Defence and Space has developed the so-called has been applied in different RF-test campaigns for the perfor-
PAMS (Portable Antenna Measurement System), which enables mance verification of large deployable mesh or panel based
pattern and gain measurements of large antennas without the reflector antennas. A typical example is shown in Fig. 6.
need of moving the test object by means of precise positioners
or scanning systems. The core idea is to utilize an especially
modified overhead crane as a coarse near-field scanner driving
a gondola with an integrated RF probe at the crane hook. The
gondola and its probe are commanded by a controller to sample
the near-field data along a roughly pre-defined trajectory above
or around the device under test. The versatility in the scan
surface allows for measurements of a motion free test antenna.
The sampling in irregular intervals requests for an innovative,
advanced near-field to far-field transformation algorithm, which
incorporates precise knowledge of the probe orientation and
position into a set of equations that is solved for the unknown
plane wave modes. This algorithm was developed at the TU
München. The exact probe location information is obtained
in up to six dimensions from a laser tracking system, which is
following a target mounted on the lower part of the gondola. Figure 6: PAMS test configuration for a large antenna structure
The sample density needs typically to be in the order of < 0.4λ ; (5m Deployable Reflector developed by Airbus Defence &
the scan speed is about 200 mm per second. Space GmbH, Friedrichshafen).
The novel RF-test system has been qualified in several fre-
quency bands between L- and Ka-Band. Therefore well-known The comparison of the measured RF-pattern with predic-
reference antennas have been measured with the PAMS as well tions based on well-known antenna analysis software routines
as in the classical Compensated Compact Range at Airbus De- such as GRASP developed by TICRA, Denmark, has revealed
fence & Space GmbH, Taufkirchen. As an example the set-up a very good agreement [72].
of a PAMS-based test campaign in Ka-Band is shown in Fig. 5.
The test object is mounted on a fixture in front of an absorber
wall being coarsely levelled horizontally. Exact levelling or
placement of the test object is not mandatory as the antenna co-
ordinate system is aligned with optical targets. The laser tracker
needs to be placed such that the respective target is always
visible during the nearfield scanning. The data resulting from
the nearfield to farfield transformation have demonstrated an
excellent level of agreement with those achieved in the classical
compact range set-up. Typically the error contribution was in
the order of -40 dB [71].

10
RoE Journal Antenna Measurement Challenges and Opportunities - 11/49

Large Satellite Antenna

Measurements at JPL
Jefferson A. Harrell, Paolo Focardi*
Jet Propulsion Laboratory, California Institute of Technol-
ogy, Pasadena CA, USA
*Corresponding author: paolo.focardi@jpl.nasa.gov

Introduction. The history of the Mesa Antenna Measurement

Facility at the Jet Propulsion Laboratory (JPL) goes back to
the early 1960s. The location was chosen due to the topology
being a mesa overlooking the JPL campus, which provides both
multi-path advantages for outdoor antenna testing and a quieter
RF environment. At the time, all measurements were Far Field
(FF), which provided relatively quick feedback of the pattern Figure 7: The CloudSat W-band offset reflector being
cuts. With lesser computational power than what is available measured in the 12.2 m (40 ft) chamber at JPL.
today, some parameters, such as the feed position on a reflector
antenna, were empirically optimized by real time adjustments.
Some of the antennas operated above 100 GHz, but most were time higher frequencies. While our facilities offer limited ac-
at or below Ka-band. Over time, as Near Field (NF) processing commodation in terms of antenna size, we are working toward
developed and NF range hardware matured, JPL’s antennas upgrading our NF scanners with the latest instrument packages
started being tested in indoor NF chambers as opposed to the to be able to handle antennas at W-band and higher. Moving to
FF ranges outdoor. The first NF chamber on the JPL Mesa was higher frequencies has the added benefit of making our cham-
a custom-conversion from an indoor 12.2 m FF range, which bers inherently larger and therefore we can accommodate larger
ultimately became a cylindrical NF chamber, optimized for apertures in terms of wavelengths.
the long and narrow antenna elements used for NSCAT, after
being used as a plane-polar NF scanner for the 4.8m Galileo
antenna. This chamber has served faithfully through many
projects, including CloudSat [73] (see Fig. 7), which operated
at W-band. Note that this NF chamber, and the one cited below,
support conventional NF data acquisition and processing. This
NF range was also used to measure Ku-band 1:10 scale models
of the AQUARIUS and SMAP [74, 75] (see Fig. 8 and 9)
instrument antennas with excellent results when compared with
predictions calculated with TICRA GRASP and Ansys HFSS.

Emerging challenges. As the demand for NF measurements

increased, an adjacent 18.3 m indoor FF chamber was chosen
to be converted into a NF chamber via the acquisition of an
NSI, Inc. 9.1 x 4.6 meters planar scanner. In 2017, this was up-
graded to include spherical NF capacity. Most of JPL’s projects
require the planar scan capability, but the spherical capability
has proven to be useful for lower gain antennas and feeds. The
Figure 8: AQUARIUS 1:10 scale model being measured in the
current frequency range of the planar scanner is 1 to 40 GHz.
However, by using custom electronics, we have obtained rea- 12.2 m (40 ft) chamber at JPL.
sonable co-pol only measurement results at 130 GHz and 168
GHz. With future JPL mission concepts planned for W-band We recently tested a prototype antenna from Tendeg that
and above, the Mesa is currently planning to acquire upgrades approached the limits of our 18.3 m chamber capabilities. The
for up to 170 GHz, and possibly higher. The necessary equip- deployable mesh reflector, which is parabolic on one axis and
ment includes RF electronics, probe fixtures, and standard gain flat on the other (Fig. 10) offers some interesting RF perfor-
horns for each band. The equipment is available from the range mance with its 7.1 m length. The unit fits inside our 6.1 x 18.3 m
manufacturer and is relatively easy to install. chamber even though it required a somewhat complex Ground
Support Equipment (GSE) structure and about two wavelengths
Future developments to satisfy these challenges. In the of spacing for the probe aperture to clear the center mounted
pursuit of always better RF performance, sensitivity, and spa- feed. The pattern, measured on the planar NF scanner at 5 GHz,
tial resolutions, current and future projects at JPL are pushing showed repeatability below -40 dB.
the antenna technology toward larger antennas and at the same On the opposite side of the spectrum, plans are underway to

11
RoE Journal Antenna Measurement Challenges and Opportunities - 12/49

Figure 10: A 7.1m long Tendeg deployable mesh reflector

being tested in the 18.3 m (60 ft) chamber at JPL.

der a contract with the National Aeronautics and Space Admin-

istration. Copyright 2023. California Institute of Technology.
Government sponsorship acknowledged.

Figure 9: SMAP 1:10 scale model being measured in the 12.2

m (40 ft) chamber at JPL.

replace the absorber in the 12.2 m chamber with material suit-

able for UHF frequencies, and to replace the aging cylindrical
range with a planar/spherical range. Several future JPL projects
are aiming to use UHF for telecom and instrument applica-
tions. The upgrade would enable the measurements currently
performed outdoors to be done indoors.

Conclusion. Adding test capabilities does not come with-

out challenges as it is always the case with indoor ranges. As
mentioned above, we are planning to upgrade our existing pla-
nar/spherical range electronics to be able to test at higher fre-
quencies, from 90 GHz to 170 GHz. While ad-hoc set-ups have
been done in the past, we need a more standard and reliable
approach for these frequencies. In our other chamber, we are
planning to add a new planar/spherical range with large absorber
to test antennas at UHF frequencies, which until now have al-
ways relied on outdoor ranges. These upgrades will enable our
Mesa Antenna Test Facility to continue to play its vital role
of ensuring mission success for JPL telecom and instrument
antennas.

Acknowledgment. The research was carried out at the Jet

Propulsion Laboratory, California Institute of Technology, un-

12
RoE Journal Antenna Measurement Challenges and Opportunities - 13/49

Robotic
Antenna Measurements
Joshua Gordon*
Communications Technology Laboratory,
National Institute of Standards and Technology,
Boulder, CO, USA
*Corresponding author: josh.gordon@nist.gov

Outline. Robots excel at manipulating objects with 6 degrees

of freedom movement. A combination of well-defined kinemat-
ics, optimized control loops, and robust mechanical architecture
make robots a versatile and capable platform for a range of
applications where controlled movement throughout space is
needed. The requirement to manipulate antennas of various
form factors accurately and precisely over a volume of space
is especially important for antenna testing. The use of robots
to address measurement challenges in today’s wireless world
has continued over the past decade with new types of systems
and antenna measurements being realized. This contribution
exposes the reader to key aspects of current and emerging di- Figure 11: The CROMMA, a hybrid-robotic antenna range.
rections of research in robotic antenna measurements so that
they can explore this exciting and continually evolving area of
metrology. nipulate antennas in 6 degrees of freedom (6Dof) and becomes
unwieldy. Conversely, the use of robotics allows for a compact
Introduction. In the context of radio-frequency metrology, the mechanical system that abides by a well described kinematic
pre-defined and well-established boundary conditions offered model to achieve 6Dof motion in a more general manner while,
by waveguides and connectors do not exist in the same way for maintaining payload, stiffness, and accuracy. This approach
free-space antenna measurements. Rather, boundary conditions breaks away from a literal metaphorical mechanical positioner
are defined through the action of dynamically placing anten- construction, opening up new antenna range design possibilities
nas at multiple poses (position and orientation) in space often and new ways of doing antenna measurements.
times to within a fraction of the operating wavelength [55, 76]-
[79]. Furthermore, the dynamic placement of antennas must Robots for Antenna Testing. In using robotics for antenna
be coordinated in space and time with other equipment within measurements a core concept is that robots provide a platform
an automation infrastructure. With the prevalence of wireless for doing so and allow one to control, program, and automate
systems, the need to test antennas using multiple measurement a multitude of measurement scenarios which can be tailored
techniques over a large range of frequencies (roughly 1 GHz- to a large set of measurement requirements. Two classes of
to-1 THz) has never been more relevant. This need puts a high robots have emerged which have proven useful for antenna
demand on antenna measurement facilities with regards to mea- testing, able to provide both positional accuracy and requisite
surement setup flexibility, working volume, payload, positional working volume among other desirable attributes. These being
accuracy, repeatability, and sampling strategy for characterizing the serial 6-axis (and 7-axis) robotic arms [80, 81] and paral-
a device under test (DUT). One of the main components of any lel 6Dof Stewart-Gough platforms (a.k.a. hexapods) [82, 83]
antenna test facility is the positioning system used to move and (See Fig. 13). Due to their fundamentally different kinematics
place antennas. The capability of the positioner dictates the (i.e., links in series vs. links in parallel) these tend to excel
type of measurements that can be performed (e.g., canonically in different ways as antenna range components, with serial
a plane, sphere, or line). Furthermore, it is paramount that robotic arms excelling at long range motion of several meters
the positioner maintain the alignment of antennas to specified with high payloads (10’s of kg) and nominal accuracy around
coordinate systems and to a specified tolerance as defined by 70 µm (shown [84] to be reduced to < 25 µm with optical
the method of measurement. The use of a system of ad-hoc feedback and calibration), and hexapods excelling at smaller
stacked motion stages such as linear slides and rotary tables has range motion (a few centimeters) but significantly higher stiff-
long been used to achieve the role of the positioner in antenna ness, smaller resolution, and better accuracy (∼ 1 µm). Hybrid
test facilities [76]. This metaphorical mechanical representation robotic manipulators [85, 86] combine attributes from both se-
derived from the literal mirroring of antenna motion from a rial and parallel robotics and provide advantages from both
specific measurement method geometry (i.e., plane, sphere), adding further flexibility and utility. Shown in Fig. 13 is a
while effective in specific cases, is limiting in its utility to ma- multi-purpose (near-field, gain extrapolation, and polarization)
mm-wave hybrid-robotic antenna range [84, 87] capable of both
1 U.S. Government Work. Not Protected by Copyright high accuracy scanning of < 25 µm and DUT alignment resu-

13
RoE Journal Antenna Measurement Challenges and Opportunities - 14/49

ment quality. Robot-based antenna range designs can be scaled

up or down in size as the kinematics remain the same regard-
less, and robot positioners sharing the same kinematics can be
swapped out for others with little or no change in performance.
In the case of multi-robot antenna ranges, the roles and duties of
individual robots can be reassigned or shared between position-
ers through simple programming, thus expanding automation
possibilities and workflow efficiency.
Application Trends: Current and Future. Robotic antenna
measurements span a wide range of emerging applications and
provide a paradigm to perform freshly-imagined configurations
and measurements. Performance boundaries are being expanded
beyond what was once accessible. The capability of scanning
hardware is no longer limiting the theoretical tools that have
been developed, thus making predicted measurements a real-
ity. The generatrix of a near-field scan surface now formable
through robotic path planning into near arbitrary geometries
for instance. This allows one to extended the viewing angle
of a planar scan (e.g., from ±45◦ to ±135◦ ) and exploit ad-
vanced processing techniques on arbitrary surfaces [90]. More
Figure 12: Photo of multi-robot antenna range (top), digital
efficient sampling strategies based on path optimization applied
twin representation (bottom).
to near-field scanning have been used to reduce scanning time
by 16% for instance, using spline-based motion sequences [91]
lution of ∼ 1 µm. Other hybrid configurations suited for lower managed by the intelligence of the robot controller. The use of
frequency applications (i.e., where far-field distances grow over robotics is also enabling one to extend antenna measurements
1 m) implement a long linear slide and multiple robotic arms far beyond the physical laboratory with concepts like digital
[88],[89] (see Fig. 12). The added slide increases the degrees twins (Fig. 12) and Model Based Systems engineering and
of freedom to 13 and allows one to achieve large scan areas in Design (MBSE/MBSD) [89] by bridging computational elec-
various orientations. tromagnetic and physical measurements. The configurability
afforded by robotics combined with MBSE/MBSD approaches
Calibration and Uncertainty Analysis. Another key aspect allow optimization and reduced uncertainties of test configu-
of robotic platforms is the ability to work directly with the un- rations not achievable with legacy static test-range configura-
derlying kinematics to optimize performance. Representations tions. Applications in 5G and 6G advanced communications
based on Denvait-Hartenberg [80], Hayati [81] parameters and are also benefiting from the use of robotics. Serial robotic
unified parameterizations for hybrid systems [86] from spatial arms are enabling compact synthetic aperture systems for im-
measurements (e.g., laser tracker data) can be employed to de- proved OTA measurements with uncertainties [92]. Multi-robot
velop models and implement calibration of robotic positioners multi-purpose testing platforms [93] able to perform a suite
with uncertainties. The ability to employ measurement-based of measurements including channel sounding and OTA for 6G
approaches for robot calibration provides a powerful frame sub-THz applications are emerging to tackle wireless testing
work for improving off-the-shelf positional accuracy, improved challenges of the future. Portable compact robotic antenna test
control and automation, and the ability to develop an uncertainty ranges are also being realized where the antenna test range
analysis based on robust robotic kinematics, thus exposing the can be brought to the device for in-vio measurements of anten-
full capability of robotic antenna testing platforms to the user. nas integrated to larger systems and test beds while combining
with other measurements like thermal testing of active phased
Practical Benefits. It is worth mentioning a few important
arrays [94]. Multi-purpose calibration facilities [88],[89] for
practical benefits that are less often discussed in the literature.
antenna gain, polarization, and electric-field strength have been
Time consuming tasks like antenna mounting and alignment are
revolutionized by hybrid robotic systems.
quicker to achieve with increased repeatability. Climbing scaf-
folding to mount and align antennas, as is the case with fixed Conclusion. Robotics brings a new paradigm giving unfet-
positioners in legacy systems, are mostly unnecessary as the tered access to 3D space with sophisticated automation and a
robots can readily be commanded to easily reachable locations multitude of possibilities for new antenna measurements, un-
and moved back into position, and antenna alignment poses can certainty analysis, and approaches. Current and future antenna
be stored in local variables in the controller. Measurement plan- measurement challenges across a wide range of applications
ning using digital twins can be done offline allowing, in essence, benefit from this approach with one’s imagination being the
multiple virtual antenna ranges to exist simultaneously. In mm- ultimate driver of what new things this platform will enable.
wave applications, setups can be simplified as entire RF systems
can be mounted on robot positioners and rotated as needed, thus
eliminating the need for rotary joints and improving measure-

14
RoE Journal Antenna Measurement Challenges and Opportunities - 15/49

Extreme Temperature Space atdielectric

microwave frequencies is commonly achieved by using low
materials and foams that have some structural capa-
bility such as dense closed cell foams. From a thermal point
Antenna Measurements of view, when aiming for low temperatures, the challenge is to
keep the temperature of the outer surface of the enclosure above
Luis Rolo*, Paul Moseley the dew point, to prevent condensation. Reaching further down
European Space Agency / ESTEC, Noordwijk, The Nether- to -100◦ C and below, the material, and its thickness, need to
lands
ensure that icing does not occur. In this case, the duration of the
*Corresponding author: luis.rolo@esa.int
test and the lowest temperature inside the enclosure also play a
role. In the hot cases, the main challenge becomes maintaining
Introduction. Operational conditions in space pose several the structural integrity of the enclosure as these materials tend to
challenges to the design of space antennas, in particular to loose their strength and become soft at high temperatures, typi-
materials, mechanical and thermal design. The former are cally above 150◦ C. When testing at temperatures above 200◦ C,
chosen carefully to withstand the harsh space environment, such such as for space antennas on board of satellites for planetary
as vacuum and large temperature excursions. Due to their nature science in close proximity to the sun (e.g. Solar Orbiter and
and function, space antennas are accommodated on the outer BepiColombo), extrapolation techniques are then used [96] and
structure of the spacecraft and as such they may experience very in some cases the tests are limited to sample tests [97] to infer
diverse conditions, depending on the application. For example, critical antenna parameters such as ohmic losses. The interest
an antenna on a satellite orbiting Mercury is exposed to a very to expand usability of such domes for relatively large antenna
distinct thermal environment when compared with a satellite dimensions and compatible with a more generalised acquisition
in an geostationary orbit around Earth. One way to ensure that scheme such as spherical near field is growing as demonstrated
the antenna will perform as expected throughout the difference by recent developments [100].
environments it will be subject to is through on-ground testing. (b) Testing under vacuum The move into a vacuum environ-
This typically happens throughout the antenna development and ment is preferred for high frequencies (due to vacuum chamber
can take place during a qualification phase, but also sometimes sizes) and allows getting a thermal environment very close to the
as early as feasibility studies and material selection phases. real scenario as the thermal flows are more representative (con-
When the temperature effects on the antenna performance ducted and radiated). Typical thermal-vacuum chambers are
knowledge are limited to the thermoelastic deformation of its closed systems, where a device can be cycled through a series of
mechanical structure and elements, one technique is to charac- temperatures and simple monitoring such as device temperature
terise such deformations inside a thermo-vacuum chamber and can be recorded. To incorporate free space RF measurement
use the results to build an electromagnetic model that represents with such a facility, there are many additional challenges to
the antenna under those conditions [95, 99]. This technique solve that depend on the facility architecture chosen.
is particularly suitable for large and complex space antennas The design of the Low-temperature Near-field Terahertz
and instruments, but its often limited to pattern and pointing Chamber (LORENTZ) brings the entire measurement system
knowledge. inside a vacuum chamber. This chamber was designed to oper-
For complete performance assessment, characterising an- ate between 80 and 400K and can cover a frequency range of
tennas under extreme temperature conditions typically entails 50 to 1500GHz. It includes a planar near-field scanner with a 1
performing antenna testing while the antenna is experiencing by 1m scan range, can accommodate instruments up to 300kg
the thermal environment. and has a 2 meter internal diameter. The facility successfully
State-of-the-art: Testing space antennas under representative measured the Engineering Model (EM) and Flight hardware for
thermal environments is seen as significant developments which the Sub-millimetre Wave Instrument (SWI) on JUICE satellite.
allow not only to cope with the various environments, but also
enable testing in various schemes such as Compact Ranges,
Near Field test facilities and also testing at material sample
level. Typical antenna parameters that are characterised under
temperature are radiation pattern and gain/ohmic losses, phase
centre estimation, group delay variation, pointing error and
Passive Inter-Modulation (PIM) performance. There are appli-
cation specific parameters which can also be evaluated, such as
Radiometric Performance, for remote sensing instruments, or
Signal-in-Space (SIS) performance for navigation payloads.
The large majority of tests under temperature fall in one of
two categories: a test setup that constrains the environment in
a transparent enclosure where the antenna under test (AUT) is
and is used in a traditional/standard anechoic chamber, and a
test setup that is integrated inside a vacuum chamber capable of
recreating the required space environment.
(a) RF transparent enclosures: Transparency of enclosures Figure 13: LORENTZ Antenna Test Facility.

15
RoE Journal Antenna Measurement Challenges and Opportunities - 16/49

It is also possible to test the antenna inside a vacuum cham- foreseeable that new developments will be necessary to sup-
ber, while keeping the measurement system in a standard lab port the developments on the testing techniques and facilities.
environment. This requires one or two RF transparent windows These might include antenna gain standards built with materials
into the facility, that allow for an antenna to transmit out of with low Coefficient of Thermal Expansion (CTE) which are
or a gaussian beam to pass through to characterize materials, relatively temperature insensitive and integration of Electro-
keeping this way the measurement system in a room pressure magnetic Modelling (EM) tools with multiscale/multiphysics
and temperature environment. The challenge with this approach simulation software to allow optimisation of complex antenna
is to design a sufficient sized RF window that is transparent in digital designs, taking its thermal environment, transient and
RF, but does not thermally load the whole chamber. This is typ- steady-state behaviours into account at an early stage.
ically done at sub-mm wave frequencies, where such windows
have been perfected for astronomical telescopes. This becomes
challenging for lower frequencies and larger devices as the size
of the window must increase. Regardless of the operational
frequency, such a window is never truly transparent and will
always introduce some loses and additional reflections into the
measurement.

Emerging and Future Needs. For the next generation of Cos-

mic Microwave Background (CMB) and far-infrared telescopes,
there is the requirement to fully quantify the systematic error of
the full end-to-end instrument as well as individual components.
At component level there are various common technologies
currently in development and each require their own test con-
figuration: dielectric lenses and Half Wave Plates (HWP) are
being considered as part of the antenna optics, but both need
to be actively cooled down to cryogenic temperatures and thus
will require dedicated testing techniques and facilities. To this
end, calibrated (but re-configurable) test benches for reflective
and refractive optics will need to be developed to allow radiated
tests of such devices and for the full instrument large cryogenic
facilities will be required, capable of reaching temperatures
down to 4K in vacuum.
The size of focal plane receivers proposed are greatly in-
creasing, with some missions proposing arrays with thousands
of detectors. The individual detector will be coupled with a
horn or lens to the sky, each of which has to have its antenna
pattern and gain characterised in representative environmental
conditions: several detector technologies, need to be cooled
below 1 Kelvin which makes testing extremely challenging.
With the growth of available large deployable reflectors in
the market, especially metal mesh-based reflectors like those
proposed for future earth observation missions such as CIMR,
there is the need to accurately calibrate performance of the an-
tenna, taking into account its thermal environment which, for
such large apertures, may include a gradient through the reflec-
tor surface. Such reflectors, when used for multi-band missions,
will also require PIM performance assessment and despite PIM
testing at feed or antenna array element is commonly done to-
day, such characterisation for large apertures will require large
thermal enclosures in an anechoic environment which are yet to
be developed.
In addition, the use of reconfigurable payloads for future
generation of navigation satellites will likely require precise
antenna phase centre and group delay knowledge for multiple
thermal scenarios and payload configurations, with unprece-
dented accuracy.
Establishing and quantifying uncertainties related with per-
formance over temperature is a challenge on its own and it is

16
RoE Journal Antenna Measurement Challenges and Opportunities - 17/49

Automotive Antenna
Measurements at VISTA
Matthias Hein1 *, Francesco Saccardi2
1 RF & Microwave Research Laboratory, Thuringian Center
of Innovation in Mobility, TU Ilmenau, Ilmenau, Germany
2 MicrowaveVision Italy SRL, Pomezia, Italy
*Corresponding author: matthias.hein@tu-ilmenau.de

Introduction. Measurements of the radiation patterns of auto-

motive antennas present a special challenge. On the one hand,
there is a large number of antennas – more than 20 for a middle-
class passenger car, the number of which will increase in view of
future technologies like automated and connected driving, 6G,
or combined terrestrial and non-terrestrial network coverage.
This requires a concept of efficient and accurate antenna testing. Figure 14: VISTA antenna measurement chamber at
On the other hand, the antennas must be characterized in their ThIMo/TU Ilmenau with multiprobe antenna measurement
installed state, since the mounting environment has a significant arch and a car raised into the center of the arch with a scissor
influence on the radiation, especially at lower frequencies. This lift (orange).
requires measurement equipment capable of accommodating
entire vehicles. This paper presents the automotive antenna test
facility VISTA - Virtual Road Simulation and Test Area at the Measurements Capabilities.
Thuringian Center of Innovation in Mobility at TU Ilmenau. (a) 3D Antenna Patterns. Radiation pattern measurements
The semi-anechoic chamber is equipped with a multiprobe arch imply the characterization of the directional variation of the
for fast antenna measurements. After a description of the setup, radiated power density and, after calibration (i.e. gain substi-
current and prospective measurement and post-processing steps tution method [55, 103]), the derivation of gain and radiation
that can be performed in VISTA are sketched. efficiency as well as derived parameters like the total radiated
power or total isotropic sensitivity. A “passive measurement” is
The Automotive Test Range VISTA. VISTA is a test facility applied if the antenna feed points are accessible. In this case,
operated at TU Ilmenau and equipped with an antenna measure- the AUT is connected to the ports of a vector network analyzer
ment system from MVG (Fig. 14) [101]. Its outer shielding (VNA), and the transmission coefficient S21 is measured for
dimensions are 16 m x 12 m x 9 m (LxWxH). Walls and ceiling each angular position. Some challenges arise when performing
are lined with 60” pyramid foam absorbers. The metallic floor and evaluating such automotive measurements. First, consid-
can be covered with absorbers of sizes suitable for the respec- ering that not the AUT alone is radiating but parts of the car
tive testing process (8” to 60”). A 6.5 m turntable is embedded chassis as well, the probes are not in the far-field (FF) for most
in the floor, which contains a 4-wheel-roller dynamometer. One frequencies of interest (e.g. mobile communications). Instead,
of the centerpieces of the chamber is an antenna measurement the measurements take place in the nearfield (NF), and thus
arch of 4 m radius. Its left part contains 111 dual-pol probes the multiprobe antenna arch acts as a spherical nearfield sys-
covering the co-elevation range from θ =0° (zenith) to 110° op- tem (SNF). Hence, the measured data must be transformed to
erating in the frequency range 400 to 6000 MHz. The right part FF with a post-processing step. This transformation is based
consists of 22 dual-pol probes with 5° angular spacing, covering on the spherical wave expansion (SWE) theory [76, 2]. The
the frequency range 70 to 400 MHz. The center of the arch conventional SWE-based NF-to-FF transformation becomes
is 2.3 m above the floor, making it necessary to raise the car critical, when the AUT is located outside the “equivalent sam-
under test on a scissor lift. The space beneath the car can be pling sphere” (ESS), which is defined by the frequency and the
lined with absorbers to emulate free-space conditions. Another angular separation of the probes on the arch [104]. For example,
alternative is to use a specific construction mimicking a per- at 6 GHz, the ESS at VISTA is approx. 3 m, hence smaller
fect electrical conductor [102] or, prospectively, more realistic than a typical vehicle. In cases where the AUT is outside this
materials like asphalt or even artificial road surfaces. The mea- sphere (i.e., mounted in the rear spoiler), advanced processing
surement usually proceeds such that the vehicle rotates on the tools [104, 105] can be used to move the ESS to the fed antenna
turntable underneath the fixed arch. For each azimuthal angle, location, in order to capture most of the radiated power (i.e.
the antenna under test (AUT) is measured along an elevation local measurement approach [104]). Another important aspect
cut with the LF or HF arch. Upon a complete turn of 0 to 360°, is the truncation of the scanning area which could compromise
the electric field strength in amplitude and phase is obtained for the measurement accuracy especially at low frequencies such
all elevation and azimuth angles of the upper hemisphere down as 70 to 400 MHz and near horizon [106]. Simply replacing
to -20° below the horizon. The co-elevation range is truncated the missing field samples with zeros generates a discontinuous
below the vehicle, i.e. θ =[110° to 180°]. field which is difficult to represent in the SWE domain due to

17
RoE Journal Antenna Measurement Challenges and Opportunities - 18/49

the limited number of available spherical wave modes [106]. (VDT) present the only solution to test automotive system in
To mitigate such type of error methods like the iterative modal virtual environment with real hardware (vehicle-in-the-loop)
filtering (IMF) or the equivalent current technique (EQC) can [110, 111]. The complexity related to a full emulation of a real-
be considered [106]. Gain calibration is yet another important istic scenario can be significantly alleviated by the combination
aspect. The above-mentioned substitution method allows the of measurements and simulations [112]. Performing electro-
calibration of the whole measurement system without the need magnetic full-wave simulations of the vehicle with antennas
of accessing individual components of the RF chain. How- integrated is not always possible due to protected intellectual
ever, the gain measurement of the reference antenna is also property rights related to antenna design and signal processing.
affected by errors, for example due to residual reflectivity of On the other hand, the effort to perform full vehicle measure-
the anechoic chamber and/or truncation of the scanning area. ments with the antenna in different positions is prohibitive. A
To improve the accuracy of the gain calibration, the known valuable alternative is to measure the antenna in a stand-alone
efficiency of the reference antenna can be used instead of its configuration, compute the equivalent currents on a Huygens
gain [103]. Indeed, since the efficiency is an integral quantity, box including that antenna and use the EQC in a simulation
it allows to smoothen the measurement errors showing up in software containing the vehicle structure [112]. Such a co-
form of ripples. More specifically, considering the upper hemi- simulation can also be applied to other use cases such as the
spherical efficiency, truncated spherical NF systems, like the evaluation of human exposure to the radiated field inside cars
one at VISTA, can be accurately calibrated [103]. If the RF [113] or virtual verification and validation of automotive radar
feed points of the antenna are not accessible, the simple VNA [114]. Evaluating the radiation performance of a vehicle over
measurement principle must be replaced by an “active measure- infinite flat dielectric ground (like asphalt) is often a test require-
ment”. This means that a signaling link to the AUT must be ment which can be met by combining SNF measurements of
established using a communication tester, for example. Instead the vehicle performed under free-space conditions with image
of measuring the amplitude and phase of the AUT in transmit theory [115, 116]. Even more complex scenarios like urban traf-
mode, which needs special consideration of phase retrieval, an fic environments can be emulated by co-simulation techniques,
alternative is to operate the AUT in receive mode and measure where the EQC of the whole car are numerically derived from
communication parameters like the reference signal received measurement and then inserted into advanced electromagnetic
power (RSRP) instead. A low mean error of 0.2 dB between simulation tools emulating the scenarios of interest [111].
VNA and RSRP measurements was shown [107].
Conclusion. The development of modern automotive anten-
(b) Over The Air Testing. Like in 3G/4G/5G cellular net-
nas and wireless transmission systems requires fast and accurate
works, over-the-air (OTA) testing, where the transmitter or re-
measurements of the whole vehicle to reduce design and test
ceiver are part of the AUT, is becoming essential also in the
cycles [111]. The state-of-the-art multiprobe spherical nearfield
automotive industry. In these cases, the feed point of the anten-
facility implemented in the automotive test range VISTA at TU
nas is not accessible, hence an active measurement setup must
Ilmenau, Germany, was described, and promising approaches
be considered. Classical NF measurement techniques cannot be
to address upcoming challenges related to the combination of
applied in conventional OTA testing because of the lack of phase
measurements and simulations and virtual verification and vali-
coherence between TX and RX. In fact, OTA tests are usually
dation in virtual environment were outlined.
performed in a far-field setup, where the plane wave condition is
achieved by placing the TX and RX antennas at sufficiently large Acknowledgment. The research presented here was funded
distance (direct FF) or by using a compact antenna test range by numerous R&D grants and supported by many people as
(CATR, indirect FF). The disadvantages of these two solutions detailed in the respective references. Substantial contributions
are mainly associated to the high costs needed to implement to this paper from Dr. Christian Bornkessel (TU Ilmenau) and
a large measurement distance in an anechoic environment or Dr. Lars Jacob Foged (Microwave Vision Italy) are gratefully
to realize an effective CATR for automotive applications. On acknowledged.
the other hand, performing automotive OTA testing at reduced
distance (e.g. 4 m as in the multiprobe system in VISTA) is an
appealing solution as long as the measurement uncertainty is not
compromised. In [108], preliminary results of OTA measure-
ments of vehicle installed antennas were presented, showing
that good accuracies can be achieved for typical automotive
OTA figures-of-merit such as partial radiated power/sensitivity.
Moreover, it was shown in [109] how the parallax compensation
technique can be exploited to improve the accuracy in case of
offset mounted antennas.
(c) Virtual Drive Testing. The impact of wireless sensing and
communication technologies on safety-relevant applications in
mobility require scenario-based safety assurance testing includ-
ing the vehicle, the road and traffic environment, as well as the
electromagnetic wave propagation. To cope with the enormous
complexity and deficiency of real drive tests, virtual drive tests

18
RoE Journal Antenna Measurement Challenges and Opportunities - 19/49

Drone-aided In-situ
Antenna Measurements
Cosme Culotta-López1 , Fabio Paonessa2 *
1 QuadSAT,Odense, Denmark
2 CNR-IEIIT,
Turin, Italy
*Corresponding author: fabio.paonessa@cnr.it

Introduction. The use of Unmanned Aerial Vehicles (UAVs)

allows for in-situ characterization of antennas in outdoor set- Figure 15: QuadSAT’s UAV-based measurement platform.
tings. Moreover, their adaptable nature provides opportunities
to conduct various diagnostic techniques that were previously
unfeasible. This article explores works carried out in this field
miniaturized power sensors [133]. In this case, the UAV shield-
until now and highlights some future potential applications.
ing from external interference represents a critical aspect [131].
Current status. In-situ measurements have been mostly re- A further approach aims at minimizing the payload complexity
ported for radio telescopes, radar and satellite antennas, mo- by carrying on the UAV only a probe antenna with some basic
bile and TV tower antenna systems. Early measurements with electronics, thus avoiding the presence of both transmitter and
modern UAVs have been reported since 2014 [117], thereafter receiver. In this case, a wired or fiber-optic link will tether the
publications succeeded with rapid growth. So far, the operating drone to a ground equipment, e.g. a Vector Network Analyzer
frequencies ranged from HF band [118] up to Q-band [119]. (VNA) [127].
Testing of radio telescopes undoubtedly represents an ideal In both cases, the measured data is generally represented by
scenario for in-situ UAV-based measurements. UAVs allow the power level received either at the AUT port or at the UAV.
the calibration and characterization of modern radio telescopes The AUT pattern calculation requires the knowledge of both
based on phased-array technology – such as the Square Kilo- the UAV attitude and position during the measurement and, of
metre Array (SKA) – at both the element and instrument lev- course, the radiation properties of the UAV-mounted payload.
els (end-to-end verification). Experiments have been reported The former is generally measured by Inertial Measurement
on a 6 m×6 m array prototype of the SKA-mid instrument at Units (IMUs) onboard the flight controller with an accuracy
350 MHz [120], on a 16-elements demonstrator of the SKA- of few degrees. As far as the UAV positioning is concerned,
low instrument from 50 MHz to 350 MHz [121], and on two modern UAVs can take advantage of Real Time Kinematic
48-elements arrays located at the Murchison Radio-astronomy (RTK) devices to reach a centimeter-level accuracy.
Observatory from 50 MHz to 320 MHz [122]. Measurements
have also been performed on operating radio telescopes such as As far as the antenna on board the UAV is concerned, wire
the LOFAR low-band array from 32 MHz to 70 MHz [123, 124] antennas represent the most practical solution at the lowest fre-
and high-band array from 124 MHz to 180 MHz [125]. As far as quencies (HF and VHF). They can be fixed to the UAV frame so
radio astronomical reflector antennas are concerned, UAV-based that their orientation coincides with the UAV attitude. In these
measurements have been reported on 5-m dishes at 1 GHz [126], cases, the onboard antenna can be strongly coupled with the
on a 15-m reflector at 2 GHz [127], and an offset-parabolic re- UAV metal parts, therefore the overall electromagnetic behavior
flector at 14.5 GHz [128]. must be accurately studied to maximize the measurement accu-
Speaking of radar and satellite antennas, measurements racy. On the other hand, the smooth pattern of electrically small
have been reported on coastal HF oceanographic radars from antennas generally mitigates the measurement errors generated
3 MHz to 50 MHz [118], on a naval surface-wave radar at by the limited IMU accuracy. As the operating frequency in-
13 MHz [129], and on a folded dipole for the Europa Clipper creases, directive and/or calibrated antennas and even horns can
mission spacecraft at 60 MHz [130]. Finally, broadcast tower be used. Directive antennas can be mounted on gimbals [126],
antenna systems have been measured at 177.5 MHz, 226 MHz [134], in this way the onboard antenna always points toward the
and 598.5 MHz [131]. AUT regardless of the UAV maneuvers, simplifying the data
processing.
As far as the UAV is concerned, the payload generally con-
sists in a radio frequency transmitter, i.e., the UAV represents The most common flying platform consists in a 4-to-8-rotor
the flying test source and the AUT operates in receiving mode. multicopter with a variable mass (generally within 3 kg) and a
An example is shown in Fig. 15. The simplest transmitter flight autonomy between 15 and 25 minutes. Fixed wing UAVs
consists in a continuous wave (CW) generator, but also more have been employed to cover large distance with long flight
complex solutions have been occasionally adopted, e.g. a pulse duration [118]. Generally speaking, multicopters provide more
transmitter in [124], whereas [126] used a flat-spectrum noise versatility in terms of flight strategies, lift capacity and possibil-
source of about 2 kg carried by a 11 kg hexacopter. ity to control the horizontal orientation of the vehicle (the yaw
Airborne receivers have been adopted in some cases; they angle), which is of great important as far as the characterization
mostly consist in portable spectrum analyzers [131, 132] or of the polarization properties of the antenna are concerned.

19
RoE Journal Antenna Measurement Challenges and Opportunities - 20/49

Emerging challenges. In the simplest measurement configu- plication of computationally-heavy NF to FF transformations.

ration, the distance between the AUT and the UAV satisfies the One of the central elements of the array was used as phase
far field (FF) criteria. This is the case of most of the cited works. reference. In [121], the phase reference antenna was instead
In the far field, the AUT radiation pattern can be easily extracted placed in the proximity of a 16-elements SKA-low prototype to
from the measured received power through the well-known Friis retrieve a near field phase pattern. A NF-to-FF transformation
equation. was then applied. Of course, this method assumes the reference
Near-field (NF) strategies become necessary when the Fraun- antenna phase pattern to be known from simulations. It is to
hofer distance is incompatible with the applicable flight regula- note that the two field components needed by the transformation
tions. The conventional approach to compute the AUT radiation were acquired with different flights and were therefore sampled
pattern from near-field data is to apply a NF to FF transforma- on different sets of points.
tion. Traditional algorithms use the phase information of the Finally, the flexibility of UAVs allow for further antenna
received signal in order to perform the transformation. This is diagnostics other than measurements of the radiation pattern,
particularly challenging in UAV-based measurements, as the fly- such as the evaluation of the tracking performance of an AUT
ing instrumentation and the ground one generally do not share for satellite communications [136]. In this case, the pass of a
a common frequency reference, impeding phase measurements. satellite is projected to a shorter distance from the AUT and
Even when obtaining a reliable phase measurement, the posi- emulated by the UAV, matching its velocity and radiated power,
tion accuracy of the RTK receivers limits the applicability of and compensating for Doppler effects. Since the position and
NF-to-FF transformations to S-band. Furthermore, transforma- attitude of the UAV are known in all moments, they can be com-
tion algorithms will deal with a highly irregular spatial domain pared to the known pointing angle of the AUT for diagnostics
compared to conventional regular grids of motorized positioners and refinement purposes. The main advantage of this method is
(e.g. Cartesian, cylindrical or spherical). the repetition of the pass at will, without needing to wait for the
In [135], the EM model of a 30-m wide LOFAR station has satellite, and even the diagnostic of tracking problems before a
been validated in the near field. In particular, the geometrical satellite is even launched.
model of the station incorporated both the array elements and
Future developments. The use of UAVs for antenna measure-
the flying test source. The measurements were then compared
ments has proven to be a valuable tool in outdoor environments.
with the transmission coefficient simulated taking into account
However, there is still room for future research and development
the real UAV flight path and attitude. This represents, however,
in this area. One possibility is to incorporate software-defined
an unconventional and computationally heavy approach.
radios (SDRs) to transmit/receive signals with the drone, or
A phaseless transformation technique has been adopted
to both transmit and receive allowing for a more comprehen-
in [133]. In that case, a UAV equipped with a power sensor
sive examination of the antenna characteristics. Additionally,
was used to measure the radiation pattern of a 2-horn array in
phaseless measurements with phase retrieval algorithms may be
S- and C-band through cylindrical scans in the near field. The
improved to allow measurements in the near-field region using
measurement setup is highly simplified as the phase information
UAVs without additional equipment for the measurement of
is not acquired (i.e., magnitude-only measurements were made).
the phase. Another possibility, assuming reception from the
However, phaseless methods generally minimize a nonlinear
UAV side, is the use of drone swarms to expedite measurements
and non-convex cost functional, potentially leading to an ill-
through the acquisition of multiple parts of the measurement
posed problem that suffers of local minima.
surface simultaneously. Furthermore, advanced measurement
A conventional approach based on a NF-to-FF transforma-
strategies can be implemented by using the processing capa-
tion with phase information has been adopted instead in [127].
bilities of the microcomputer onboard the UAV, thus using the
A VNA on the ground fed the AUT (a 15-m parabolic reflec-
live processing of the measurement data for the modification
tor at 2 GHz) while the received signal was transmitted from
of the measurement path. Finally, the use of RTK-aided IMUs
the UAV to the ground thanks to a RF-over-Fiber (RFoF) link,
to increase attitude accuracy in yaw/pitch/roll could lead to
obtaining reliable phase measurement regardless of the UAV
uncertainties below 1◦ , further improving the quality of the
movement. A laser tracker was present to obtain sufficiently
collected data. These possibilities illustrate the potential for
accurate position data, and the Fast Irregular Antenna Field
continued advancement and innovation in the field of antenna
Transformation Algorithm (FIAFTA) was used to deal with the
measurements using UAVs.
irregular sampling grid.
A further approach aims at applying NF to FF transfor-
mation avoiding the need of tethered flights. It uses an addi-
tional known reference antenna to retrieve the phase informa-
tion. A common receiver samples the signals from both the
AUT and the reference antenna, allowing to compute the phase
difference and eliminating the phase drift caused by the fre-
quency offset between receiver and transmitter. In [125], the
radiation patterns of the LOFAR high-band array have been
measured in the proximity of the beam axis, exploiting the
near-field focusing method. Such a technique avoids both the
time-consuming λ /2 sampling of the aperture field and the ap-

20
RoE Journal Antenna Measurement Challenges and Opportunities - 21/49

Fast Antenna Measurement mmWave frequencies is more pronounced. To overcome this

challenge, modern devices now come equipped with multiple
arrays placed in different parts of the device to ensure full cov-
Techniques for 5G Personal erage is achieved through selection. This effect can be tested
using head and/or hand phantoms or life person testing of the
Devices and Beyond devices as shown in Fig. 16.
Lars Jacob Foged*
Microwave Vision Group, MVG, Pomezia, Italy
*Corresponding author: lars.foged@mvg-world.com

Introduction. Modern personal devices, such as mobile phones,

are equipped with a range of antennas and sensors to ensure op-
timal performance and connectivity. The evaluation of devices
has undergone significant changes, with regards to performance
metrics assessment such as coverage/radiation patterns both
with and without user interference.
Early testing of 1G/2G/3G devices often involved mechani-
cally rotating the device while it was held in a test fixture. This
resulted in high levels of interaction with the fixture, potentially Figure 16: Measurements of shadowing effect by user and
distorting the results. The launch of one of the most popular efficiency of array switching on 5G enabled device at mmWave
smartphones in history in 2010 was a pivotal moment for the frequencies [137].
testing industry. The incident, known as ”antenna-gate” brought
attention to the importance of precise representation of devices Another testing challenge is MIMO OTA testing which aims
during testing, as well as a comprehension of the effects of user to determine system-level parameters such as data throughput
interference on performance. This paved the way for systems with received power at a device in a realistic and complex
based on non-invasive device positioners enabling access to scenario using a standardized channel model in a multi-probe
both standalone and user-influenced performances. This contri- anechoic chamber (MPAC). To achieve this, a channel emula-
bution provides an overview of current testing methods used for tor and a probe array are used to reproduce the RF and spatial
the evaluation of 4G/5G/6G devices. contributions to the signal, respectively. The RF contribution
OTA Testing. Traditionally, the transmit and receive perfor- consists of various elements such as modulation, polarization,
mances of wireless devices have been measured by directly temporal delay, and doppler profile, while the spatial contribu-
connecting the device to test equipment. However, this is no tion is represented by plane waves with a certain angular spread.
longer possible in today’s technology where transmitter/receiver A typical test setup is shown in Fig. 17. More details on MIMO
are part of the device. Instead, testing is done through a remote OTA testing using MPAC solution is outlined in [138].
connection using over-the-air (OTA) methods.
Test parameters for personal communication devices are
defined in the far field (FF), which presents a challenge for OTA
testing. Ensuring a high-quality FF condition in the test envi-
ronment is essential, and compact test environments with low
spatial losses are generally preferred. Striving to achieve mea-
surements that closely approximate the FF condition presents a
paradox, as most communications supported by antennas and
devices occur at a finite distance and often in the near field
(NF). However, the FF state provides a convenient reference
condition that enables traceable and comparable results from dif-
ferent measurement ranges and systems. In OTA testing, spatial Figure 17: Measurements of MIMO OTA performances using
power quantities related to radiated power and device sensitiv- a typical MPAC solution.
ity are typically measured to characterize the transmitting and
receiving properties of a device. At mmWave frequencies the required number of probes for
a full MPAC solution becomes prohibitive as the electrical size
Testing Challenges. In [137], it was reported that the human
of the device growths. This is a key challenge for the testing.
body can significantly impact the performance of personal de-
A potential solution is to limit the angular space of the testing
vices, particularly in mmWave frequencies. This is due to the
reducing the overall number of probes required
strong shadowing effect that occurs, which reduces the cover-
age efficiency of the phased array. Unlike lower frequencies Testing Solutions and Challenges. The standard NF and
like GSM and UMTS, where fields can curve around the user FF measurement solutions used for personal device testing are
and still provide decent 3D coverage, the shadowing effect at described in [55, 76]. The techniques, as shown in Fig. 18,

21
RoE Journal Antenna Measurement Challenges and Opportunities - 22/49

are commonly referred to as direct FF, indirect FF and NF Future Developments. The time to market is a critical factor
techniques. in personal communication. The development of new technolo-
gies necessitates thorough testing, with advancements in mea-
surement speed, accuracy, and convenience serving as important
driving forces. These factors actively propel the progress of
testing methodologies. With the rapid evolution of technology,
there is a clear and notable shift moving away from traditional
testing techniques such as [55] and [76]. Instead, there is a grow-
ing emphasis on adopting more specialized methodologies that
focus on optimizing device performance in realistic scenarios.
This shift is primarily driven by the increased demands arising
from research and development (R&D) activities. Conducting
realistic user experience testing has become indispensable in
the process of device development. Such testing plays a key
role in identifying areas that require improvement, ultimately
Figure 18: Overview of testing solutions for personal devices. leading to the enhancement of overall device performance.
As both testing and numerical modelling techniques are con-
Due to the lower interaction with the test fixture, NF tech- tinuing to develop individually to meet new design challenges,
niques have generally been preferred for testing of personal measurements post processing technology is under refinement
devices designed to radiate in all directions. NF techniques to facilitate the use of a measured data on specific devices as
require a full scan of the measurement surface to determine sources in larger scale simulations. This integration of mea-
performance. For this reason, the multiprobe systems shown surements and numerical techniques in R&D development is
in Fig. 16 were developed. Measurement speed is significantly becoming increasingly important, as it improve development
increased from the substitution of the mechanical probe move- speed and efficiency [139, 111].
ment with the electronically scanned multi-probe array. Public safety using personal communication devices is also
At mmWave frequencies the higher directivity of the an- an area important area of new developments. Until recently the
tenna arrays incorporated in the devices allow testing using standard approach was to mimic device and user interaction us-
indirect FF techniques such as Compact Antenna Test Range ing phantoms and measure the power dissipation within human
(CATR) or Plane Wave Generators (PWG). The advantage of tissue equivalent materials. These legacy methods are likely to
these systems is speed, from the possibility to determine FF per- continue to be in use for type approval testing of devices below
formance in a given direction from the measurements of a single 6GHz. Newer methods are already now preferred for R&D pur-
point [76, 139, 111]. An implementation of the PWG designed poses based on multi-sensor techniques giving access to much
specifically to be able to perform device testing including live faster, reliable testing and data specifically useful for further
or phantom users is shown in Fig. 19. development of the devices. At frequencies beyond 6GHz, the
legacy methods are no longer useful, but as personal communi-
cation move to higher frequencies new testing method to ensure
public safety are required. These methods are mainly based on
post-processing of the measured device and are currently under
development. They aim to determine if power densities in the
close vicinity of the devices are within acceptable health limits
when operated.
By incorporating new testing methods, we can not only
verify but also achieve the optimum performance of personal
communication devices.
Conclusion. Modern device testing methods prioritize fast
and accurate performance measurement with minimum of inter-
action with test fixture and including the user’s impact on device
functionality. This important shift in testing methodology will
lead to improved device performance and more reliable results
and increased safety for consumers.
Figure 19: Personal device testing including live user
experience using movable PWG. Additional PWGs can be
mounted to emulate simultaneous connectivity points.

Other test solution is the reverberating chamber [55]. It

allows to determine integral performance values of devices but
does not provide information on the spatial performance or user
interference.

22
RoE Journal Antenna Measurement Challenges and Opportunities - 23/49

Over-the-Air Testing of
Communication Systems
Wei Fan1 *, Michael Foegelle2
1 National Mobile Communications Research Laboratory, 2014: LTE, 2×2 MIMO 2016: LTE, multi-band 2×2 MIMO
School of Information Science and Engineering, Southeast
University, Nanjing, China
2 ETS-Lindgren, Inc. USA

*Corresponding author: weifan@seu.edu.cn

Introduction. Wireless radio access technologies continue

to evolve, as technologies like 5G and Wi-Fi 6 reach main- 2018: LTE, multi-band
stream adoption and the industry begins to look towards 6G 4×4 MIMO 2020: Qualcomm mmWave antenna module
and Wi-Fi 7. To increase the available bandwidth and overall Figure 20: Cabled testing for LTE phones and integrated
performance, these systems utilize adaptive radio systems with antenna for 5G FR2 phones.
multi-antenna technologies that use digital, analog, or hybrid
beamforming mechanisms as well as employing concepts like
massive MIMO, in addition to exploiting wider bandwidths in spatial multiplexing and transmit diversity, are determined by
both incumbent spectrum allocations and newly opened regula- the propagation channels as well as the antenna design. To
tory bands in the microwave and millimeter-wave (mm-Wave) tackle this problem, OTA testing of 2 × 2 MIMO terminals was
frequency ranges. Radio endpoints typically undergo thorough initiated in [142, 143], with a focus on introducing realistic
conformance and interoperability testing to ensure that they and representative spatial fading channel conditions. Using a
meet the specifications of the design standard for the given radio base station emulator (BSE) and a channel emulator (CE) with
access technology. Traditionally, most of these tests have been appropriate measurement setups in the anechoic chamber, OTA
performed using a conducted setup to evaluate the radio perfor- radiated performance testing evaluates the wireless system’s per-
mance independently from that of the antenna(s) in the system. formance and reliability in a controlled laboratory environment
Over-the-air (OTA) testing was introduced initially as a final that emulates realistic real-world fading channel conditions.
step to evaluate the radiated transmit and receive performance Several MIMO OTA methods were proposed and extensively
of the device through the antennas [141]. Later, the introduc- discussed, including the later standardized radiated two-stage
tion of multiple-input/multiple-output (MIMO) technologies (RTS) and the multiprobe anechoic chamber (MPAC) solutions
made it impossible to isolate the device performance from the [142, 143, 146, 147]. Throughput (i.e. data-rate) was selected
radio environment in which it operates, resulting in the need as the FoM for MIMO OTA testing. 3GPP TR 38.827 specifi-
for RF environment simulation that could evaluate the behavior cation studies the performance metrics, measurement method-
of a complex antenna system over the air [142, 143]. With ologies, channel models, and validation procedure for MIMO
the advent of advanced phased array technologies at mmWave performance evaluation of 5G mobile terminals [145]. The
and microwave frequencies (for base stations), the radio perfor- test methods, originated from 4G MIMO OTA were extended
mance has become much more tightly tied to the behavior of to support OTA testing of 5G mobile terminals. Specifically,
the antenna system [144, 145, 146]. As the radio transceiver the MPAC solution has been selected as the reference testing
becomes more compact and highly integrated, cabled access method for 5G frequency range 1 (FR1) UEs up to 4 × 4 MIMO,
to the radio circuitry becomes impractical if not impossible, while the RTS method can be utilized as well if consistent re-
leading to the need to perform all of the traditional conformance sults with the reference MPAC solution can be achieved. Fig.
test cases through the antenna system using an OTA testing ap- 21(a) and (b) show the setup for 5G FR1 UE OTA testing with
proach. A photo of why testing will move towards OTA mode is MPAC and RTS methods, respectively [148, 149]. For 5G FR2
shown in Fig. 20. Advanced radio technologies also pose huge UEs, only the simplified 3D MPAC solution has been selected,
challenges on their OTA performance and conformance testing, as shown in Fig. 21(c) [148], with limited capability to emulate
due to the high system complexity, high implementation cost, spatial channels. Although arbitrary channel models can be
long measurement time, and high measurement uncertainty. emulated by the RTS method in principle, the RTS method has
Current State-of-Art. OTA testing of wireless device perfor- been only validated for 2 × 2 MIMO mobile handsets in FR1.
mance was initialized for single antenna (single-input single- OTA performance testing of base station under realistic fading
output, SISO) mobile terminals in CTIA and 3GPP, where conditions is still in its infancy in the standardization.
figures of merits (FoMs) total radiated power (TRP) and to- OTA radiated conformance testing evaluates a wireless sys-
tal isotropic sensitivity (TIS) were selected to characterize the tem’s capability with respect to the transmitter (e.g. power,
transmit and receive capability of the mobile terminals [141]. signal quality, unwanted emission, etc) and the receiver (e.g.
However, SISO OTA testing was deemed not sufficient to cap- dynamic range, sensitivity, selectivity, and blocking, etc) [150,
ture the performance of MIMO-enabled mobile terminals, since 151]. Several measurement systems have been employed for
the performance enhancement introduced by multi-antenna, e.g. radio frequency (RF) OTA metric measurement for directional

23
RoE Journal Antenna Measurement Challenges and Opportunities - 24/49

figuration and beam-steering capability of 5G radios) and high

system cost (due to the demanding and expensive testing cham-
ber environment). The current RF testing methods measure the
beams sequentially with a beam-lock function until all available
beams are measured. Considering that the antenna array pattern
becomes more directive with a larger array configuration, the
entire radiation pattern measurement is very time-consuming
since very fine spatial resolution is required over the sphere
(a) (b) enclosing the antenna for each beam-steered state. Besides the
radiation characteristics, it is also of importance to efficiently
calibrate the large-scale antenna arrays in practical measure-
ment setups. The conventional “probe and park” solution with
the help of mechanical scanner is slow and inaccurate [158].
Recently, array diagnosis and calibration in the all-on mode has
been actively discussed to improve the measurement efficiency
and accuracy [159].
Another challenge is the measurement deviations introduced
by practical setup, e.g. power-only measurements, limited mea-
(c)
surement range, non-anechoic chamber environment, hardware
Figure 21: Measurement setup for OTA testing of 5G NR: (a)
impairment, etc. It is highly desirable that we can perform the
The MPAC setup for 4 × 4 MIMO terminals at FR1 [148]. (b)
transmitter and receiver conformance testing of the commu-
The wireless cable setup for 4 × 4 MIMO terminals at FR1
nication radios in cost-effective measurement setups in a fast
[149]. (c) The MPAC setup for 2 × 2 MIMO terminals at FR2
and accurate manner. Miniaturization of testing chamber is
[148].
essential to reduce setup cost, which has attracted great research
attention recently, e.g. mid-field solutions, which aim to recon-
testing, i.e. direct far-field method (DFF), compact antenna struct the far-field results from mid-field measurements using
testing range (CATR), plane wave generator (PWG) and near- compensation or extrapolation algorithms [160]; metamaterial
field to far-field transformation (NF-FF) techniques and non- enabled absorber design, which can significantly reduce the
directional testing (e.g. TRP and TIS measurement in rever- absorber dimension while maintaining the same performance
beration chamber) [152]. The required measurement distance as traditional absorber [161]; algorithms to reduce echo in the
to meet the far field assumption in the DFF setup increases test environment so that we can retrieve target results in non-
significantly for 5G radios as the antenna electrical dimension anechoic measurements [162].
continues to grow, which makes it impractical for massive de- As more integrated antennas will be employed in future
ployment due to physically large testing setup and link bud- radios, it becomes essential to have radiated OTA access to the
get issues. CATR, which can reduce the measurement range digital receivers in the radio systems. This has been achieved
compared to the DFF, has been employed for BS testing cur- for mobile terminals supporting 2 × 2 and 4 × 4 MIMO schemes.
rently. However, it still suffers from high cost and large test However, it will become problematic as the MIMO order gets
chamber as the BS antenna aperture gets larger. Recently, a larger. For the MPAC system, the number of required probe
multi-feed CATR was designed to generate plane wave from antennas (thereby associated CE resource) will increase as more
several distinct directions, which is promising for radio resource antennas are employed in the DUT, leading to expensive sys-
management (RRM) measurements of 5G radios [153]. PWG tem design. To ensure the channel reconstruction accuracy, we
offers a shorter measurement range and lower cost compared to typically have to sacrifice channel emulation flexibility, as done
the CATR. However, its supported bandwidth cannot cover the in FR2 MIMO OTA testing [145]. For the wireless cable solu-
required 1.7 GHz to 7.2 GHz due to the limitation of the phase tion [147], the condition number of the transfer matrix between
shifters. Furthermore, its application in the mmWave bands is probe antenna ports and DUT antenna ports will get large as
still not mature. Significant efforts have been taken to make the the number of antennas grow, making it difficult to achieve in
PWG a reality, e.g. support for ultra-wideband bandwidth and practice (as it requires inversion of the transfer matrix). It is
mmWave frequency band, generation of a spectrum of oblique currently still an open challenge.
incident plane waves, PWG in non-anechoic deployment en- Conclusion. In conclusion, we briefly revisited the current
vironment and cost-reduction by reducing complexity of the standardization for over-the-air testing of communication de-
PWG feeding network [154, 155, 156, 157]. NF-FF solutions vices. We explained the new radio technologies introduced for
can further reduce the measurement range. However, it suffers the future radio systems and new challenges on over-the-air test-
from long measurement time, and it cannot be directly used for ing introduced by the new antenna technologies. Over-the-air
modulated signal measurements. testing is seen inevitable for future integrated radio systems.
However, strong efforts are required from academia, industry
Emerging Challenges and Future Development. The key
and standardization to make the OTA testing more accurate,
bottlenecks in standardization for conformance OTA testing are
fast, and cost-effective.
the long measurement time (due to a large-scale antenna con-

24
RoE Journal Antenna Measurement Challenges and Opportunities - 25/49

mmWave Active Antenna superposition of the measured beams generated by a codebook

forms the so called Radiation Pattern Envelope (RPE) which,
while being a non-physical radiation pattern, is of the utmost im-
Systems Measurements - portance for emission compliance verification and interference
assessment during network planning activities.
an Overview Given the number of measurements required for an accurate
characterization of mm-wave antenna systems and the reduced
Claudio Massagrande* , Roberto Flamini space required for high-frequency measurements, it is prefer-
Huawei Technologies Italia Srl, Milan Research Center, Mi- able to perform these in an anechoic chamber instead of an open
lan, Italy field range. Contrary to older antenna systems, 5G systems not
*Corresponding author: claudio.massagrande1@huawei.com only require a more comprehensive characterization, but also
require a more complicated assortment of wires and cables in
Introduction. Active Antenna Systems (AAS) have received order to power and configure the Antenna Under Test (AUT).
larger attention lately with the introduction of 5th generation Extreme care must be paid to how these wires and cables en-
(5G) systems. With a huge paradigm shift with respect to previ- ter the anechoic chamber, since every opening in the room’s
ous generations, 5G systems rely on actively tracking the users shielding degrades its immunity from external sources of EM
within the cell, with a dedicated beam generated by a phased noise. Also extremely important is to ensure that none of these
array, in order to maximize the radiated energy to the individ- cables and wires impede in any way the rotational and transla-
ual users and to exploit Space Division Multiplexing Access tional movement of the measurement axis inside the chamber
(SDMA) technology within the cell, thus enabling frequency as this could result into permanent damage of the measurement
reuse. Moreover, in order to support the increased data rate, system or, in the best case, measurement failure. Finally, care
larger portions of the spectrum in the mm-wave range (Ka-band must be taken that the digital control signals of the VNAs and
and E-band) have been reallocated from point to point legacy PSs do not interfere with the measurement, this is not a trivial
systems to wireless access to the 5G network, making larger concern since the spectral content of high speed digital signals
channels available to users [163]. may contain high order harmonics which extend well within the
In a Base Station (BS), the Active Antenna System can mm-wave spectral range.
be seen as a composition of three main parts: an analog core
(usually composed by a set of transceivers - TRXs), a Radio Dis- Antenna measurement typology. The limited room available
tribution Network and an Antena array. The TRXs, composed in anechoic chambers typically offer the possibility to perform
by Variable Gain Amplifiers (VGA) and Phase Shifters (PS) either Compact Antenna Test Range (CATR) or Near Field (NF)
required for antenna reconfiguration, are very often integrated measurement. CATR measurements are faster, as they are an
within the same unit as the antenna, eliminating the physical indoor replica of open range measurements, with a properly
interface between them, and forcing the characterization of the designed reflector positioned in the opposite end of the chamber
active control section along with the purely radiating section of generating a flat phase front over the quiet zone (where the AUT
the antenna. is measured) when illuminated by the source antenna, but they
In this work we will focus on the challenges in the field of only provide the far field radiation pattern along one cut on each
antenna measurement brought about by these new systems and pass. NF measurements are longer, they can take up to a few
discuss some approaches on how to best address these issues. hours, as a surface enclosing the AUT needs to be sampled at an
interval compatible with the antenna wavelength and aperture
Antenna measurement set-up. Previous generations of wire- size, but they provide a complete description of the fields around
less telecommunication equipment all relied, for the most part, the antenna and, from this information, the full 3D far field
on fixed antenna systems to transmit and receive their signals. pattern of the antenna can be derived by applying a spherical
Point to point radio links in the Ka-band employed, in the vast wave expansion of the source fields [76]. It is worth noting
majority of cases, fixed Cassegrain type double reflectors an- that, when performing NF measurements, it is of the utmost
tennas of standard sizes (30 cm, 60 cm main reflector diameter) importance to know the exact position of the sampling probe
while sub-6GHz base station antennas for network access em- over the antenna in order to properly compute the spherical
ployed fixed arrays. The lack of pattern re-configurability of wave expansion. Any inaccuracy in the sampling point creates
these antennas and their passive nature made their electrical an unwanted phase contribution to the source fields which will
characterization relatively simple by performing standard an- affect an accurate far field pattern reconstruction. The positioner
tenna measurements of fully passive devices, either in anechoic accuracy requirement obviously scales with the measurement
chambers or in open field ranges. 5G antennas, on the other frequency, since a position error in the range of 1/100th of a
hand, require a more complicated measurement setup as well wavelength is normally considered as the acceptable limit. For
as an accurate characterization of a whole set of configurations, E-band systems this requirement translates into a positioner
each corresponding to one specific beam pointing direction. accuracy in the range of 40 µm from the nominal sampling
This set of configurations is usually referred to as a codebook position. In Fig. 22, an example of an Active Antenna System
and the beams it generates are usually referred to as a Grid measured by means of a Near Field test range is shown.
of Beams. The individual beams within the grid can be opti- The full knowledge of the fields around the antenna is es-
mized in order to achieve the best network performance. The sential in identifying, for example, faulty elements within the

25
RoE Journal Antenna Measurement Challenges and Opportunities - 26/49

Figure 23: Measured radiation patterns of calibrated and

un-calibrated active array versus target pattern (numerically
Figure 22: Near Field measurement setup of a mm-wave obtained with ideal excitation set).
antenna system. The control electronics is on the backside of
the unit properly shielded by absorbing material in order to
minimize its impact on the measurement. Antenna measurement standards. The novel architecture
of active antennas introduced by 5G systems have also had an
impact on long-lasting standards of antenna measurement [55].
array or to verify if all the radiators are properly excited, and it Integrated active antennas with the radiating section on the top
is not available by CATR measurement. NF measurements are layer of a circuit board and distributed amplification on the
also useful when beam forming techniques are used to generate backside of the board, with no accessible interface to character-
a more complex pattern than a single beam, for example if nulls ize the antenna, have required a new set of Key Performance
are required in certain position for interference cancellation, in Indicators (KPI) to be defined for antenna characterization.
this case a single 2D CATR scan may not be able to provide Traditional parameters such as Gain and Directivity have
all the relevant information regarding the pattern generated by been replaced by Total Radiated Power (TRP) and Equivalent
the array. It would not be fair, however, to disregard CATR Isotropic Radiated Power (EIRP) as it is not trivial to separate
measurements as a valid option for 5G antenna systems. Indeed, the contributions by the antenna to those by the active devices
they are extremely useful for specific situations such as a fast when performing Over The Air (OTA) measurements [164]. As
pattern verification along defined planes or during production a matter of fact, in integrated active antennas the EIRP (that
stages where it would not be conceivable to dedicate a few is defined as the antenna realized gain times the net power
hours of precious chamber time to one single system. In the accepted by the antenna) can be measured with the support of
end it is important to recognize the strength and drawbacks of properly characterized reference antenna and power source. On
both measurement methods and to choose the most suited one the contrary the accepted power and therefore the realized gain
according to the specific need. of the antenna, cannot be accurately calculated. In this case the
AAS gain declared by the Original Equipment Manufacturer
Antenna calibration. The information derived by a NF scan (OEM) is essentially derived as the EIRP divided by the output
can also be used to perform the calibration of the phased array power configured in the system.
control elements, which is fundamental for the proper function-
ing of the active antenna. In fact, the control elements of the Conclusion. In this brief contribution, the application sce-
array (named beam-formers) are essentially mm-wave VGAs nario for mm-wave active antenna systems within the 5th gen-
with Gain in the range of 30dB placed after mm-wave PSs ca- eration ecosystem was discussed. We then explored the issues
pable of achieving full 360° of phase rotation, both VGAs and that arise when setting up the measurement of an integrated
PSs usually being controlled by a 5 or 6 bits word giving them active antenna system operating in the mm-wave range in an
32 or 64 independent states. anechoic chamber, the conditions which influence the choice
of CATR or NF measurement and the factors which can affect
Since mm-wave devices are affected by a high degree of pa-
the quality of the measurement. The need for an appropriate
rameters dispersion, it is very unlikely that all the beam-formers
calibration of the active antenna system was briefly discussed
in a single system align perfectly to feed the respective radiat-
and finally an overview of how traditional measurement stan-
ing elements with the desired phase and amplitude combination
dards are impacted by the new integrated active systems and we
synthesized by the base band (Fig. 23), even choosing to neglect
presented the new standards introduced by the NGMN (Next
any additional misalignment generated by the Beam Forming
Generation Mobile Networks) recommendation.
Network (BFN). The NF sampling of the fields in front of each
element provide the information needed to equalize all the radi-
ators within the array and to compensate for the dispersion of
the respective control chains.

26
RoE Journal Antenna Measurement Challenges and Opportunities - 27/49

Millimeter and related phaseless processing algorithms [167]. This limitation

applies to source/antenna integrated systems dedicated to com-
munications or automotive radar applications.
Sub-millimeter Antenna
RF Sub-system. Generating and detecting mmW signals lie
Measurements at the heart of making measurements at these high frequen-
cies. A limited ability to generate power or to amplify that
Laurent Le Coq1 *, Daniel Janse van Rensburg2 power becomes a severely restrictive aspect. Also, path loss,
1 IETR, University of Rennes, Rennes, France albeit free space or guided wave loss, limits the minimum de-
2 NSI-MI tectable signal at the receiver. A proven technique to largely
Technologies, Suwanee, GA, USA
overcome these guided path loss limitations is frequency up
*Corresponding author: laurent.le-coq@univ-rennes.fr
and down conversion at well selected locations in the RF sub-
system [168] by using a distributed RF sub-system. Using this
Introduction. Antenna testing at frequencies greater than 30 type of RF design also has the added benefit of minimizing the
GHz are typically regarded as millimeter and sub-millimeter mmW components (including receivers operating at interme-
(mmW) testing. The test techniques deployed at these fre- diate frequencies (IF)) required to make such high frequency
quencies do not fundamentally differ from those used at lower measurements. It also opens the path for a high degree of mod-
frequencies, but typically vary in terms of implementation. RF ularity, in terms of banded mmW solutions. Addressing the RF
instrumentation and mechanical implementation is driven by power and loss problem focuses on magnitude-only measure-
significantly higher RF loss in guided wave paths and tighter ments and although this would address the majority of test cases,
mechanical tolerancing required to control phase uncertainty the ubiquity of near-field measurement test systems [169] that
driven by the shorter wavelengths. Radiation characterization require a complex signal to be measured, make the detecting
concerns any type of antenna, but specific topics have to be of phase in mmW measurement systems an essential capability.
mentioned as they refer to unique research activities or develop- Fortunately, the use of frequency conversion techniques and
ment of industrial applications, and these highlight the principal coherent IF receivers make this possible. However, the detec-
constraints for testing: embedded antennas (Antenna In Pack- tion of signal phase at mmW frequencies is highly affected by
age, Antenna On Chip)[165] and mono/multi-beams antennas temperature variation, mechanical vibration, cable flexing and
[166] for example. In the first case interaction between the overall RF stability and these factors become uncertainty induc-
measurement system and antenna under test is key. In the sec- ing parameters into mmW measurement systems [170]. In order
ond case, mechanical and motion accuracies dictate the effort to mitigate coupling between the Antenna Under Test (AUT)
required. In both cases, several solutions, mixing software and and the measurement probe, Electro Optic Systems have been
hardware topics, have been proposed to overcome existing limi- developed [171, 172]. Even if further development is needed,
tations of classical measurement systems, or to implement new their implementation in next generation measurement systems
measurement systems appropriate for mmW frequencies. will offer a clear benefit for near-field applications.
Antenna technology. Compared to low frequency range ap- Test Facility Types. Test facility types can be divided into
plications, mmW range applicable technologies are deeply im- amplitude-only and complex signal classes. In many cases
pacted by physical size of the antennas and the relevant feed amplitude-only test systems will be instrumented with phase
structure losses. The use of coaxial cables and connectors is capable RF sub-systems (e.g. Vector Network Analyzers) but
limited, due to their maximum effective frequency of use, their the phase would simply be discarded. However, it is possible
losses, and their physical size compared to the antenna size. In to also use scalar instruments like power meters or spectrum
fact, most commercial RF connectors are electrically large at analyzers which do not allow for a complex measurement to be
mmW frequencies, inducing parasitic effects that can adversely performed and the purpose would be to reduce RF sub-system
affect the antenna performance. A solution to this problem is cost. Examples of such test systems would include: Direct
integration of the antenna in or on the operational platform. far-field illumination range (FF): This type of range represents
This has led to on-chip integration of antennas, which compli- the simplest (conceptually) test solution where the transmitter
cates testing of the antenna. This either requires special probes and receiver is separated by a distance such that the AUT ex-
that are used to excite the antenna at chip level (which often periences an effective far-field condition. An appropriate RF
requires microscopic positioning of the probe) or a system-level sub-system designed for the range length and a mechanical posi-
test approach where the antenna cannot be separated from the tioner allowing for the desired motion of the AUT during testing
accompanying sub-system (also referred to as active antenna is required. This solution is very practical and in common use
testing). In the THz domain, integrating frequency genera- in industry today up to frequencies of 1 THz. Compact Range
tion modules (diodes and photo-diodes) on the antenna is a (CR): This type of range can be viewed as a direct far-field
research topic of interest. This integration introduces a spe- illumination range where the range length is reduced by using
cific problem for radiation characterization as the RF signal a feed/reflector combination to create a suitable test zone. The
cannot be synchronized to classical RF systems, or optical- instrumentation required is identical to that of the FF range.
based measurement systems do not provide phase information However, the biggest challenge for this type of range is the
(frequency-domain spectroscopy). This constraint requires the fidelity of the reflector being used. Machining and finishing a
use of specific magnitude-only measurement procedures and reflector surface to support mmW frequencies is challenging

27
RoE Journal Antenna Measurement Challenges and Opportunities - 28/49

(e.g. surface accuracy of < ±30 µm RMS to support 200 GHz

[173]) and CR systems supporting measurements above 100
GHz are not common in industry (Fig. 24 shows a machined
CR reflector with near-optical surface finish). Despite this strin-
gent constraint, this type is of main interest as it enables direct
analysis of the measured data [174] and can be coupled to an
interpolation algorithm to provide efficient 3D characterization
[175].

Figure 24: CR reflector with measured ±50µm RMS surface

tolerance offering a reasonable optical reflection.
Figure 25: Articulating arm SNF scanner.
Hologram: To overcome CR limitations but following the
same philosophy of generating a plane wave in a limited physi- foundation stability. Both of these factors impact RF phase sta-
cal space to create a well-defined test area volume (the Quiet bility most (since motion induced translates to phase variation)
Zone) in which the antenna under test must be located during and this aspect requires tight control of these environmental
the measurement process. An example solution has been devel- parameters. Since most mmW test solutions push the bound-
oped using a transmission-type hologram [176]. This solution aries of mature RF technology today, measurement uncertainty
mitigates the geometrical accuracy needed by the collimating is often higher than what is desired. Implementation therefore
element but impacts the frequency range of measurements (i.e. revolves around understanding and controlling the factors most
limits the bandwidth). Plane wave generator (PWG): Genera- adversely driving the measurement uncertainty budget. Using
tion of a plane wave for test applications can be achieved using the uncertainty budget as a guide during implementation is key
CR techniques or arrays of discrete radiators. Practical exam- to success.
ples of the latter have not been commercially implemented at
mmW frequencies. Recent 5G developments require the design Chambers/Anechoic Environments. Radiation characteri-
of new measurement systems to characterize spatial beam recon- zation is impacted by any source of corruption of the signal
figurability in a multi path link scenario and CR techniques are to be measured. Close attention must be paid to the near and
deployed to address this [177]. This application is in the lower far surroundings of the AUT. Specific material treatments must
part of the mmW band and it can be expected that higher fre- be made to minimize specular and multiple reflections in the
quency coverage will be required in the near future. Near-Field measurement chamber. Relative size of the AUT and the facility
range (NF): Near-field test systems at higher frequencies are positioner requires that special attention be paid to the region
principally limited by the ability to measure RF phase accurately. close to the AUT. For both problems, a classical solution is to
Technology is available today to measure complex RF signals use absorbing material, which then requires characterization of
in excess of 1 THz. However, maintaining RF phase stability this material in the mmW range [182, 183].
during antenna or NF probe motion becomes the next major
Conclusion. Antenna testing at mmW frequencies do not fun-
challenge and this is true for planar [178], spherical [179, 180]
damentally differ from those at lower frequencies. However,
and multi-axis robotic test systems [181]. Figure 25 shows a
high RF losses, short wavelengths and the electrical size of
SNF scanner implementation suitable for mmW antenna testing.
commercially available cables and connectors make special im-
Errors. Measurement uncertainty associated with the various plementations essential. These can often be as simple as tighter
test systems described can be determined through a Range As- mechanical tolerances, but can also be more sophisticated accu-
sessment and reported in an industry norm 18-term framework racy enhancements through software correction techniques or
as described in [3]. Parameters that typically drive the overall phase-less NF solutions. This short write-up attempts to give
uncertainty are mechanical position accuracy of both the NF the reader a broad view of the solutions that are available in
probe and the AUT [170], RF sub-system stability and dynamic industry today or are being pursued through research to expand
range as well as RF/LO cable performance. Additional factors the boundaries of possibilities in the future.
that also impact high frequency test systems are temperature and

28
RoE Journal Antenna Measurement Challenges and Opportunities - 29/49

Phaseless Antenna conventional indirect off-axis holography are summarized in

[189]. One is the modified hologram, based on the removal of
the autocorrelation terms of the hologram before the filtering
Measurements process (for example by means of an extra measurement to char-
acterize the amplitude of EAUT ), deriving in some advantages as
Benjamin Fuchs1 *, Fernando Las-Heras2
1 Federal
the diminishing of overlapping, the reduction of the bandwidth
Office of Communications, Biel, Switzerland in the k-space (and thus, the sampling requirements) and the
2 Department of Electrical Engineering, University of Oviedo, reduction of size of the setup. Other is the synthesized reference
Gijón, Spain
field off-axis holography [190], in which the reference branch is
*Corresponding author: fuchsbenjamin81@gmail.com
composed by a phase shifter that synthesizes a plane wave that
is added to the acquired field of the AUT using a power com-
Introduction. Antenna measurement based on near field to far biner at the receiver frontend (an alternative way for generating
field transformation requires precise knowledge of amplitude such phase delays can be performed by mechanical shifts of the
and phase data on a prescribed surface. Phaseless or phase probe antenna). Progress in broadband antenna characterization
retrieval or amplitude-only techniques typically refer to far- has been set with the broadband off-axis indirect holography, in
field pattern characterization when only amplitude data of the which a frequency interference hologram is composed at each
near field is available due to the high impact of phase error acquisition point from measurements on a frequency range, be-
in high frequency measurements [185] (e.g., flexing cables at ing compatible with non-redundant sampling techniques and
frequencies above 100 GHz) or due to a complete lack of phase- suitable for non-canonical acquisition surfaces. Another alter-
measurement hardware. Even, phaseless techniques can be used native is the use of a reference signal produced by the AUT and
as alternative to probe position error correction methods [184]. collected by a second probe moving simultaneously over the
One approach of phase retrieval techniques classification would near-field scanning surface. The complex signals measured by
lead to three categories: four magnitudes techniques, indirect the two probes are summed, both in phase and in quadrature, by
holography techniques, and multiple scan techniques by means two hybrids mounted on the same circuit board. The squared
of iterative and optimization schemes. Others can be seen as amplitudes are detected by four diodes [187] and subsequently
combination or variations of such techniques. In the four mag- processed to retrieve the unknown phase.
nitudes techniques, the phase difference between two complex
signals can be determined from four magnitude measurements.
The two signals can be selected as probe and reference channel,
two probes at different positions, two orthogonal polarizations,
etc. With this technique, the phase difference can be determined
from the magnitude values of such signals and linear combi-
nations of such signals and phase delayed representations of
such signals. A general description as well as hardware and
calibration limitations can be found in [186]. Implementations
of multi-probe cylindrical near-field have been proposed [187].
Indirect off-axis holography, also known as Leith-Upatnieks
holography [188], is an interferometric technique based on the
acquisition of the amplitude of two or more interfering signals,
which is adapted from optical holography to amplitude-only
antenna measurements. In indirect off-axis holography, the mea-
surement setup includes not only the AUT and a probe antenna
but also a reference branch with a “reference antenna” (Fig.
26). In addition to the signal transmitted by the AUT (EAUT),
a fraction of the generator signal, using a directional coupler,
is transmitted by a reference antenna (Er) separated from the
AUT. To balance the amplitudes of the signals radiated from
both branches and, therefore, increase the dynamic range of the Figure 26: Indirect off-axis holography setup for phase
acquired signal (hologram), an attenuator or amplifier is usually retrieval antenna measurement [189].
inserted in the reference branch.
Regarding the two-scan techniques, one of first approaches,
h(x, y) = |EAUT (x, y) + Er (x, y)|2 (5) the Gerchberg-Saxton algorithm [191], alternated substitutions
Both signals are received by the probe antenna to create the and back-substitutions of the magnitude using Fourier transform.
hologram (h in Eq. 5) from which the complex pattern of the In [192], the algorithm is formulated as alternating projections
AUT (and, hence, its phase) can be retrieved by means of appro- onto convex or non-convex sets and the “stagnation” interpreted
priate filtering. This conventional indirect holography requires as due to local minima of the objective functional. The men-
a complete characterization of the radiated field of the reference tioned approaches inspired more refined phaseless antenna test-
antenna. Some variations and improved techniques over the ing methods ([193], [194]). The two-scan techniques exploit

29
RoE Journal Antenna Measurement Challenges and Opportunities - 30/49

two amplitude acquisitions and can be grouped in two families tion of phaseless samples over multiples frequencies [202] are
of algorithms. The first class aims at recovering the missing efficient approaches. Leveraging prior knowledge about the
phase of the data while the second sets up a relation between antenna phaseless measurement problem, such as the geome-
the antenna features and the amplitude patterns and then recov- try of the measurement setup and/or the characteristics of the
ers the features from the measured data. Both classes involve unknowns to be retrieved, enables to go beyond the bare phase
iterative procedures. The first class of algorithms (e.g. [194]) retrieval formulation. These so-regularized data fitting prob-
starts from an initial guess for the field phase of the first pattern; lems can be easier to solve. Recent works have shown that deep
then, the complex field is “propagated” to the second scan and learning methods can be used to solve a wide range of inverse
the new numerically obtained phase is attached to the second problems, including non-linear phase retrieval problems [203].
measured amplitude pattern; finally, the estimated complex field However, if the power of data in fitting models is unarguable,
for the second scan is “back-propagated” to the first scan and the most reliable and accurate phase retrieval algorithms to
the process is iterated. The procedure stops when the propa- come will likely be a combination of data-centric deep learning
gated/backpropagated and measured amplitude patterns become and model-based optimization having an understanding of the
close enough. Such an approach is named Plane-To-Plane (PTP) underlying physics.
and the two scans typically involve measurements over two dif-
Conclusion. Due to their many benefits, phaseless measure-
ferent scanning surfaces. For the latter class ([195, 196]), the
ments have pervaded a wide range of applications and stimu-
unknown is a feature of the antenna to be characterized, for ex-
lated the development of various experimental configurations
ample, its fields at the aperture or equivalent currents around the
and reconstruction algorithms. In the context of antenna char-
antenna. A radiating model links the unknown to the measured
acterization, phaseless measurements typically refer to the far-
amplitude patterns and the distance between the latter and the
field pattern characterization from the measurement of magnitude-
numerically predicted one is minimized. Measurements over
only near field. They avoid the need to accurately measure
two scanning surfaces [195] or by two probes scanning the same
the phase of the electromagnetic field radiated by the antenna
surface have been proposed to achieve uniqueness. Differently
under test, a step that can be delicate and costly, especially
shaped antennas with different scanning geometries have been
at high frequencies. The price to pay to retrieve the missing
considered [196]. Regarding optimization techniques, the phase
phase is the need to perform more measurements than neces-
retrieval problem can be also expressed as the solution of the
sary for a complex antenna characterization. Thus, phaseless
following equation:
approaches typically resort to several measurements (e.g., the
four magnitude and multiple scan techniques) or require the
|Ax| = |b| (6)
use of a known reference antenna (e.g., the indirect holography
where x is the source representation of the antenna (typically approach). Optimization methods have also been developed
equivalent currents [196] or spherical wave coefficients [197], b to solve the notoriously difficult phase retrieval problem and
the collected measurement samples, and A some discretization various formulations, initializations and regularizers leverag-
of the radiation operator. Eq. 6 can be used to model any an- ing prior knowledge have been proposed to help converging to
tenna phaseless measurement problem in matrix notation. This the optimal solution. Finally, despite the significant improve-
facilitates the application of non-linear optimization methods to ments made in phaseless antenna measurements, there is still
find the solution vector x. Some forms of the iterative two-scan work to be done in order to reach the maturity and uncertainty
method have been formulated as an optimization problem of a estimation of complex antenna characterization.
properly defined functional ([195, 196]).
Acknowledgment. To Olav Breinbjerg, Angelo Liseno, Fer-
Emerging challenges and future developments. Phaseless nando Rodrı́guez, Thomas Eibert, Ana Arboleya, Jaime Laviada,
antenna measurements continue to draw a lot of attention fueled Yuri Alvarez, Laurent Le Coq and Nicolas Mézières for their
by the ever-higher operation frequency of antennas and recent useful discussions on the topic.
advances made in applied mathematics including machine learn-
ing. Optimization methods based on convex formulations [198]
and non-convex gradient-based techniques combined with a
proper initialization have been proposed this last decade to
efficiently solve, in an automatic fashion, large-scale phase
retrieval problems. The implementation of most of these tech-
niques is readily available online [199] and can be plugged-in
electromagnetic modeling tools to efficiently solve phaseless
antenna measurement problems [200]. However, none of these
techniques can generally prevent from falling into sub-optimal
solutions for Fourier-based magnitude-only measurements, such
as phaseless antenna measurements. Oversampling can help
converging to the optimal solution, as well as increasing the
number of independent measurements. For that purpose, the
modification of the sensing matrix via various AUT positions
[201] or different measurement surface scans, or the acquisi-

30
RoE Journal Antenna Measurement Challenges and Opportunities - 31/49

Implantable Antenna
Characterization
Sergio Castelló-Palacios1 , Anja K. Skrivervik2 *
1 Institute
of Telecommunications and Multimedia Applica-
tions (iTEAM), Universitat Politècnica de València (UPV),
Valencia, Spain
2 Microwave and Antenna Group (MAG), Ecole Polytechnqiue

Fédérale de Lausanne (EPFL), Lausanne, Switzerland

*Corresponding author: anja.skrivervik@epfl.ch

Introduction. The characterization of implantable antennas

comprises many challenges, as their final place of action is Figure 27: Setup for the measurement of the performance of an
within a living host body [204, 205]. It is thus not possible for on-body device with an arm phantom.
evident practical reasons to measure antenna performances in
their final setup, in which only system aspects can be determined
(e.g. the success of a remote powering or telemetric link). The to be adapted to the type of guiding structure (differential or
knowledge of the characteristics of the antenna are nevertheless signal to ground) that will be used to connect it to measure-
important in the design and validation phase of an implant, as ment equipment, which is very hard to achieve for electrically
it is of utmost importance to ascertain that the latter work well small structure. As a result, in the typical case where a coax-
before implanting it into a living host. ial cable is used for measuring the antenna, a spurious current
This issue is addressed by the use of simulation models at circulates on the mantle of cable. In the case of an antenna
different stages of the creation implantable antennas, which are radiating into free space, the effect of this current will mainly
known as phantoms. In the first stages of the antenna devel- affect the measurement of the radiation characteristics (gain,
opment, simulation models or virtual phantoms are included efficiency and pattern). The antenna’s reflection coefficient or
within the software to optimize its characteristics. The purpose input impedance measurement will suffer less, as the chassis of
is to replicate the environment the antenna is expected to work the measurement equipment will absorb the spurious current. In
in by means of adding voxel models with the electromagnetic the case of an antenna implanted in a lossy phantom, there will
properties of the involved tissues. This feature is essentially potentially also be a major effect of the spurious cable current
solved and it is a matter of resolution and computing time what on the measurement of the input impedance or reflection coef-
affects the quality of the simulations. After manufacturing the ficient: Indeed, the medium surrounding the cable will short
antenna, its validation must be carried out in the real scenario circuit the spurious current, and reflect it back to the antenna.
or in a realistic one instead. Physical or experimental phan- This may severely affect the measured values [217].
toms emerge at this point when the measurements in the real Emerging challenges. An important challenge in the char-
environment is not possible. There are three main aspects of acterization of Implantable antennas is that there is today no
the communication link that can be assessed thereby: the radi- clear consensus in the community on the key performance indi-
ation characteristics of the antenna, i.e., pattern and efficiency cators (KPI) of such antennas: Indeed, for antennas radiating
[206, 207]; the impact of the body over the radio link in the into free space, it is general accepted that an antenna can be
communications channel, i.e., the path loss [208, 209]; and the described e.g. by its input impedance (including bandwidth),
specific absorption rate (SAR) [210, 211] of the signals on the gain pattern, polarization and radiation efficiency. All these
body tissues, which must meet the ICNIRP guidelines [212]. KPIs enable to link the antenna to a more complex system. A
Fig. 27 shows a setup for measuring the radiation pattern and simple illustration is given by the Friis’ formula giving the link
efficiency of an on-body antenna with an arm phantom in an budget between a transceiver and a receiver:
anechoic chamber.
Beside the issue of the host body, implantable antennas are
λ 2
 
usually physically small, and thus often small with respect to   
the wavelength, which leads to notorious difficulties in measure- Pr = Ptr gr gtr 1 − |Γr |2 1 − |Γtr |2 χ pol
2
(7)
4πd
ments [213, 214, 215, 216]. The problem in measuring elec-
trically small antennas are the so-called cable currents, which where P represents the power, g the antenna gain, Γ the antenna
are at the origin of spurious unwanted radiation. In order to input reflection coefficient, χ pol the polarization mismatch, λ
understand the origin of this difficulty, remember that at low the wavelength and d the distance between antennas. The sub-
frequency components are connected using wires, the latter scripts r and tr denote the receiver and transceiver, respectively.
having negligible dimensions compared to the wavelength. At All the terms in this equation except for the last are linked to
microwave frequencies, components are interconnected using antenna characteristics, while the last is linked to the channel
transmission lines having non-negligible dimensions compared (in this case free space) and its effect on the losses. Everything
to the wavelength. A component (e.g. an antenna) needs thus becomes much more complex when implantable antennas and

31
RoE Journal Antenna Measurement Challenges and Opportunities - 32/49

systems are considered, as it becomes difficult to decouple the Future developments to satisfy these challenges. Ideally,
antennas from the channel. Indeed, let us consider the scenario future developments would provide means to decouple the an-
of Fig. 28: tenna characteristics from the host body, at least to some extent,
in order to obtain a well-defined interface between antenna and
propagation channel, as exists in free space. This would enable
to separate the modeling and measurement of in-body channel
propagation on one hand and of antennas on the other. The
main difficulty to overcome is linked to the coupling between
the reactive near field surrounding the antenna and the host
body, which generates a non-negligible part of the losses [217].
Those losses depend of course on the antenna and the surround-
ing medium, but also on the depth of implantation. First models
to compute these losses have been proposed in literature for the
case of deep implants [217, 219, 220]. In latter case indeed, we
can suppose that near field zone around the antenna is entirely
Figure 28: Implanted antennas scenario.
located inside the body. In [217], a close form approximation
for the near field losses for implanted loops and dipoles, and
The link budget between the implanted and external nodes in [219] the authors propose the very interesting concept of
depends on many factors (antenna characteristics, in body prop- intrinsic radiation efficiency to account for the near field losses.
agation, out of body propagation, losses) and it is very difficult This work needs to be extended to case of shallow implants,
to decouple the internal antenna from the channel. Indeed, the where the antenna near field area reaches outside the host-body,
gain, efficiency and polarization of the implanted antenna de- and to more complex antennas. Moreover, models for the reflec-
pend not only on the antenna itself, but also on the host body. tion at the body-free-space interface also need to be investigated.
One way to solve this issue is to define the set formed by the As the studies pursue to be as realistic as possible and un-
antenna and the host body as being the actual antenna, and derstanding the phenomena of the electric fields inside the body,
define the gain, efficiency and polarization of this, from which phantoms are evolving to multilayer models [221, 222]. Most
a link budget to the off-body node can be obtained. Unfortu- works are limited to the use basic constituents due to the avail-
nately, this is not satisfactory as new antenna characteristics ability and ease of preparation. However, different chemicals
need to be obtained for each new position in the host body, or can improve the model versatility to be adjusted to different
each shape and dimension of the latter. Comparison between tissues [218]. Creating the shape of the tissues is currently
different antennas become thus very difficult. fostered thanks to the 3D printing capabilities. In this sense,
From the experimental assessment side, the main challenge authors are taking advantage of the wide range of materials
is to replicate the electromagnetic properties of the different currently available for creating phantoms [223]. The printed
body tissues within the whole frequency band of the transmis- models act as shells for the liquid or semisolid phantoms. This
sion range of the antenna. The way to achieve this goal is approach deals with two problems, the shape of the tissues,
matching the relaxation frequency of the target, which is related which is certainly key in the scenario, and the lifetime enlarge-
to how the molecules can keep polarizing at the same rate as ment due to the enclosure. However, there are two challenges
the electric field. This applies to the relative permittivity (εr ), here to be addressed: the influence of this external shell in the
since relative permeability (µr ) of body tissues is approximately effective relative permittivity and the chemical compatibility
equal to 1. This was not a problem in the past due to the narrow of the printed material and the phantom composition. The first
bandwidths that were used, enough for covering the needs at issue is usually resolved by using materials with low relative
the time. However, the current trend of rising the data rate and permittivity values (mostly polymers), and the second one has
cutting down the latency are moving the technologies to the not been comprehensively targeted due to the aforementioned
use of higher bandwidths. In [218], a collection of wideband basic ingredients. As a wider variety of reagents are used, their
phantoms is presented and further on reproduced in [209] for chemical compatibility with the surrounding materials will be
channel modeling. Another concern from the scientific com- analyzed in terms of degradation.
munity is the phantoms lifetime, considering that these take
Conclusion. The characterization of implanted antennas is
time for preparing them or are costly, if purchased. It is mainly
far from trivial, and many questions still remain open. Most
related to the way of storing them and the type of material used
critically, a consensus of relevant KPIs and measurement envi-
in the phantom elaboration. Solids are normally more durable
ronment is needed by the community. The phantoms will play
in time without losing properties, whereas liquids and gels re-
an essential role in the validation of these antennas when these
quire careful storage. One could think that then, solid phantoms
need to be tested in realistic scenarios.
are more convenient, but most tissues contain a considerable
amount of water, which normally request liquids to imitate their Acknowledgment. The authors are grateful for the fruitful
properties. Once these issues are settled, the challenge lies on discussions in the CA20120 (INTERACT) COST Action meet-
the mechanical arrangement of different phantoms for the setup. ings.
This can lead to liquid diffusion between different layers that
change the dielectric properties of the different tissues.

32
 RoE Journal Antenna Measurement Challenges and Opportunities - 33/49

New Measurement Ranges Combined Ranges. The future of antenna range design offers
a combination of two or more of these technologies into a single
range. About 10 years ago, a tapered range with spherical and
and Instrumentation planar near field capabilities was commissioned (Fig. 29).

Vince Rodriguez1 *, Bruce Williams2

1 NSI-MI Technologies, Suwanee, Georgia, USA
2 Orbital Outpost X, Mountain View, California, USA
*Corresponding author: vince.rodriguez@ametek.com

Introduction. In this paper the authors explore the needs that

are evolving for both the antenna range or anechoic chamber
and the instrumentation required for measuring newer antennas.

Antenna Ranges. Probably the main issue affecting antenna Figure 29: A combined tapered, SNF and PNF range seen
measurement ranges is that there is no universal solution. It is from the SNF probe location. A dielectric lens is placed at the
true that you can measure any antenna in a far field configura- end of the taper to increase the quiet zone size.
tion. However, this will require in some cases a test range more
than 1 km long. Take a 40 λ size antenna operating at 1 GHz. In addition, a variation of the compact range where a dielec-
The lower limit of the far field region is 1600 λ or about 480 tric lens is used instead of a mirror to create the plane wave, was
m [55]. Clearly an outdoor range will be required, however the introduced to improve the tapered range at higher frequencies
path loss will be extremely high, requiring higher transmitted [227]. Since then, it has become common to see combinations
power. That higher power comes with the regulatory issues of compact range and near field, or compact range and far field
that accompany high-power broadcasting at certain frequencies. (Fig. 30), where the ranges share the same AUT (antenna under
For the last 60 years, the world of antenna measurements has test) positioner.
orbited around the problem of measuring larger antennas in
smaller ranges. Near to far field transforms and compact range
development began in the 1960s [224, 225] to contend with the
longer far field distances of electrically large antennas. On the
other side of the spectrum are the small, lower frequency an-
tennas, that while not electrically large still require long ranges.
For example, a 0.5 λ antenna at 200 MHz requires at least a 5 λ
distance, or 7.5 m [76]. This is doable in a typical indoor range,
but it requires a very large room lined with lots of expensive
absorber. To solve these problems, tapered anechoic chambers
were introduced in the 1960s as well. All these methodologies
Figure 30: A compact range (left) and far field range (right)
are applicable for certain antennas and frequency ranges. Com-
share the same AUT positioner. This combined range extends
pact Ranges are ideal in the 2 GHz to 100 GHz range. At lower
the lowest useful test frequency down to 500 MHz.
frequencies the size of the reflector becomes the main driver
for the cost. At frequencies above 40 GHz, the tolerance of the
These ranges are more flexible as they allow for testing a
reflector finish is the driver. Near-Field Systems can be used
variety of antennas, at a variety of frequencies, with smaller
from lower frequencies, as low as 200 MHz, and as high at
uncertainties since the most suitable approach can be used for
the 500 GHz range. At lower frequencies the AUT-to-probe
the specific antenna under test. Future ranges will certainly
separation and the size of the probe still require a large facility,
follow this path of combining technologies.
and at higher frequencies positioning accuracy and stable phase
measurement are the critical issues. Common near-to-far-field Instrumentation. The modern antenna range has become far
transforms require accurate phase measurements, and although more complex, especially in the past two decades. The simple
“phaseless” transform algorithms have been studied, they are range of the past is inadequate for today’s antennas that are no
not in a mature stage [76]. The phase reference issue is a big longer separable from the electronics behind them. More so-
problem for testing some integrated antennas such as active phisticated tests are needed to characterize these combinations
arrays with digital outputs and small wireless devices. Tapered of antennas, amplifiers, switches, and frequency converters.
ranges are ideal for lower frequencies (100 MHz to 2 GHz), The antenna range has now become a “subsystem test range,”
with the lowest frequency being limited mainly by RF absorber providing capabilities far beyond simple pattern and gain mea-
technology [226]. Although tapered ranges have been used up surements. Similarly, the AUT has now become the UUT (Unit
to 40 GHz, they have a limit on the electrical antenna size with Under Test) to address the wide range of functions and require-
feed performance and location becoming more critical with ments that can be imagined for devices that integrate electro-
rising frequency [226]. magnetics, mechanics, electronics, firmware, software and other

33
RoE Journal Antenna Measurement Challenges and Opportunities - 34/49

technologies. These integrated antenna subsystems can take on Agile Steerable Antennas. New applications for RF and
any number of functions depending on the supported product microwave signals, including Wi-Fi, 5G, IoT, and advanced
which may be a communicator, transponder, radar, or jammer. weather radar, have brought about new growth in low-cost but
Range instrumentation now must support a balance of compet- very capable steered-beam antenna technologies. These agile
ing objectives. Some of these are defined by the UUT: wideband antennas can sequentially or even simultaneously form a multi-
operation, higher test frequencies, frequency-converting and tude of spatially diversified RF pathways. This offers system
non-converting AUTs, high-power RF radiation, pulse mode designers a way to optimally use limited bandwidth and power
operation, and interleaved transmit/receive requirements [228]. to service an ever-increasing number of locations, devices, or
Others are application-driven: a need for accurate pattern, gain, users. At the antenna range, of course, there is a cost in the com-
EIRP, and G/T results based on measured data [229]. Yet others plexity. If the antenna can form sixty-four (or six hundred forty)
are driven by complexity, cost, and risk considerations: all- distinct beams, the range technician will likely need to test all of
at-once acquisitions incorporating multi-frequency, multi-port, them. The current explosion of fast control and reconfiguration
dual-pol, and multi-state measurements, and the need to collect capabilities presents challenges: test coverage, data volume,
all these measurements in the least amount of time. While this measurement timing and electromechanical coordination [231]]
level of complexity was once the domain of only the leading- (Fig. 32).
edge firms, it has become common across the middle of the
industry as the pace of antenna system development increases
alongside innovations in electronics, computation, and RF tech-
nology.
Frequency Converting Antennas. It has now become com-
mon to see frequency conversion as an inseparable function
of many UUT. For example, inspection of a home satellite
dish feed assembly (Fig. 31) reveals an integral “block down-
converter” that converts Ku- or Ka-band frequencies down to
around 1 GHz to send to a receiver. The downconverter includes
its own internal Local Oscillator (LO), mixer, and filters.

Figure 32: Typical timing sequence for a steerable antenna test

using an AUT-specific control sequencer with a standard VNA
for the related RF measurements.

Control schemes for agile antennas vary considerably; there

are no standard methods, signals, or interfaces. This is likely to
continue as differing requirements drive antenna designers to
differing solutions. The future for agile-antenna range testing
lies in adapting each of the various control schemes to a well-
defined “in-between” interface, a task best done by the AUT
control designers. The standard interface, published by the test
Figure 31: An integrated feedhorn and low-noise block system designers, gives the range unfettered access to AUT
downconverter (FLNB) used for satellite television reception. capabilities for testing without requiring intimate knowledge of
the internal control scheme. Putting the control scheme behind
Simple as it is, this antenna-plus-converter offers some chal- a standard interface has the added benefit of protecting the AUT
lenge to the range instrument designer. Small dish antennas designer’s work which is often proprietary or confidential.
typically used with these feeds are ideal candidates for test-
Conclusion. Flexibility is going to be the name of the game
ing on a planar near field range, but near field testing requires
moving forward when it comes to range design. If the antenna
amplitude and phase information to calculate far-field perfor-
engineer cannot accept additional uncertainty or error when
mance. The internal LO in this unit poses a challenge: how
using non-ideal techniques then the go-to approach will be to
to get coherent phase readings given internal LO drift. Addi-
build ranges that can support several different techniques: far
tionally, assuming the phase drift problem has been resolved,
field, tapered range, compact range, or near field measurements.
the challenge is how can it be measured the G/T (Gain-over-
On the instrumentation side, the equipment, knowledge, and
Temperature) performance for such an antenna without access
techniques needed to make increasingly complex measurements
to the antenna’s RF output port [230]. There are plenty of
of agile multi-state and multi-function antennas will be key to
more-complex antenna systems that present similar challenges.
staying relevant.
Integration of inseparable amplifiers, frequency converters, and
even digitizers is accelerating in the satellite, military, scientific,
and even consumer product spaces.

34
RoE Journal Antenna Measurement Challenges and Opportunities - 35/49

EurAAP and IEEE ments,” originally published in 2012 [76]. This latter standard,
dedicated to near-field measurement methods, was created to
complement the existing IEEE Std 149 and is currently under
Standardization and revision by a working group consisting of members of AMTA,
WG5, and IEEE [235]. Planned publication is in early 2024.
Facility Comparison
Lucia Scialacqua1 *, Jeff Fordham2,3
1 Microwave Vision Italy, Pomezia, Italy
2 NSI-MITechnologies, Suwanee, Georgia
3 IEEE Standard Association

*Corresponding author: lucia.scialacqua@mvg-world.com

Introduction. Antenna measurements represent a crucial phase

during the process of design, development, and verification of
any antenna system to ensure the compliance with all speci-
fications. In recent years, emergence and adaptation of new
complex antenna technologies has led to specifications even
more stringent to comply. This drives continuous research on
new measurement technologies to guarantee a required specific
accuracy during testing.
Of importance to characterization of antennas are: a consis- Figure 33: Illustration of primary and secondary field
tent and reliable methodology of testing, and assurances of the illumination in a compact antenna test range.
quality of the measurement system used. Confidence in these
two areas is assured by utilizing facilities that follow a set of The motivation of the activities of maintenance of these stan-
clear standards backed by an accredited quality process. dards is IEEE-SA policies specifying that a standard is valid
for ten years after its approval. A standard cannot be simply
Standardization. The Institute of Electrical and Electronics reaffirmed, even if the standard is still considered relevant. The
Engineers (IEEE), through its Standards Association (SA) has policies specify that a working group must review and revise
published many standards governing topics associated with an- it, or the standard will expire. There are more than 50 dedi-
tennas and antenna measurements. The IEEE SA is supported cated people from industry, academia, and institutions, who are
in the antenna technology field by the Antenna and Propagation members of the aforementioned working groups, contributing
Society (APS) Standards Committee (SC). This committee is to update and revision.
charged with the development and maintenance of standards
related to antennas and radar cross section. The standardiza- Facility Comparison. An important support is provided by
tion of procedures and parameters in antenna measurements is the outcomes of measurement campaigns aimed at the inter-
crucial. The efforts of the IEEE are accomplished in collabo- comparison of antenna measurements since 2005, performed
ration with the work of global working groups consisting of by EurAAP WG5 [236]. Facilities participating to such cam-
members from the IEEE APS-SC, the Antenna Measurement paigns have the possibility to demonstrate their measurement
Techniques Association (AMTA), and the European Associa- proficiency both for internal use and to get or maintain official
tion on Antennas and Propagation (EurAAP) working group accreditations, to standards such as ISO 17025 [237]. Intercom-
on measurements (WG5). Two standards addressing the def- parison campaigns can be extremely useful in the estimation of
inition of terms associated with antennas are: IEEE Std 211 measurement uncertainties that characterize a measurement en-
[232], and IEEE Std 145 [233]. Of these IEEE Std 211 was vironment and method [55, 76]. This can be achieved thanks to
most recently updated in 2018 and IEEE Std 145 is undergoing the measurement of highly accurate reference antennas. During
revision with a target publication date in early 2024. IEEE Std the proposed period of facility comparison campaigns, the se-
149 “Recommended Practice for Antenna Measurements” [55], lected reference antenna travel among the participating facilities
is the primary standard covering most antenna measurement whereby the data is collected. Upon completion of the activity
techniques. This standard was most recently updated and pub- the data is analysed and made available to the participants and
lished in 2022 [234] by a working group consisting of members usually targeting to publish the outcomes in peer-reviewed jour-
from the IEEE, the AMTA, and WG5. The previous version of nals. Some of the reference antennas tested in the comparison
the standard was released in 1979 and many updates, and mod- campaigns are illustrated in Fig. 34. The measurement post-
ern measurement methods were added to the standard, such as processing of the intercomparison campaigns, for the linear
reverberation chamber measurements, RF measurements using array, horn, and reflector antenna shown in Fig. 34, consists
drones, sensitivity measurements for receive antenna systems in the computation of a reference pattern and an associated
and compact range techniques as illustrated in Fig. 33. Also Equivalent Noise Level (ENL) as reported in [238]. An exam-
important to the field of antenna measurements is IEEE Std ple of multiple pattern acquisition is shown in Fig. 35. The
1720 “Recommended Practice for Near-Field Antenna Measure- reference pattern is computed from several independent mea-

35
RoE Journal Antenna Measurement Challenges and Opportunities - 36/49

surements by the different participants and its correlation with the input reflection coefficient at the waveguide flange; the co-
each measurement is expressed through the ENL. and cross-polar radiation patterns in different planes and the
forward hemisphere; the direction of the maximum co-polar
pattern; and the 1σ -uncertainty of directivity and gain data. Pre-
liminary results have been presented during the 2022 European
Conference on Antennas and Propagation (EuCAP) [247]. The
post-processing and analysis of all data has been finalized and
the new updated results will be presented afterwards in relevant
conferences and journals.
Emerging Challenges. A major challenge for comparison
campaigns is to introduce increasingly complex antennas as
test objects that are used in the latest and most modern tech-
nologies/applications. Measuring at increasingly challenging
frequency bands attracts new participants of the campaigns
who want to test the capabilities of their measurement facilities.
Moreover, increasing the number of participants in measure-
ment campaigns means increasing the variety of the type of
various test systems. This can lead to an enrichment of knowl-
edge in antenna measurement procedures, a benefit for the entire
community. In parallel, it is important to always keep updated
the defined standards on antenna measurements due to the de-
velopments of new emerging technologies and applications.
Future developments to satisfy these challenges. Future
activities on comparison campaigns consist of proposing new
Figure 34: Examples of reference antennas measured during challenging antennas. The selection of a low-directivity antenna
recent intercomparison campaigns. would be a new challenge to meet the measurement needs of
increasingly less directional antennas, such as in the automotive
Additional figures of merit that can enhance the comparison industry. A future challenge is also to create a stronger bridge
are the Birge Ratio and Escore as reported in [9]. The results of scientific collaboration between the community of antenna
of the above described intercomparison campaigns have been measurements and that one operating with numerical simula-
published in peer-reviewed journals, see references [238, 239, tions. One way is to broaden comparison campaigns not only to
240, 241, 242, 243]. measured data, but also to compare the measured reference with
simulations made with different numerical methods. Experts on
numerical simulations can benefit from the experience of all the
long standardization work of so many years on measurements,
while experts on antenna measurements will be more confident
of the results by simulation tools that were used to make the
design of antennas, then manufactured and to be tested. In
support of the need for less directional intercomparison testing
and comparisons to simulation, the APS-SC is currently un-
dergoing feasibility studies of bi-conical antennas. This work
is being done under an IEEE Project Authorization Request
(PAR) P2816 “Recommended Practice for Computational Elec-
tromagnetics Applied to Modeling and Simulation of Antennas”
[248]. Initial antennas have been prototyped by the APS-SC Ad
Figure 35: Example of comparison of radiation patterns, a hoc Group on Antenna Measurements and are in preliminary
reflector fed by a broadband horn in Fig. 34. stages of testing. Publication of results are planned for future
conferences. Activities to define and maintain standards on an-
Another relevant comparison campaign has been carried tenna measurements will always be continuous in order that the
during 2019-2022, with the mm-wave Validation Standard an- community can benefit up-to-date reference to new emerging
tenna (mm-VAST) shown in Fig. 33. The mm-VAST antenna technologies and applications.
has been developed by DTU and TICRA in the framework of
Acknowledgment. The authors would like to thank all the
an ESA project [244, 64, 245, 246]. The measurements have
organizations for spending effort to contribute to the different
been conducted in three operational configurations: 19.76 GHz,
activities on antenna measurements and to EurAAP for support
37.80 GHz, and 48.16 GHz. The objective is to ensure ac-
in keeping existing this working group.
curate measurements of the next generation communication
antennas in the bands K, Ka, Q, and V. Collected data comprise

36
RoE Journal Antenna Measurement Challenges and Opportunities - 37/49

Advanced Antenna Material algorithms and standards are necessary for system calibration.
Unfortunately, these standards are not readily available. In
addition, due to the increased number of unknowns for these
Measurements materials, calibration and measurement time increases dramati-
cally. The challenges are compounded if measurements in the
Michael Havrilla*
field or in remote locations are necessary.
Department of Electrical and Computer Engineering, Air
Materials exhibiting strong spatial dispersion present many
Force Institute of Technology, Wright-Patterson Air Force
Base, USA4
challenges as well (weak spatial dispersion is accommodated
*Corresponding author: michael.havrilla@afit.edu more easily [263]). Spatially dispersive media have mate-
rial properties that are dependent upon the direction/angle in
which an electromagnetic wave is impinging and constitutive
Introduction. Recent advances in fabrication techniques (e.g., relations involving spatial derivatives or integrals. As antici-
3D printing [249]) has allowed ever-exotic materials to be con- pated, Maxwell’s equations are more difficult to solve, although
structed for electromagnetic and antenna applications. Materi- Fourier transforms may offer some advantages. The key chal-
als that are nonlinear, bianisotropic, spatially dispersive, time lenge is to develop an appropriate theoretical material model
varying, and nonreciprocal, for example, show promise for and to acquire a measurement system capable of multi-angle
enhancing electromagnetic control and antenna performance measurements. As in the nonisotropic case, calibration and mea-
[250, 251, 252, 253, 254, 255, 256, 257, 258, 259, 260]. surement time also increases, which can challenge budgetary
In order to understand how electromagnetic waves inter- and allocation resources.
act with these complex media, material property tensors must
Time-varying media have constitutive relations that are typ-
be quantitatively determined. Although electromagnetic ma-
ically not amenable to Fourier transformation, thus analysis,
terial properties may, in some cases, be predicted via compu-
modeling, and measurement must be performed directly in the
tational modeling or by utilizing a lumped-parameter model,
time domain. In addition, spacetime modulated media can have
experimental measurements are most often employed for the
material properties that change rapidly over time. As a result,
material characterization and validation process [261]. Elec-
the traditional swept-frequency network analyzer (NWA) mea-
tromagnetic characterization of simple (i.e., linear, isotropic,
surements are rendered practically useless. The challenges are
homogeneous, and time-invariant) media is well established
compounded if spacetime modulation is extremely rapid, since
[262]. However, characterization of the types of complex me-
these materials are Lorentz-transformed into generally bian-
dia mentioned above, requires advanced material measurement
isotropic media relative to an observer in a rest frame [258, 259].
techniques and subsequently creates many challenges, which
Another challenge is the electromagnetic characterization of
are discussed next.
materials having inherent irregular shapes (e.g., molded materi-
Emerging Challenges. The electromagnetic measurement als for antenna applications). What experimental measurement
and characterization of complex media presents many chal- system is most appropriate in this case and what method of
lenges that are not encountered in simple media. In nonlinear theoretical analysis should be used? Although many challenges
media, harmonic generation (and nonreciprocal behavior) typ- exist for characterizing modern materials, many technological
ically occurs. Due to this nonlinear behavior, the traditional advances are being made that are capable of accommodating
linear S-parameters (scattering parameters) cannot be employed these challenges, as is discussed in the next section.
in the material extraction process. It can be debated if it even
makes sense to assign electromagnetic properties to such a Advances in Technology to Meet Challenges. Many techno-
material and whether it would be of use in practice. Equally logical, as well as manufacturing, advances have been made in
challenging is the determination of the conditions of unique- recent years that make it possible to meet the challenges of the
ness in the material extraction process. Perhaps even more electromagnetic measurement and characterization of modern
challenging is developing a proper nonlinear model for a given materials.
material. Almost all modern network analyzers now have nonlinear
The measurement of anisotropic and bianisotropic media measurement capability. Although many nonlinear models exist,
also presents many challenges, both theoretically and experi- X-parameters have become relatively well known [264]. These
mentally. Maxwell’s equations can become quite complicated parameters make it possible to measure harmonics of the excita-
and difficult to solve, especially if a Green’s function develop- tion frequency, thus making it a possible avenue to characterize
ment is needed, which is often the case when performing non- nonlinear materials.
destructive evaluation. The number of measurements required Several advances have been made that make the fabrication
increases (for example, characterization of an anisotropic biax- and measurement of anisotropic and bianisotropic media feasi-
ial material requires six measurements), thus field applicators ble. First, the advent of 3D printing, as mentioned previously,
having sufficient diversity are required. In addition, optical has now been utilized for fabrication of complex media. As 3D
activity or Faraday rotation effects occur, thus an experimen- printing capabilities increase, such as 3D printing of metals and
tal system must have co- and cross-polarization measurement magnetic media, ever-more exotic materials will be realized. In
capability. Consequently, known co- and cross-polarization addition, these 3D printing technologies now make it easier to
rapidly prototype co- and cross-polarization calibration stan-
4 U.S. Government Work. Not Protected by Copyright. dards. Regarding measurement, precision robots (supporting

37
RoE Journal Antenna Measurement Challenges and Opportunities - 38/49

quad-ridged horn antennas, for example) now make it possible terization is a critical technology that will undoubtedly enable
to obtain rapid and accurate multi-measurements. This data the invention of new material characterization probes and the
collection is assisted via many-port (e.g., 8-port, 16-port, etc.) measurement of antenna materials in ever-more complicated
network analyzers. In addition, these modern network analyzers environments.
have common and differential mode capabilities, thus enhancing Although the emphasis above is on technological advances
measurement diversity. Portable NWA’s now make it possible to that have enabled new advanced measurement techniques, theo-
perform these measurements in the field or in remote locations. retical developments have also aided in accommodating nonsim-
Multi-axis hot wire cutting machines allow for the fabrication of ple media. For example, scalar potential formulations have lead
new diverse measurement systems (see, for example, [261] and to more compact electromagnetic field representations, lead-
Fig. 36). Electronic calibration kits, available from most NWA ing to computationally efficient techniques for characterizing
vendors, reduce measurement time, thereby meeting budgetary nonisotropic media [266]. Future advances in theoretical elec-
and allocation challenges. tromagnetics will likely continue to aid in the development of
novel advanced material measurement capabilities.
Conclusion. Interest in advanced materials has flourished in
recent years, being enabled by manufacturing and fabrication
capabilities as well as the desire to have more control over the
electromagnetic field. These modern materials are being inte-
grated in many antenna systems to enhance performance metrics.
Advanced material measurements play (and will continue to
play) a vital role in antenna system modeling and prediction of
operational capabilities. Although many challenges remain in
the characterization of advanced materials, these challenges are
being met by recent technological advances.
Acknowledgment. The views expressed in this article are
those of the author and do not reflect the official policy or
position of the United States Air Force, Department of Defense,
or the U.S. Government.

Figure 36: Rectangular-to-Square Waveguide Biaxial Material

Characterization System.

The robotic technology mentioned above can also aid in

the characterization of spatially-dispersive media since multi-
angle measurements can be obtained in a relatively straight-
forward manner. Improvements in the generation of fast pulse-
generation technology provides a path to enable the charac-
terization of modern spacetime metamaterials. In addition,
improvements in computer resources and precision scanning
technology have made it possible to scan irregular-shaped mate-
rials and perform material property extraction via computational
electromagnetics [265]. Indeed, computational material charac-

38
RoE Journal Antenna Measurement Challenges and Opportunities - 39/49

References [13] S. F. Gregson, Z. Qin, C. G. Parini, “Compressive sensing

in massive MIMO array testing: a practical guide,” IEEE
[1] A. D. Yaghjian, “An overview of near-field antenna mea-
Transactions on Antennas and Propagation, Vol. 70, No. 9,
surements,” IEEE Transactions on Antennas and Propa-
pp. 7978–7988, Sep. 2022.
gation, Vol. 34, No. 1, pp. 30–45, Jan. 1986.
[14] R. Solimene, M. A. Maisto, and R. Pierri, “Sampling
[2] J. E. Hansen, Spherical Near-field Antenna Measurements,
approach for singular system computation of a radiation
London, UK: Peter Peregrinus Ltd., 1988.
operator,” Journal of the Optical Society of America A,
[3] C. Parini, S. F. Gregson, J. McCormick, D. Janse van Vol. 36, pp. 353–361, Mar. 2019.
Rensburg, and T. F. Eibert, Theory and Practice of Modern [15] J. Kornprobst, A. Paulus, J. Knapp, and T. F. Eibert,
Antenna Range Measurements, Vol. 1 and 2, London, UK: “Phase retrieval for partially coherent observations,” IEEE
IET SciTec Publishing, 2020. Transactions on Signal Processing, Vol. 69, pp. 1394–
[4] O. M. Bucci and C. Gennarelli, “Application of nonredun- 1406, 2021.
dant sampling representations of electromagnetic fields to [16] C. G. Parini, S. F. Gregson, A. K. Brown, “Untethered
NF-FF transformation techniques,” International Journal near-field drone based antenna measurement system for
on Antennas and Propagation, Vol. 2012, Art. No. 319856, microwave frequencies using multiple reference antennas
pp. 1–14, 2012. for phase and drone location recovery,” IET Microwaves,
[5] P. Petre and T. K. Sarkar, “Planar near-field to far-field Antennas and Propagation, Vol. 16, No. 13, pp. 798–811,
transformation using an equivalent magnetic current ap- 2022.
proach,” IEEE Transactions on Antennas and Propagation, [17] A. Paulus and T. F. Eibert, “Near-field far-field transforma-
Vol. 40, No. 11, pp. 1348–1356, Nov. 1992. tions with unknown probe antennas,” TechRxiv. Preprint,
[6] Y. Alvarez, F. Las-Heras, and M. R. Pino, “Reconstruction https://doi.org/10.36227/ techrxiv.21542082.v1, 2022.
of equivalent currents distribution over arbitrary three- [18] D. Ostrzyharczik, J. Knapp, J. Kornprobst, and T. F. Eibert,
dimensional surfaces based on integral equation algo- “Inverse source solutions with Huygens-surface conform-
rithms,” IEEE Transactions on Antennas and Propagation, ing distributed directive spherical harmonics expansions,”
Vol. 55, No. 12, pp. 3460–3468, Dec. 2007. IEEE Transactions on Antennas and Propagation, Vol. 71,
No. 2, pp. 1684–1696, Feb. 2023.
[7] T. F. Eibert and C. H. Schmidt, “Multilevel fast multi-
pole accelerated inverse equivalent current method em- [19] T. F. Eibert, “Multiple-frequency preconditioned itera-
ploying Rao-Wilton-Glisson discretization of electric and tive inverse source solutions,” IEEE Transactions on Mi-
magnetic surface currents,” IEEE Transactions on An- crowave Theory and Techniques, Vol. 71, No. 7 pp. 2842-
tennas and Propagation, Vol. 57, No. 4, pp. 1178–1185, 2853, 2023.
Apr. 2009. [20] O. M. Bucci, C. Gennarelli and C. Savarese, “Representa-
[8] J. L. Araque Quijano and G. Vecchi, “Field and source tion of electromagnetic fields over arbitrary surfaces by a
equivalence in source reconstruction on 3D surfaces,” finite and nonredundant number of samples,”IEEE Trans-
Progress in Electromagnetics Research, Vol. 103, pp. 67– actions on Antennas and Propagation, Vol. 46, No. 3, pp.
100, 2010. 351-359, Mar. 1998.
[9] J. Kornprobst, A. Paulus, T. F. Eibert, R. A. M. Mauer- [21] C. Gennarelli, F. Ferrara, R. Guerriero, and F. D’Agostino,
mayer, “Method-of-moments modeling of conducting ob- “Non-redundant near-field to far-field transformation tech-
jects within the fast irregular antenna field transformation niques”, Scitech Pub., 2023.
algorithm,” 41th Annual Symposium of the Antenna Mea- [22] A. Capozzoli, C. Curcio, A. Liseno and P. Vinetti, “Field
surement Techniques Association (AMTA), San Diego, CA, sampling and field reconstruction: a new perspective”,
Oct. 2019. Radio Science, Vol. 45, No.6, pp. 1-31, Dec. 2010.
[10] J. Knapp and T. F. Eibert, “Near-field antenna pattern [23] A. Capozzoli, C. Curcio, and A. Liseno, “On the Optimal
measurements in highly-reflective environments,” IEEE Field Sensing in Near-Field Characterization”, Sensors,
Transactions on Antennas and Propagation, Vol. 67, No. 9, Vol. 21, No. 13, 4460, pp. 1-18.
pp. 6159–6169, Sep. 2019. [24] R. Cornelius and D. Heberling, “Analysis of Sam-
[11] M. D. Migliore, “A simple introduction to compressed pling Grids for Spherical Near-Field Antenna Measure-
sensing/sparse recovery with applications in antenna mea- ments,” Progress In Electromagnetics Research Sympo-
surements,” IEEE Antennas and Propagation Magazine, sium (PIERS), pp. 923–927, 2015.
Vol. 56, No. 2, pp. 14–26, Apr. 2014. [25] S. Pivnenko and T. Laitinen, ”Spherical near-field antenna
[12] B. Hofmann, O. Neitz, and T. F. Eibert, “On the mini- measurements with theta-scanning with reduced number
mum number of samples for sparse recovery in spherical of phi-steps,” 2009 Antenna Measurement Techniques As-
antenna near-field measurements,” IEEE Transactions on sociation Symposium (AMTA), pp. 383–388, 2009.
Antennas and Propagation, Vol. 67, No. 12, pp. 7597– [26] C. Culotta-López, D. Heberling, A. Bangun, A. Behboodi
7610, Dec. 2019. and R. Mathar, “A Compressed Sampling for Spherical

39
RoE Journal Antenna Measurement Challenges and Opportunities - 40/49

Near-Field Measurements,” in 2018 AMTA Proceedings, [38] L. Scialacqua, L. J. Foged, F. Mioc, and F. Saccardi, “Link
2018. between measurement and simulation applied to antenna
scattering and placement problems,” European Confer-
[27] A. Capozzoli, C. Curcio, G. D’Elia, A. Liseno, “Phaseless
ence on Antennas and Propagation, 2017.
antenna characterization by effective aperture field and
data representations”, IEEE Transactions on Antennas [39] B. Xu, Z. Ying, L. Scialacqua, A. Scannavini, L. J. Foged,
and Propagation, Vol. 57, No. 1, pp. 215-230, Jan. 2009. T. Bolin, K. Zhao, S. He, and M. Gustafsson, “Radiation
performance analysis of 28 GHz antennas integrated in
[28] J. A. Gordon, D. R. Novotny, M. H. Francis, R. C. 5G mobile terminal housing,” IEEE Access, Vol. 6, pp. 48
Wittmann, M. L. Butler, A. E. Curtin and J. R. Guerri- 088–48 101, 2018.
eri, ”Millimeter-Wave Near-Field Measurements Using
Coordinated Robotics,” IEEE Transactions on Antennas [40] C. Cappellin, A. Frandsen, and O. Breinbjerg, “Applica-
and Propagation, Vol. 63, pp. 5351-5362, 2015. tion of the swe-to-pwe antenna diagnostics technique to an
offset reflector antenna,” IEEE Antennas and Propagation
[29] R. Moch and D. Heberling, ”Robot-Based Antenna and Magazine, Vol. 50, No. 5, pp. 204–213, 2008.
Radar Measurement System at the RWTH Aachen Univer-
[41] K. Persson and M. Gustafsson, “Reconstruction of equiva-
sity,” 2020 Antenna Measurement Techniques Association
lent currents using a near-field data transformation - with
Symposium (AMTA), 2020.
radome applications,” Progress in Electromagnetics Re-
[30] A. Capozzoli, L. Celentano, C. Curcio, A. Liseno, S. search, Vol. 54, pp. 179–198, 2005.
Savarese, “Twice optimised near-field scanning system [42] J. L. A. Quijano and G. Vecchi, “Improved-accuracy
for antenna characterization,” IET Microwaves, Antennas source reconstruction on arbitrary 3-D surfaces,” IEEE
and Propagation, Vol. 14, No. 3, pp. 163-173, Feb. 2020. Antennas and Wireless Propagation Letters, Vol. 8, pp.
[31] R. Cornelius, T. Salmerón-Ruiz, F. Saccardi, L. Foged, 1046–1049, 2009.
D. Heberling and M. Sierra-Castañer, ”A comparison of [43] E. Jørgensen, D. W. Hess, P. Meincke, O. Borries, C.
different methods for fast single-cut near-to-far-field trans- Cappellin, and J. Fordham, “Antenna diagnostics on planar
formation”, IEEE Antennas and Propagation Magazine, arrays using a 3D source reconstruction technique and
Vol. 56, No. 2, pp. 252-261, April 2014. spherical near-field measurements,” Proceedings of The
[32] F. Rodrı́guez-Varela, B. Galocha-Iraguen, M. Sierra- European Conference on Antennas and Propagation, 2011,
Castaner, “Single-Cut Phaseless Near-Field Measurements pp. 2547–2550.
using Specialized Probes”, Proceedings of the AMTA Sym- [44] C. Cappellin, S. Pivnenko, E. Jørgensen, and P. Meincke,
posium, 2022. “Array diagnostics, spatial resolution and filtering of un-
desired radiation with the 3d reconstruction algorithm,”
[33] E. Jørgensen, P. Meincke, C. Cappellin, and M. Sabba-
Proceedings of the 35th ESA Antenna Workshop, Noord-
dini, “Improved source reconstruction technique for an-
wijk, The Netherlands, 2013.
tenna diagnostics,” Proceeedings of the 32nd ESA Antenna
Workshop, ESTEC, Noordwijk, The Netherlands, 2010. [45] J. Kornprobst, R. A. Mauermayer, O. Neitz, J. Knapp,
and T. F. Eibert, “On the solution of inverse equivalent
[34] L. Foged, L. Scialacqua, F. Saccardi, J. A. Quijano, G. surface-source problems,” Progress In Electromagnetics
Vecchi, and M. Sabbadini, “Practical application of the Research, Vol. 165, pp. 47–65, 2019.
equivalent source method as an antenna diagnostics tool
[46] O. Borries, M. H. Gæde, P. Meincke, A. Ericsson, E.
[amta corner],” IEEE Antennas and Propagation Maga-
Jørgensen, D. Schobert, and E. Gandini, “A fast source
zine, Vol. 54, No. 5, pp. 243–249, 2012.
reconstruction method for radiating structures on large
[35] E. Jørgensen, P. Meincke, and C. Cappellin, “Advanced scattering platforms,” Antenna Measurement Techniques
processing of measured fields using field reconstruction Association Symposium, 2021.
techniques,” Proceedings of European Conference on An- [47] A. Ericsson, O. Borries, M. H. Gæde, P. Meincke, E.
tennas and Propagation, 2011, pp. 3880–3884. Jørgensen, and E. Gandini, “Fast source reconstruction
[36] L. Ciorba, G. Virone, F. Paonessa, M. Righero, E. De of large reflector antennas for space applications,” IEEE
Lera Acedo, S. Matteoli, E. C. Beltran, P. Bolli, G. Gior- International Symposium on Antennas and Propagation,
danengo, G. Vecchi, A. Magro, R. Chiello, O. A. Pev- 2022.
erini, and G. Addamo, “Large horizontal near-field scanner [48] L. Scialacqua, F. Mioc, L. J. Foged, G. Giordanengo, M.
based on a non-tethered unmanned aerial vehicle,” IEEE Righero, and G. Vecchi, “Diagnostics on electrically large
Open Journal of Antennas and Propagation, Vol. 3, pp. structures by a nested skeletonization scheme enhance-
568–582, 2022. ment of the equivalent current technique,” European Con-
[37] J. L. Araque Quijano, L. Scialacqua, J. Zackrisson, L. ference on Antennas and Propagation, 2020.
J. Foged, M. Sabbadini, and G. Vecchi, “Suppression of [49] L. Scialacqua, F. Mioc, L. J. Foged, G. Giordanengo, M.
undesired radiated fields based on equivalent currents re- Righero, and G. Vecchi, “Equivalent source technique pro-
construction from measured data,” IEEE Antennas and cessing of broadband antenna measurements,” European
Wireless Propagation Letters, Vol. 10, pp. 314–317, 2011. Conference on Antennas and Propagation, 2020.

40
RoE Journal Antenna Measurement Challenges and Opportunities - 41/49

[50] L. Scialacqua, F. Scattone, A. Giacomini, L. Foged, and Propagation, Vol. 60, No. 8, pp. 3875-3891, Aug. 2012,
F. Mioc, “Advanced diagnostics on a large array by the doi: 10.1109/TAP.2012.2201125.
equivalent current technique,” 2021 Antenna Measurement [63] M. Alberica Saporetti, L. Foged, M. Sierra Castaner, S.
Techniques Association Symposium (AMTA), 2021. Pivnenko, R. Cornelius and D. Heberling, ”Description
[51] L. Scialacqua, F. Mioc, F. Scattone, L. J. Foged, G. Gior- and Results: Antenna Measurement Facility Comparisons
danengo, M. Righero, and G. Vecchi, “High performance [Measurements Corner]”, IEEE Antennas and Propaga-
processing with equivalent currents: Results on a very tion Magazine, Vol. 59, No. 3, pp. 108-116, Jun. 2017, doi:
large measured antenna,” European Conference on Anten- 10.1109/MAP.2017.2686097.
nas and Propagation, 2022. [64] S. Pivnenko, et al., “DTU-ESA millimeter-wave validation
[52] T. H. Loh (Editor), Metrology for 5G and Emerging Wire- standard antenna – requirements and design”, Proceedings
less Technologies, The Institution of Engineering and of the 36th Annual Symposium of the Antenna Measure-
Technology, Dec. 2021, doi: 10.1049/PBTE099E. ment Techniques Association (AMTA), Oct. 2014.
[53] A. C. Newell, “Error analysis techniques for planar near- [65] J. R. Guerrieri, et al., ‘International Comparison of Ku-
field measurements,” IEEE Transactions on Antennas and Band Standard Gain Horn Characterization’. EuCAP 2014,
Propagation, Vol. 36, No. 6, pp. 754–768, Jun. 1988, doi: Apr. 2014, doi: 10.1109/EuCAP.2014.6902369.
10.1109/8.1177. [66] S. Pivnenko, et al., “Comparison of Antenna Measure-
[54] P. Pelland, G. Hindman and A. Newell, “Advances in au- ment Facilities With the DTU-ESA 12 GHz Validation
tomated error assessment of spherical near-field antenna Standard Antenna Within the EU Antenna Centre of Ex-
measurements,” The 7th European Conference on Anten- cellence”, IEEE Transactions on Antennas and Propa-
nas and Propagation (EuCAP 2013), Gothenburg, Swe- gation, Vol. 57, No. 7, pp. 1863–1878, Jul. 2009, doi:
den, 2013, pp. 793-797. 10.1109/TAP.2009.2021934.
[55] IEEE Std 149-2021, IEEE Recommended Prac- [67] O. Breinbjerg, et al., “The 2019-2022 ESA-EurAAP Fa-
tice for Antenna Measurements, Feb. 2022, doi: cility Comparison Campaign with the DTU-ESA mm-
10.1109/IEEESTD.2022.9714428. VAST Antenna — Mid-Term Status,” 2022 16th Eu-
[56] A. C. Newell, et al., “Accurate measurement of an- ropean Conference on Antennas and Propagation (Eu-
tenna gain and polarisation at reduced distances by an CAP), Madrid, Spain, 2022, pp. 1-5, doi: 10.23919/Eu-
extrapolation technique,” IEEE Transactions on Anten- CAP53622.2022.9769507.
nas and Propagation, Vol. 21, No. 4, Jul. 1973, doi: [68] J. A. Gordon et al., ”Millimeter-Wave Near-Field Measure-
10.1109/TAP.1973.1140519. ments Using Coordinated Robotics,” IEEE Transactions
[57] T. H. Loh, ”Non-invasive measurement of electrically on Antennas and Propagation, Vol. 63, No. 12, pp. 5351-
small ultra-wideband and smart antennas”, Invited Paper, 5362, Dec. 2015, doi: 10.1109/TAP.2015.2496110.
the 9th Loughborough Antennas and Propagation Confer- [69] C. Culotta-López et al., ”On the Uncertainty Sources of
ence (LAPC 2013), Loughborough, UK, Nov. 2013, pp. Drone-Based Outdoor Far-Field Antenna Measurements,”
456 – 460, doi: 10.1109/LAPC.2013.6711941. 2021 Antenna Measurement Techniques Association Sym-
[58] ITU-R. Minimum Requirements, Related to Technical posium (AMTA), Daytona Beach, FL, USA, 2021, pp. 1-6,
Performance for IMT-2020 Radio Interface(s). Report ITU- doi: 10.23919/AMTA52830.2021.9620638.
R M.2410. Nov. 2017. [70] T. Gemmer, “Generalised Test-Zone Field Compensation”,
[59] K. A. Remley, et al., “Measurement Challenges for Apprimus Verlag, 2022, ISBN 978-3-98555-083-8.
5G and Beyond: An Update from the National Insti- [71] A. Geise, O. Neitz, J. Migl, H.J. Steiner, T. Fritzel, C.
tute of Standards and Technology”, IEEE Microwave Hunscher, T. Eibert, “A Crane-based Portable Antenna
Magazine, Vol. 18, No. 5, pp. 41–56, Jul. 2017, doi: Measurement System – System Description and Valida-
10.1109/MMM.2017.2690882 tion”, IEEE Transactions on Antennas and Propagations,
[60] 3GPP TR 38.810 V16.6.1 (2020-11), NR; Study on test Vol. 67, No. 5, pp. 3346-57, May 2019.
methods, Nov. 2020. [72] C. Cappellin, M. Lori, A. Geise, C. Hunscher, L.
[61] H. Kong, et al, ”Mid-field OTA RF test method: new Datashvili, “Predicted and Measured Antenna Patterns
developments and performance comparison with the com- of the European Large Deployable Reflector”, EuCAP
pact antenna test range (CATR)”, European Conference 2022, Madrid.
on Antennas and Proagation (EuCAP 2020), Copen- [73] E. Im, S. L. Durden, S. Tanelli, “CloudSat: The Cloud
hagen, Denmark, pp. 1-5, Mar. 2020, doi: 10.23919/Eu- Profiling Radar Mission”, Proceedings of the 2006 CIE
CAP48036.2020.9135921. International Conference on Radar, October 16-19, 2006,
[62] P.-S. Kildal, et al., ”Characterization of Reverberation Shanghai, China.
Chambers for OTA Measurements of Wireless Devices: [74] P. Focardi, J. Harrell, J. Vacchione, “Large/Complex
Physical Formulations of Channel Matrix and New Un- Antenna Performance Validation for Spaceborne
certainty Formula”, IEEE Transactions on Antennas and Radar/Radiometric Instruments”, Proceedings of

41
RoE Journal Antenna Measurement Challenges and Opportunities - 42/49

IEEE Antennas and Propagation Society International near-field measurements,” 2017 IEEE Conference on An-
Symposium, Orlando, FL, July 7-13, 2013. tenna Measurements & Applications (CAMA), Tsukuba,
Japan, 2017, pp. 1-4, doi: 10.1109/CAMA.2017.8273367.
[75] P. Focardi, P. Brown, J. Vacchione, J. Harrell, “SMAP:
Performance Verification and Testing of a Challenging [89] D. M. Lewis, J. Bommer, G. E. Hindman and S. F. Greg-
Instrument Antenna”, Proceedings of the 9th European son, ”Traditional to Modern Antenna Test Environments:
Conference on Antennas and Propagation, Lisbon, Portu- The Impact of Robotics And Computational Electromag-
gal, April 12-17, 2015. netic Simulation on Modern Antenna Measurements,”
2021 15th European Conference on Antennas and Propa-
[76] IEEE Std 1720-2012, IEEE Recommended Practice
gation (EuCAP), Dusseldorf, Germany, 2021, pp. 1-5, doi:
for Near-Field Antenna Measurements, Dec. 2012, doi:
10.23919/EuCAP51087.2021.9411068.
10.1109/IEEESTD.2012.6375745.
[90] D. J. van Rensburg B. Walkenhorst Q. Ton and J. De-
[77] E. S. Gillespie (ed.), “Special issue on near-field scan-
mas, ”A robotic near-field antenna test system relying
ning techniques,” IEEE Transactions on Antennas and
on non-canonical transformation techniques”, Antenna
Propagation, Vol. AP-36, pp. 727-901, June 1988.
Measurement Techniques Association Symposium (AMTA)
[78] A. Newell, R. Baird and P. Wacker, ”Accurate measure- Proceedings, 2019.
ment of antenna gain and polarization at reduced distances
[91] R. Moch, D. Heberling, ”Fast Spline-Based Antenna Mea-
by an extrapolation technique,” IEEE Transactions on An-
surements for Robotic Test Ranges via Pointwise Probe
tennas and Propagation, Vol. 21, No. 4, pp. 418-431, July
Correction,” 2022 16th European Conference on Antennas
1973, doi: 10.1109/TAP.1973.1140519.
and Propagation (EuCAP), Madrid, Spain, 2022, pp. 1-5,
[79] A. G. Repjar, A. C. Newell, and D. T. Tamura, “Extrapola- doi: 10.23919/EuCAP53622.2022.9769401.
tion Range Measurements for Determining Antenna Gain
[92] A. J. Weiss et al., ”Configuration and Control of a
and Polarization,” NBS Tech Note 1311, August 1987.
Millimeter-Wave Synthetic Aperture Measurement
[80] R. S. Hartenberg, J. Denavit, ”A kinematic notation for System with Uncertainties,” 2020 95th ARFTG
lower pair mechanisms based on matrices”, J. Appl. Mech. Microwave Measurement Conference (ARFTG),
1955, 77, 215–221 Los Angeles, CA, USA, 2020, pp. 1-4, doi:
[81] S. Hayati, K. Tso, G. Roston, ”Robot geometry calibra- 10.1109/ARFTG47271.2020.9241381.
tion”, Proceedings of the IEEE International Conference [93] J. Y. Kwon, C. Cho, and W. Chung, ”Robot-based an-
on Robotics and Automation, Philadelphia, PA, USA, tenna measurement system for 6G frequency band,” 2023
24–29 April 1988; Vol. 2, pp. 947–951 The Korean Institute of Electromagnetic Engineering and
Science (KIEES) Winter Conference, Feb. 2023, pp. 1.
[82] D. A. Stewart, ”Platformwith Six Degrees of Freedom”,
Proc. Inst. Mech. Eng. 1965, 180, 371–386. [94] P. A. Slater, J. M. Downey, M. T. Piasecki, and B. L.
Schoenholz, “Portable laser guided robotic metrology sys-
[83] V. E. Gough, ”Universal tyre test machine”, Proceedings
tem,” 2019 Antenna Measurement Techniques Association
of the FISITA 9th International International Automobile
Symposium (AMTA), IEEE, Oct. 2019.
Technical Congress, London, UK, 30 April–5 May 1962.
[95] L. F. Rolo, M. H. Paquay, R. J. Daddato, J. A. Parian,
[84] J. A. Gordon et al., ”Millimeter-Wave Near-Field Measure-
D. Doyle and P. de Maagt, ”Terahertz Antenna Tech-
ments Using Coordinated Robotics,” IEEE Transactions
nology and Verification: Herschel and Planck - A Re-
on Antennas and Propagation, Vol. 63, No. 12, pp. 5351-
view,” IEEE Transactions on Microwave Theory and Tech-
5362, Dec. 2015, doi: 10.1109/TAP.2015.2496110.
niques, Vol. 58, No. 7, pp. 2046-2063, July 2010, doi:
[85] T. K. Tanev, ”Kinematics of a hybrid (parallel–serial) robot 10.1109/TMTT.2010.2050179.
manipulator”, Mech. Mach. Theory 2000, 35, 1183–1196.
[96] V. Sánchez et al., ”Measurement of bepicolombo mis-
[86] B.L. Moser, J. A. Gordon, and A. J. Petruska, “Unified sion Medium Gain Antenna parameters,” 2016 10th Eu-
parameterization and calibration of serial, parallel, and ropean Conference on Antennas and Propagation (Eu-
hybrid manipulators,” Robotics, Vol. 10, No. 4, p. 124, CAP), Davos, Switzerland, 2016, pp. 1-3, doi: 10.1109/Eu-
Nov. 2021. CAP.2016.7481695.
[87] J. A. Gordon et al., ”An All-Metal, 3-D-Printed CubeSat [97] L. Rolo et al., ”Characterization of Reflectivity Losses in
Feed Horn: An assessment of performance conducted Space Reflector Antennas at Temperatures above 350°C”,
at 118.7503 GHz using a robotic antenna range,” IEEE 2015 AMTA Symposium, Long Beach, California, USA,
Antennas and Propagation Magazine, Vol. 59, No. 2, pp. November 2015.
96-102, April 2017, doi: 10.1109/MAP.2017.2655574.
[98] Y. Wang, C. Wang, P. Lian, S. Xue, J. Liu, W. Gao, Y.
[88] D. R. Novotny, J. A. Gordon, M. S. Allman, J. R. Guerrieri Shi, Z. Wang, K. Yu, X. Peng, B. Du and S. Xiao, Effect
and A. E. Curtin, ”Three antenna ranges based on artic- of Temperature on Electromagnetic Performance of Ac-
ulated robotic arms at the national institute of standards tive Phased Array Antenna. Electronics, 2020, 9, 1211.
and technology: Usability for over-the-air and standard https://doi.org/10.3390/electronics9081211.

42
RoE Journal Antenna Measurement Challenges and Opportunities - 43/49

[99] M. D. Stegman, M. Fedyk and S. Kuehn, ”Solar Thermal [111] Modern Automotive Antenna Measurements Edited by
Vacuum testing of deployable mesh reflector for model cor- Lars J. Foged and Manuel Sierra Castañer, Artech House,
relation,” 2010 IEEE Aerospace Conference, Big Sky, MT, 2022.
USA, 2010, pp. 1-15, doi: 10.1109/AERO.2010.5446961.
[112] L. J. Foged, L. Scialacqua, F. Saccardi, F. Mioc, D.
[100] A. Giacomini, V. Schirosi, A. Martellosio, L. J. Foged, Tallini, E. Leroux, U. Becker, J. L. Araque Quijano, G.
C. Feat, J. Sinigaglia, S. Leroy, F. Viguier, M. Moscetti Vecchi, “Bringing Numerical Simulation and Antenna
Castellani, D. Cardoni, A. Maraca, F. Rinalducci and L. Measurements Together,” 2014 IEEE Antennas and Prop-
Rolo, ”Spherical Near-Field Measurements of Satellite agation Society International Symposium, The Hague,
Antennas at Extreme Temperatures”, 2019 AMTA Sympo- Netherlands, 2014.
sium, San Diego, California, USA, October 2019.
[113] L. Scialacqua, F. Mioc, E. Toniolo and L. J. Foged, “As-
[101] P. Noren, Ph. Garreau, L. J. Foged, “State of the art sessment of Automotive SAR from Direct Measurements
spherical near-field antenna test systems for full vehicle and Post Processing,” 2022 16th European Conference on
testing,” EuCAP 2012, Prague, Czech Republic, 2012. Antennas and Propagation (EuCAP), Madrid, Spain, 2022,
[102] M. E. Asghar, C. Bornkessel, R. Stephan and M. A. pp. 1-4, doi: 10.23919/EuCAP53622.2022.9769387.
Hein, “Automotive Antenna Measurements in the Installed
[114] S. B. Jayapal Gowdu, P. Aust, A. Schwind, F. Hau, and
State Under Variable Boundary Conditions,” EuCAP
Matthias A. Hein, “Evaluation of scenario-based automo-
2022, Madrid, Spain, 2022, pp. 1-4, doi: 10.23919/Eu-
tive radar testing in virtual environment using real driving
CAP53622.2022.9769602.
data,” 2022 IEEE 25th International Conference on Intelli-
[103] F. Saccardi, et al., “Accurate Calibration of Truncated gent Transportation Systems (ITSC), Macau, China, 2022,
Spherical Near Field Systems with Different Ground DOI: 10.1109/ITSC55140.2022.9922366.
Floors using the Substitution Technique”, AMTA 2019,
San Diego, USA, 2019. [115] F. Saccardi, F. Mioc, A. Giacomini, L. J. Foged, “Es-
timation of the Realistic Ground Effect in Free-Space
[104] F. Saccardi, F. Rossi, F. Mioc, L. J. Foged, P.O. Iversen,
Automotive Measurements,” AMTA 2018, Williamsburg,
“Application of the Translated-SWE Algorithm for the
Williamsburg, VA, USA, 2018.
Characterization of Antennas Installed on Cars Using a
Minimum Number of Samples,” AMTA 2017, Atlanta, GA, [116] F. Saccardi, F. Mioc, L. J. Foged, P. O. Iversen, J. Estrada
USA, 2017. “Estimation of Real Ground Effect in Absorber-Based Au-
[105] F. Wollenschläger, L. J. Foged, M. E. Asghar, C. tomotive Measurements: Experimental Validation,” APS
Bornkessel and M. A. Hein, ”Spherical Wave Expansion 2020, Montréal, Canada, 2020.
Applied to the Measured Radiation Patterns of Automo- [117] G. Virone et al., “Antenna Pattern Verification System
tive Antennas in the Installed State in the GHz Range,” Based on a Micro Unmanned Aerial Vehicle (UAV)”,
EuCAP 2019, Krakow Poland, 2019. IEEE Antennas and Wireless Propagation Letters, Vol.
[106] F. Saccardi, F. Rossi, L. Scialacqua, L. J. Foged, “Trun- 13, pp. 169-172, 2014.
cation Error Mitigation in Free-Space Automotive Partial [118] L. Washburn, E. Romero, C. Johnson, C. Gotschalk and
Spherical Near Field Measurements,” AMTA 2017, At- B. Emery, “Antenna calibration for oceanographic radars
lanta, GA, USA, 2017. using aerial drones”, 2016 IEEE Conference on Antenna
[107] P. Berlt, F. Wollenschlaeger, C. Bornkessel and M. Measurements & Applications (CAMA), Syracuse, NY,
A. Hein, “Reliable Derivation of Automotive Antenna USA, 2016, pp. 1-4.
Gain Patterns from LTE Communication Parameters,”
[119] F. Paonessa et al., “Design and Verification of a Q-Band
2018 IEEE 87th Vehicular Technology Conference (VTC
Test Source for UAV-Based Radiation Pattern Measure-
Spring), Porto, Portugal, 2018, pp. 1-5, doi: 10.1109/VTC-
ments”, IEEE Transactions on Instrumentation and Mea-
Spring.2018.841749.
surement, Vol. 69, No. 12, pp. 9366-9370, Dec. 2020.
[108] A. Scannavini, F. Mioc, K. Rutkowski, M. Mercier, F.
Saccardi and L. J. Foged “Experimental validation of au- [120] I. Farhat, D. Cutajar, K. Z. Adami, C. Sammut and
tomotive OTA measurements at close distance,” EuCAP J. Abela, “Characterization of 36 Meter Square Mid-
2022, Madrid, Spain, 2022. Frequency Radio Astronomy Prototype Antenna Array”,
2018 IEEE Conference on Antenna Measurements & Ap-
[109] F. Mioc, F. Saccardi A. Scannavini, L. J. Foged, K. plications (CAMA), Vasteras, Sweden, 2018, pp. 1-3.
Rutkowski “Reduced Distance OTA Testing Methodolo-
gies for Automotive Applications,” AMTA 2022, Denver, [121] L. Ciorba et al., “Large Horizontal Near-Field Scan-
CO, USA, 2022. ner Based on a Non-Tethered Unmanned Aerial Vehicle”,
IEEE Open Journal of Antennas and Propagation, Vol. 3,
[110] F. Saccardi, A. Scannavini, L. Scialacqua, L. J. Foged
pp. 568-582, 2022.
N. Gross, A. Gandois, S. Dooghe, P. O. Iversen “Virtual
Drive Testing based on Automotive Antenna Measure- [122] F. Paonessa et al., “SKA-Low Prototypes Deployed in
ments for Evaluation of Vehicle-to-X Communication Per- Australia: Synoptic of the UAV-Based Experimental Re-
formances,” AMTA 2019, San Diego, USA, 2019. sults”, Radio Science Letters, Vol. 3, pp. 1-5, 2020.

43
RoE Journal Antenna Measurement Challenges and Opportunities - 44/49

[123] P. Di Ninni et al., “Electromagnetic Analysis and Ex- [135] P. Bolli, G. Pupillo, F. Paonessa, G. Virone, S. J. Wijn-
perimental Validation of the LOFAR Radiation Patterns”, holds and A. M. Lingua, “Near-Field Experimental Ver-
International Journal of Antennas and Propagation, Vol. ification of the EM Models for the LOFAR Radio Tele-
2019, pp. 1-12, 2019. scope”, IEEE Antennas and Wireless Propagation Letters,
Vol. 17, No. 4, pp. 613-616, April 2018.
[124] A. Nelles et al., “Calibrating the absolute amplitude scale
for air showers measured at LOFAR”, Journal of Instru- [136] C. Culotta-López, S. Skeidsvoll, A. Buchi and J. Es-
mentation, Vol. 10, No. 11, Nov. 2015. peland, “Use of UASs for Outdoor Diagnostics of Large
Antennas”, 2022 Antenna Measurement Techniques Asso-
[125] G. Virone et al., “Measurement of the LOFAR-HBA ciation Symposium (AMTA), 2022, pp. 1-6.
beam patterns using an unmanned aerial vehicle in the near
field”, Journal of Astronomical Telescopes Instruments [137] Zhinong Ying et al., ”Study of phased array in UE for 5G
and Systems, Vol. 8, No. 1, pp. 1-14, Nov. 2021. mm wave communication system with consideration of
user body effect,” 10th European Conference on Antennas
[126] C. Chang, C. Monstein, A. Refregier, A. Amara, A. and Propagation (EuCAP), Davos, Switzerland, 2016.
Glauser and Sarah Casura, “Beam Calibration of Radio
Telescopes with Drones”, Publications of the Astronom- [138] A. Scannavini, L. J. Foged, N. Gross, L. Hentila, J.
ical Society of the Pacific, Vol. 127, No. 957, pp. 1131- Virtala and K. Raju, ”Test zone characterization for the
1143. multiprobe anechoic chamber setup (MPAC),” 10th Euro-
pean Conference on Antennas and Propagation (EuCAP),
[127] S. Punzet et al., “Fully Coherent UAV-Based Near-Field Davos, Switzerland, 2016.
Measurement and Transformation of the S67-15 m Ground
[139] M. Sierra Castañer, Lars J. Foged. Post-processing tech-
Station Antenna at the German Space Operations Center in
niques in Antenna Measurements. Scitech Publishing
Weilheim”, 2022 16th European Conference on Antennas
(IET). 2019. ISBN: 9781785615375.
and Propagation (EuCAP), 2022, pp. 1-5.
[140] L. Scialacqua, F. Mioc, A. Scannavini and L. J. Foged,
[128] C. Culotta-López et al., “On the Uncertainty Sources of ”Simulated and Measured Power Density using Equiva-
Drone-Based Outdoor Far-Field Antenna Measurements”, lent Currents for 5G applications,” IEEE International
2021 Antenna Measurement Techniques Association Sym- Symposium on Antennas and Propagation Montreal, QC,
posium (AMTA), 2021, pp. 1-6. Canada, 2020.
[129] Y. Tian, B. Wen, Z. Li, Y. Yin and W. Huang, “Analysis [141] Test Plan for Wireless Device Over-the-Air Performance,
and Validation of an Improved Method for Measuring HF Revision 3.3, CTIA Certification Standard, Oct. 2013.
Surface Wave Radar Antenna Pattern”, IEEE Antennas
and Wireless Propagation Letters, Vol. 18, No. 4, pp. 659- [142] 3GPP TR 37.977, Universal Terrestrial Radio Access
663, April 2019. (UTRA) and Evolved Universal Terrestrial Radio Access
(E-UTRA); Verification of radiated multi-antenna recep-
[130] Y. Hussein, J. Miller, V. Garkanian and E. Decrossas, tion performance of User Equipment (UE), V16.0.0, Jul.
“Far-Field Pattern Measurement and Simulation of VHF 2020.
Antenna at 60 MHz for Europa Clipper mission”, 2018
[143] “Test plan for 2x2 downlink MIMO and transmit diver-
IEEE Symposium on Electromagnetic Compatibility, Sig-
sity over-the-air performance,” CTIA Certification, Wash-
nal Integrity and Power Integrity (EMC, SI&PI), 2018, pp.
ington, D.C., Tech. Rep. Version 1.0, Aug. 2015.
94-98.
[144] Y. Qi, G. Yang, L. Liu, J. Fan, A. Orlandi, H. Kong,
[131] J. Schreiber, “Antenna pattem reconstitution using un- W. Yu, and Z. Yang, “5G over-the-air measurement chal-
manned aerial vehicles (UAVs)”, 2016 IEEE Conference lenges: Overview,” IEEE Trans. Electromagn. Compat.,
on Antenna Measurements & Applications (CAMA), 2016, Vol. 59, No. 6, pp. 1661–1670, Dec. 2017.
pp. 1-3.
[145] 3GPP TR 38.827, Study on Radiated Metrics and Test
[132] C. Wang, B. Yuan, W. Shi and J. Mao, “Low-Profile Methodology for the Verification of Multi-Antenna Recep-
Broadband Plasma Antenna for Naval Communications in tion Performance of NR User Equipment (UE), V16.4.0,
VHF and UHF Bands”, IEEE Transactions on Antennas Sep. 2021.
and Propagation, Vol. 68, No. 6, pp. 4271-4282, June
2020. [146] W. Fan, P. Kyösti, M. Rumney, X. Chen, and G. F. Ped-
ersen, “Over the-air radiated testing of millimeter-wave
[133] M. Garcı́a-Fernández et al., “Antenna Diagnostics and beam-steerable devices in a cost-effective measurement
Characterization Using Unmanned Aerial Vehicles”, IEEE setup,” IEEE Communications Magazine, Vol. 56, No. 7,
Access, Vol. 5, pp. 23563-23575, 2017. pp. 64–71, Jul. 2018.
[134] C. Culotta-López, S. Skeidsvoll, V. Costin, A. Buchi and [147] F. Zhang, W. Fan, and Z. Wang, “Achieving wireless
J. Espeland, “On Gain Measurements Using the Substi- cable testing of high-order MIMO devices with a novel
tution Method with UASs”, 2021 IEEE Conference on closed-form calibration method,” IEEE Transactions on
Antenna Measurements & Applications (CAMA), Antibes Antennas and Propagation, Vol. 69, No. 1, pp. 478–487,
Juan-les-Pins, France, 2021, pp. 51-54. Jan. 2021.

44
RoE Journal Antenna Measurement Challenges and Opportunities - 45/49

[148] H. Gao et al., “Over-the-Air Performance Testing of 5G and High Absorption Rate,” IEEE Transactions on
New Radio User Equipment: Standardization and Chal- Electromagnetic Compatibility, to be published, doi:
lenges,” IEEE Communications Standards Magazine, Vol. 10.1109/TEMC.2022.3212916.
6, No. 2, pp. 71-78, Jun. 2022. [162] L. Foged et al., “Echo suppression by spatial-filtering
[149] W. Fan,M. Li, Z. Wang, and F. Zhang, “Wireless Cable techniques in advanced planar and spherical near-field
Testing for MIMO Radios: A Compact and Cost-effective antenna measurements [AMTA Corner],” IEEE Antennas
5G Radio Performance Test Solution,” Accepted in IEEE and Propagation Magazine, Vol. 55, No. 5, pp. 235–242,
Transactions on Antennas and Propagation, 2023 Oct. 2013.
[150] Study on Test Methods, document 3GPP TR 38.810 [163] 3GPP TR 38.815, “New frequency Range for NR (24.25-
v16.6.1, Sep. 2020. 29.5 GHz)” Release 15.
[151] “Radio frequency (RF) requirement background for ac- [164] NGMN Recommendation on Base Station Active An-
tive antenna system (AAS) base station (BS) radiated re- tenna System Standards.
quirements,” 3GPP, ETSI, Sophia Antipolis, France, Tech.
[165] Y. P. Zhang and D. Liu, ”Antenna-on-Chip and Antenna-
Rep. TR 37.843, 2017.
in-Package Solutions to Highly Integrated Millimeter-
[152] X. Chen et al., “Reverberation chambers for over-the-air Wave Devices for Wireless Communications,” IEEE Trans-
tests: An overview of two decades of research,” IEEE actions on Antennas and Propagation, Vol. 57, No. 10, pp.
Access, Vol. 6, pp. 49129–49143, 2018. 2830-2841, Oct. 2009.
[153] C. Rowell, B. Derat and A. Cardalda-Garcı́a, “Multiple [166] K. Tekkouk et al., ”Recent and future research trends
CATR Reflector System for Multiple Angles of Arrival in planar multi-beam antennas in the millimeter wave
Measurements of 5G Millimeter Wave Devices,” IEEE range at IETR-France,” 2016 International Symposium on
Access, Vol. 8, pp. 211324-211334, 2020. Antennas and Propagation (ISAP), Okinawa, Japan, 2016,
[154] Y. Zhang, Z. Wang, X. Sun, Z. Qiao, W. Fan, and J. Miao, pp. 124-125
“Design and implementation of a wideband dual-polarized [167] F. Rodrı́guez Varela, J. Fernandez Álvarez, B. Galocha
plane wave generator with tapered feeding nonuniform Iragüen, M. Sierra Castañer and O. Breinbjerg, ”Numeri-
array,” IEEE Antennas Wireless Propagation Letters, Vol. cal and Experimental Investigation of Phaseless Spherical
19, No. 11, pp. 1988–1992, Nov. 2020. Near-Field Antenna Measurements,” IEEE Transactions
[155] A. Tankielun, A. Belkacem, M. Akinci, M. Celik, H. on Antennas and Propagation, Vol. 69, No. 12, pp. 8830-
Bartko, and B. Derat, “Quiet zone verification of plane 8841, Dec. 2021
wave synthesizer using polar near-field scanner,” in Proc. [168] D.J. Janse van Rensburg and B.M. Williams, “Modular
14th Eur. Conf. Antennas Propag. (EuCAP), Mar. 2020, frequency up conversion for antenna ranges”, Antenna
pp. 1–5. Systems Conference, Dallas, TX, Oct 2010.
[156] F. Scattone et al., “Production Measurement of 5G Mil- [169] D. Janse van Rensburg, G.E. Hindman, “An Overview
limeter Wave Plane Wave Generators,” in Proc. 14th Eur. of Near-Field Sub-millimeter Wave Antenna Test Appli-
Conf. Antennas Propag. (EuCAP), Mar. 2022, pp. 1–3. cations”, COMITE Conference, Czech Republic, 23-24
[157] O. A. Iupikov, P. S. Krasov, A. A. Glazunov, R. Maaskant, April, 2008
J. Fridén and M. V. Ivashina, “Hybrid OTA Chamber for [170] L. Le Coq, N. Mézières, P. Leroy and B. Fuchs, “Some
Multidirectional Testing of Wireless Devices: Plane Wave Contributions for Antenna 3D Far Field Characterization
Spectrum Generator Design and Experimental Demonstra- at Terahertz”. Sensors 2021, 21, 1438.
tion,” IEEE Transactions on Antennas and Propagation,
[171] Y. Tanaka and S. Hisatake, ”Cylindrical Near-Field Mea-
Vol. 70, No. 11, pp. 10974 - 10987, Nov. 2022.
surement and Far-Field Characterization of 300-GHz Band
[158] K. Hassett, “Phased array antenna calibration measure- Antenna Based on an Electrooptic Measurement Tech-
ment techniques and methods,” in Proc. Eur. Conf. Anten- nique,” 2022 Antenna Measurement Techniques Associ-
nas Propag. (EuCAP), 2016, pp. 1–4. ation Symposium (AMTA), Denver, CO, USA, 2022, pp.
[159] F. Zhang, W. Fan, Z. Wang, Y. Zhang, and G. F. Pedersen, 1-4.
“Improved over-the-air phased array calibration based on [172] H. Füser, M. Bieler and A. Geise, ”Using THz frequency
measured complex array signals,” IEEE Antennas Wireless combs generated from unstabilized femtosecond lasers
Propagation Letters, Vol. 18, No. 6, pp. 1174– 1178, 2019. for near- and far-field characterization of GHz and THz
[160] H. Kong, Z. Wen, Y. Jing, and M. Yau, “Midfield over- emitters,” 2012 37th International Conference on Infrared,
the-air test: A new ota rf performance test method for 5g Millimeter, and Terahertz Waves, Wollongong, NSW, Aus-
massive mimo devices,” IEEE Transactions on Microwave tralia, 2012, pp. 1-3.
Theory and Techniques, Vol. 67, No. 7, pp. 2873–2883, [173] J. Migl, H.-J. Steiner, J. Habersack, C. H. Schmidt, C.
2019. Hoetzel, “Challenging Compact Range Applications Fea-
[161] S. S. Zhekov, P. Mei, G. F. Pedersen and W. turing Recent Developments”, AMTA 37th Annual Meet-
Fan, “Hybrid Absorber With Dual-Band Performance ing and Symposium, Long Beach, CA, USA, Oct 2015.

45
RoE Journal Antenna Measurement Challenges and Opportunities - 46/49

[174] G. Forma et al., ”Radiation-Pattern Measurements and Antenna Measurement Techniques Association (AMTA)
Predictions of the PLANCK RF Qualification Model Symposium. 2016.
[AMTA Corner]”, IEEE Antennas and Propagation Mag-
[187] S. Costanzo, G. Di Massa, and M. D. Migliore, “A
azine, Vol. 51, No. 6, pp. 213-219, Dec. 2009;
Novel Hybrid Approach for Far-Field Characterization
[175] Gomez-Torrent, A.; Vigueras, M.G.; Le Coq, L.; Mah- from Near-Field Amplitude-Only Measurements on Arbi-
moud, A.; Ettorre, M.; Sauleau, R.; Oberhammer, J. “A trary Scanning Surfaces,” IEEE Transactions on Antennas
Low-Profile and High-Gain Frequency Beam Steering and Propagation, Vol. 53, No. 6, pp. 1866-1874, June
sub-THz Antenna Enabled by Silicon Micromachining”, 2005.
IEEE Transactions on Antennas and Propagation, 2020,
2, 672–682. [188] E. N. Leith and J. Upatnieks. Reconstructed wave-
fronts and communication theory. J. Optical Soc. Am.,
[176] A. Karttunen et al., ”Antenna Tests With a Hologram- 52(10):1123–1128, Oct. 1962.
Based CATR at 650 GHz,” IEEE Transactions on Anten-
nas and Propagation, Vol. 57, No. 3, pp. 711-720, March [189] A. Arboleya, Y. Alvarez, J. Laviada, F. Las-Heras, “Phase
2009. recovery in antenna measurements”, Post-processing Tech-
niques in Antenna Measurement, Editors: Manuel Sierra
[177] C. Rowell, B. Derat and A. Cardalda-Garcia, ”Multiple
Castaner; Lars J. Foged. IET Digital Library. Book. Sept
CATR Reflector System for 5G Radio Resource Manage-
2019.
ment Measurements,” 2021 15th European Conference
on Antennas and Propagation (EuCAP), Dusseldorf, Ger- [190] J. Laviada, A. Arboleya, Y. Álvarez, C. Garcı́a-González,
many, 2021, pp. 1-5. F. Las-Heras, “Phaseless Antenna Diagnostics Based on
[178] D.J. Janse van Rensburg and G. Hindman, “Sub- Off-Axis Holography with Synthetic Reference Wave,”
millimeter wave planar near-field antenna testing”, EU- IEEE Antennas and Wireless Propagation Letters, Vol. 13,
CAP Conference, Berlin, Germany, March 2009. 43-46, 2014.

[179] D.J. Janse van Rensburg, “Factors limiting the upper [191] R. W. Gerchberg, and W. O. Saxton, “A practical al-
frequency of mm-wave spherical near-field test systems”, gorithm for the determination of phase from image and
EUCAP Conference, Lisbon, Portugal, April 2015. (In- diffraction plane pictures,” Optik, Vol. 35, pp. 237–246,
vited paper) 1972.
[180] D.J. Janse van Rensburg and P. Betjes, “Structural Cor- [192] A. Levi, H. Stark, “Image restoration by the method
rection of a Spherical Near-Field Scanner for mm-Wave of generalized projections with application to restoration
Applications”, AMTA 37th Annual Meeting and Sympo- from magnitude,” J. Opt. Soc. Am. A, Vol. 1, n. 9, pp.
sium, Long Beach, CA, USA, Oct 2015. 932-943, Sept. 1984.
[181] Novotny, D.; Gordon, J.; Coder, J.; Francis, M.; Guerri- [193] O. M. Bucci, G. D’Elia, G. Leone, and R. Pierri, “Far-
eri, J., ”Performance evaluation of a robotically controlled field pattern determination from the near-field amplitude
millimeter-wave near-field pattern range at the NIST,” on two surfaces,” IEEE Transactions on Antennas and
7th European Conference on Antennas and Propagation, Propagation , Vol. 38, No. 11, pp. 1772–1779, Nov. 1990.
pp.4086 -4089, 8-12 April 2013.
[194] R. G. Yaccarino, Y. Rahmat-Samii, “Phaseless bi-polar
[182] Z. Chen, “Common Microwave Absorbers Eavaluation planar near-field measurements and diagnostics of array
in W-band (75-110GHz)”, AMTA 39th Annual Meeting antennas,” IEEE Transactions on Antennas and Propaga-
and Symposium, Atlanta, GA, USA, Oct 2017. tion, Vol. 47, No. 3, pp. 574-583, Mar. 1999.
[183] J. Säily, A. V. Räisänen, “Studies on Specular and
[195] A. Capozzoli, C. Curcio, G. D’Elia, A. Liseno, “Phase-
Non-specular Reflectivities of Radar Absorbing Materials
less antenna characterization by effective aperture field
(RAM) at Submillimetre Wavelengths”, Vol. 258, Helsinki
and data representations,” IEEE Transactions on Antennas
University of Technology Radio Laboratory publications,
and Propagation, Vol. 57, No. 1, pp. 215-230, Jan. 2009.
ISSN 1456-3835.
[184] P. K. Agrawal, ”A method to compensate for probe posi- [196] Y. Álvarez, F. Las-Heras, M. R. Pino, ”The sources recon-
tioning errors in an antenna near field facility,” Proceed- struction method for amplitude-only field measurements,”
ings of IEEE Antennas Propag. Society International Sym- IEEE Transactions on Antennas and Propagation, Vol. 58,
posium, Albuquerque, NM, pp. 218–221, May 1982. No. 8, pp. 2776-2781, Aug. 2010.

[185] J. Tuovinen, A. Lehto, and A. Raisanen, “Phase mea- [197] C. H. Schmidt, S. F. Razavi, T. F. Eibert, Y. Rahmat-
surements of millimeter wave antennas at 105–190GHz Samii, “Phaseless spherical near-field antenna measure-
with a novel differential phase method,” IEE Proceedings ments for low and medium gain antennas,” Adv. Radio
H—Microwaves, Antennas and Propagation, Vol. 138, No. Sci., Vol. 8, pp. 43-48, 2010.
2, pp. 114–120, Apr. 1991. [198] E. J. Candès, Y. C. Eldar, T. Strohmer, V. Voroninski,
[186] O. Breinbjerg, J. Fernández-Álvarez, “Phaseless Near- “Phase retrieval via matrix completion,” SIAM Rev., Vol. 2,
Field Antenna Measurement Techniques – An Overview,” pp. 225–251, Nov. 2015.

46
RoE Journal Antenna Measurement Challenges and Opportunities - 47/49

[199] R. Chandra, Z. Zhong, J. Hontz, V. McCulloch, C. [211] Y. Okano, K. Ito, I. Ida, and M. Takahashi, “The SAR
Studer, T. Goldstein, “Phasepack: A phase retrieval li- evaluation method by a combination of thermographic
brary,” Proc. 51st Asilomar Conf. Signals, Syst., Comput., experiments and biological tissue-equivalent phantoms,”
pp. 1617–1621, Nov. 2017. IEEE Transactions on Microwave Theory and Techniques,
Vol. 48, No. 11, pp. 2094–2103, 2000.
[200] B. Fuchs, M. Mattes, S. Rondineau, L. Le Coq, “Phase-
less Near Field Antenna Measurements from Two Surface [212] International Commission on Non-Ionizing Radiation
Scans - Numerical and Experimental Investigations,” IEEE Protection (ICNIRP), “ICNIRP RF EMF guidelines 2020,”
Transactions on Antennas and Propagation, Vol. 68, No. 2020.
3, pp. 2351-2322, March 2020. [213] A. Skrivervik, J.F. Zürcher, O. Staub, and J. Mosig, “PCS
[201] J. Fernández Álvarez, M. Mattes, O. Breinbjerg, “Phase antenna design: the challenge of miniaturization,” IEEE
retrieval for spherical near-field measurements using two Antennas and Propagation Magazine, Vol. 43, No. 4, pp.
antenna positions,” Antenna Measurement Techniques As- 12–27, 2001.
sociation Symposium (AMTA), pp. 1–6, 2021. [214] K. Ito, K. Furuya, Y. Okano, and L. Hamada, “Develop-
ment and characteristics of a biological tissue-equivalent
[202] J. Knapp, A. Paulus, J. Kornprobst, U. Siart, T. F. Eibert,
phantom for microwaves,” Electronics and Communica-
”Multifrequency Phase Retrieval for Antenna Measure-
tions in Japan (Part I: Communications), Vol. 84, No. 4,
ments,” IEEE Transactions on Antennas and Propagation,
pp. 67–77, 2001.
Vol. 69, No. 1, pp. 488-501, Jan. 2021.
[215] A. Alomainy, Y. Hao, Y. Yuan, and Y. Liu, “Modelling
[203] J. Dong, L. Valzania, A. Maillard, T. Pham, S. Gigan, M. and characterisation of radio propagation from wireless
Unser, ”Phase Retrieval: From Computational Imaging implants at different frequencies,” 2006 European Confer-
to Machine Learning: A tutorial,” IEEE Transactions on ence on Wireless Technology, 2006, pp. 119–122.
Signal Processing, Vol. 40, No. 1 pp. 45-47, Jan. 2023.
[216] O. Staub, J.F. Zürcher, and A. Skrivervik, “Some con-
[204] K. S. Nikita, Handbook of biomedical telemetry. Hobo- siderations on the correct measurement of the gain and
ken, NJ, USA: Wiley, 2014. bandwidth of electrically small antennas,” Microwave and
[205] A. Kiourti and K. S. Nikita, “A review of implantable Optical Technology Letters, Vol. 17, No. 3, pp. 156–160,
patch antennas for biomedical telemetry: Challenges and 1998.
solutions [Wireless Corner],” IEEE Antennas and Propa- [217] A. Skrivervik, M. Bosiljevac, and Z. Sipus, Antennas
gation Magazine, Vol. 54, No. 3, pp. 210–228, 2012. for implants: design and limitations. The Institution of
Engineering and Technology (IET), 2022, ch. 5, pp. 73–
[206] L. Vähä-Savo, C. Cziezerski, M. Heino, K. Haneda, C.
105.
Icheln, A. Hazmi, and R. Tian, “Empirical evaluation of
a 28 GHz antenna array on a 5G mobile phone using [218] S. Castelló-Palacios, C. Garcia-Pardo, A. Fornes-Leal, N.
a body phantom,” IEEE Transactions on Antennas and Cardona, and A. Vallés-Lluch, “Tailor-made tissue phan-
Propagation, Vol. 69, No. 11, pp. 7476–7485, 2021. toms based on acetonitrile solutions for microwave appli-
cations up to 18 GHz,” IEEE Transactions on Microwave
[207] N. H. Abd Rahman, Y. Yamada, and M. S. Amin Nordin, Theory and Techniques, Vol. 64, No. 11, pp. 3987– 3994,
“Analysis on the effects of the human body on the perfor- 2016.
mance of electro-textile antennas for wearable monitor-
[219] Y. El-Saboni, D. E. Zelenchuk, G. A. Conway, and W. G.
ing and tracking application,” Materials, Vol. 12, No. 10,
Scanlon, “Assessing the intrinsic radiation efficiency of
2019. [Online]. Available: https://www.mdpi.com/ 1996-
tissue-implanted UHF antennas,” IEEE Transactions on
1944/12/10/1636
Antennas and Propagation, Vol. 68, No. 1, pp. 491–499,
[208] M. Suzan Miah, A. N. Khan, C. Icheln, K. Haneda, and 2020.
K.-I. Takizawa, “Antenna system design for improved [220] M. Manteghi and A. A. Y. Ibraheem, “On the study of the
wireless capsule endoscope links at 433 MHz,” IEEE near-fields of electric and magnetic small antennas in lossy
Transactions on Antennas and Propagation, Vol. 67, No. media,” IEEE Transactions on Antennas and Propagation,
4, pp. 2687–2699, 2019. Vol. 62, No. 12, pp. 6491–6495, 2014.
[209] C. Andreu, S. Castelló-Palacios, C. Garcia-Pardo, A. [221] M. Barbi, C. Garcia-Pardo, A. Nevárez, V. Pons Beltrán,
Fornes-Leal, A. Vallés-Lluch, and N. Cardona, “Spatial and N. Cardona, “UWB RSS-based localization for cap-
in-body channel characterization using an accurate UWB sule endoscopy using a multilayer phantom and in vivo
phantom,” IEEE Transactions on Microwave Theory and measurements,” IEEE Transactions on Antennas and Prop-
Techniques, Vol. 64, No. 11, pp. 3995–4002, 2016. agation, Vol. 67, No. 8, pp. 5035–5043, 2019.
[210] S. Oh, Y.-C. Ryu, G. Carluccio, C. T. Sica, and C. M. [222] S. Costanzo, V. Cioffi, A. M. Qureshi, and A. Borgia,
Collins, “Measurement of SAR-induced temperature in- “Gellike human mimicking phantoms: Realization proce-
crease in a phantom and in vivo with comparison to numer- dure, dielectric characterization and experimental valida-
ical simulation,” Magnetic Resonance in Medicine, Vol. tions on microwave wearable body sensors,” Biosensors,
71, No. 5, pp. 1923–1931, 2014. Vol. 11, No. 4, 2021.

47
RoE Journal Antenna Measurement Challenges and Opportunities - 48/49

[223] F. Zhang, H. Zhang, H. Zhao, Z. He, L. Shi, Y. He, N. 2022 IEEE International Symposium on Antennas and
Ju, Y. Rong, and J. Qiu, “Design and fabrication of a Propagation and USNC-URSI Radio Science Meeting (AP-
personalized anthropomorphic phantom using 3D printing S/URSI), Denver, CO, USA, 2022, pp. 663-664.
and tissue equivalent materials,” Quantitative imaging in [236] https://www.euraap.org/Activities/wg/detailed-
medicine and surgery, Vol. 9, No. 1, p. 94, 2019. presentation-for-each-wg/measurements.
[224] D. Hess, ”Near-Field Measurement Experience at Scien-
[237] www.isoiec17025.com.
tific Atlanta,” Proceedings of the Antenna Measurement
Techniques Association (AMTA 1991), Boulder, CO, Octo- [238] M. A. Saporetti et al., ”X/Ku/Ka-band high Gain Reflec-
ber 7-11, 1991. tor Antenna Intercomparison Campaign Results,” 2021
Antenna Measurement Techniques Association Symposium
[225] R. C. Johnson, ”Antenna Range for Providing a Plane
(AMTA), 2021.
Wave for Antenna Measurements”. United States Patent
3,302,205, 31 January 1967. [239] M. Saporetti, L. J. Foged, L. Scialacqua, F. Saccardi,
[226] V. Rodriguez, Anechoic Range Design for Electromag- M. Mercier, J. Lui, F. Z. Xue, L. T. Chen, B.L.Bu, “An-
netic Measurements, Boston, MA: Artech House, 2019 tenna Measurement System Verification by Correlation
(ISBN 978-1-63081-537-0). with Facility-Comparison Campaign in the Framework of
the European Association of Antennas and Propagation”,
[227] V. Rodriguez, S. Matitisine, T. T. Chia, P. Lagoiski, L. International Wireless Symposium, March 2014.
Matytsine, M. Matytsine and P. K. Kan, “A Cone Shaped
Tapered Chamber for Antenna Measurements Both in Near [240] M.A.Saporetti, L.J. Foged, A. Scannavini, S. Weitz, J.
Field and Far Field in the 200 MHz to 18 GHz Frequency Kunisch, “Measurements and Simulations of MIMO 2x2
Range and Extension of the Quiet Zone using an RF Lens,” Reference Antennas”, EuCAP2016, Davos Switzerland,
Journal of the Applied Computational Electromagnetics April 2016.M. A. Saporetti, L. Foged, M. Sierra Castaner,
Society, Vol. 28, No. 12, pp. 1162 1170, December 2013. S. Pivnenko, R. Cornelius, D. Heberling, ”Description and
Results: Antenna Measurement Facility Comparisons,”
[228] K. Hassett and B. Williams, ”Multiplexed Pulsed Trans- IEEE APM, June 2017.
mit and Receive RF Measurement System for Active
Phased Array Testing,” Proceedings of the Antenna Mea- [241] M.A.Saporetti, L.J. Foged, A. A. Alexandridis, F. las
surement Techniques Association (AMTA) Symposium, Heras, C. López, Y. Kurdi, M. Sierra Castañer, “Interna-
November 2011. tional Facility Comparison Campaign at L/C Band Fre-
quencies”, AMTA, October 2017.
[229] E. A. Barry, P. N. Betjes, D. J. van Rensberg, P. Pelland,
”Implementation and Validation of a Satellite Payload Test [242] M. A. Saporetti et al “Recent Developments in Interna-
Suite for Planar Near Field Test Ranges,” Proceedings of tional Facility Comparison Campaigns”, IEEE Antenna
the Antenna Measurement Techniques Association (AMTA) Measurements Technique Association symposium, San
Symposium, October 2021. Diego (CA), October 2019.
[230] R. T. Cutshall, B. T. Walkenhorst, J. Dobbins, J. A. Frek- [243] M. A. Saporetti et al, “L, S, and C bands medium gain
ing, and B. Williams, ”A Review of the CW-Ambient Tech- ridge horn intercomparison campaign results”, IEEE Eu-
nique for Measuring G/T in a Planar Near-Field Antenna ropean Conference on Antenna Propagation, Copenhagen,
Range,” Proceedings of the Antenna Measurement Tech- March 2020.
niques Association (AMTA) Symposium, October 2019. [244] O.S. Kim, “Millimeter Wave VAlidation STan-
[231] B. Williams, E. Park, R. T. Cutshall, J. Dobbins, and J. A. dard (mm-VAST) antenna. ESA Contract No.
Freking, ” A Dual-Band High Power PNF Range with In- 4000109866/13/NL/MH. Executive Summary”, Technical
terleaved T/R and Pulse Synchronization,” Proceedings of Report, R784, Department of Electrical Engineering,
the Antenna Measurement Techniques Association (AMTA) Technical University of Denmark, September 2015.
Symposium, October 2021. [245] O.S. Kim et al., “DTU-ESA Millimeter-Wave VAlida-
[232] IEEE Std 211-2018, ”IEEE Standard Definitions of tion STandard Antenna (mm-VAST) – Detailed Design”,
Terms for Radio Wave Propagation,” Feb. 2019. Proceedings of the 9th European Conference on Antennas
[233] IEEE Std 145-2013, ”IEEE Standard for Definitions of and Propagation, EuCAP 2015, Lisbon, Portugal, April
Terms for Antennas,” March 2014. 2015.
[234] J. Fordham, L. J. Foged, V. Rodriguez, J. Dobbins and V. [246] O. S. Kim, S. Pivnenko, O. Breinbjerg, R. Jørgensen, N.
Monebhurrun, ”Publication of the Antenna Measurement Vesterdal, K. Branner, P. Berring, C. Markussen, and M.
Standard IEEE Std™ 149,” 2022 IEEE International Sym- Paquay, “DTU-ESA Millimeter-Wave Validation Standard
posium on Antennas and Propagation and USNC-URSI Antenna (mm-VAST) – Manufacturing and Testing”, Pro-
Radio Science Meeting (AP-S/URSI), Denver, CO, USA, ceedings of the 37th Annual Meeting and Symposium of
2022, pp. 661-662. the Antenna Measurement Techniques Association, Long
[235] L. J. Foged, J. Dobbins, V. Rodriguez, J. Fordham and V. Beach, California, October 2015.
Monebhurrun, ”Revision of IEEE Std. 1720-2012: Recom- [247] O. Breinbjerg et al., ”The 2019-2022 ESA-EurAAP Facil-
mended Practice for Near-Field Antenna Measurements,” ity Comparison Campaign with the DTU-ESA mm-VAST

48
RoE Journal Antenna Measurement Challenges and Opportunities - 49/49

Antenna — Mid-Term Status,” 2022 16th European Con- [261] A. Knisely, M. Havrilla, and P. Collins, “Biaxial
ference on Antennas and Propagation (EuCAP), Madrid, anisotropic sample design and rectangular to square
Spain, 2022. waveguide material characterization system,” in 2015 9th
[248] https://standards.ieee.org/ieee/2816/7620/. International Congress on Advanced Electromagnetic Ma-
terials in Microwaves and Optics (METAMATERIALS),
[249] C.R. Garcia, J. Correa, D. Espalin, J.H. Barton, R.C. 2015, pp. 346–348.
Rumpf, R. Wicker, and V. Gonzalez, “3D printing of
[262] E. Rothwell, J. Frasch, S. Ellison, P. Chahal, and R.
anisotropic metamaterials,” Progress in Electromagnet-
Ouedraogo, “Analysis of the nicolson-ross-weir method
ics Research Letters, Vol. 34, pp. 75–82, 2012.
for characterizing the electromagnetic properties of engi-
[250] J. Li, Y. Huang, L. Ma, Z. Yang, W. Lv, and G. Wen, neered materials,” Progress in Electromagnetics Research,
“Nonlinear metasurface antenna radome for power protec- Vol. 157, pp. 31–47, 2016.
tion application,” in 2019 8th Asia-Pacific Conference on
[263] A. Serdiukov, I. Semchenko, S. Tertyakov, and A. Si-
Antennas and Propagation (APCAP). IEEE, 2019, pp.
hvola, Electromagnetics of bi-anisotropic materials - The-
650–651.
ory and Application. Australia: Gordon and Breach
[251] A. Li, Z. Luo, H. Wakatsuchi, S. Kim, and D. Sieven- Science Publishers, 2001, Vol. 11.
piper, “Nonlinear, active, and tunable metasurfaces for
[264] D.E. Root, J. Verspecht, J. Horn, and M. Marcu, X-
advanced electromagnetics applications,” IEEE Access,
Parameters: Characterization, Modeling, and Design of
Vol. 5, pp. 27 439–27 452, 2017.
Nonlinear RF and Microwave Components. Cambridge:
[252] M.L. Bouknia, C. Zebiri, D. Sayad, I. Elfergani, M. Cambridge University Press, 2013.
Matin, M. Alibakhshikenari, A.G. Alharbi, Y.F. Hu, R. [265] A.K. Amert and K.W. Whites, “Characterization
Abd-Alhameed, J. Rodriguez, F. Falcone, and E. Limiti, of small specimens using an integrated computa-
“Analysis of gyrobianisotropic media effect on the input tional/measurement technique,” in 2013 IEEE Antennas
impedance, field distribution and mutual coupling of a and Propagation Society International Symposium (AP-
printed dipole antenna,” Scientific Reports, Vol. 12, No. 1, SURSI), 2013, pp. 706–707.
p. 10355, 2022.
[266] M.W. Hyde IV, M.J. Havrilla, and A.E. Bogle, “Non-
[253] I. V. Lindell, A. H. Sihvola, S. A. Tretyakov, , and destructive determination of the permittivity tensor of a
A. J. Viitanen, Electromagnetic Waves in Chiral and Bi- uniaxial material using a two-port clamped coaxial probe,”
Isotropic Media. Boston: Artech House, 1994. IEEE Transactions on Microwave Theory and Techniques,
[254] Mackay, T.G. and A. Lakhtakia, Electromagnetic Vol. 64, No. 1, pp. 239–246, 2016.
Anisotropy and Bianisotropy: A Field Guide. New Jersey:
World Scientific, 2010.
[255] Z. Guo, H. Jiang, and H. Chen, “Hyperbolic metama-
terials: From dispersion manipulation to applications,”
Journal of Applied Physics, Vol. 127, No. 7, p. 071101,
2020.
[256] R. Emadi, R. Safian, and A.Z. Nezhad, “Investigation
of saturation phenomena in spatially dispersive graphene-
based photoconductive antennas using hot-carriers theory,”
IEEE Journal of Quantum Electronics, Vol. 53, No. 5, pp.
1–8, 2017.
[257] D. Ramaccia, A. Toscano, and F. Bilotti, “Time-varying
metamaterials and metasurfaces for antennas and propaga-
tion applications,” Journal of Physics: Conference Series,
Vol. 2015, No. 1, p. 012121, 2021.
[258] C. Caloz and Z. Deck-Léger, “Spacetime metamaterials -
part I: General concepts,” IEEE Transactions on Antennas
and Propagation, Vol. 68, No. 3, pp. 1569–1582, 2020.
[259] ——, “Spacetime metamaterials, part ii: Theory and
applications,” IEEE Transactions on Antennas and Propa-
gation, Vol. 68, No. 3, pp. 1583–1598, 2020.
[260] A. Nagulu, X. Ni, A. Kord, M. Tymchenko, S. Garikap-
ati, A. Alù, and H. Krishnaswamy, “Chip-scale floquet
topological insulators for 5g wireless systems,” Nature
Electronics, Vol. 5, No. 5, pp. 300–309, 2022.

49