Fu Liu , Do-Hoon Kwon , and Sergei Tretyakov

Reflectarrays and
Metasurface Reflectors
as Diffraction Gratings
A tutorial.

R
econfigurable reflectors have significant potential in INTRODUCTION
future telecommunication systems, and approaches to During the past few years, many research groups have been
the design and realization of full and tunable reflec- studying the possible use of reconfigurable intelligent surfaces
tion control are now actively studied. Reflectarrays, (RISs) in future wireless communication systems [1], [2], [3],
being the classical approach to realize scanning reflectors, are [4], [5], [6], [7], [8], [9], [10]. The main functionality of these
based on the phased-array theory (the so-called general- RISs is to reflect incident waves (coming from a specified
ized reflection law) and physical optics approxima- direction or directions) into desired directions.
tion of the reflection response. To overcome Basically, this is the same function as usu-
the limitations of the reflectarray technol- ally realized by reflectarray antennas. Most
ogy, researchers actively study inho- commonly, reflectarrays are used as flat
mogeneous metasurfaces, using the or conformal equivalents of parabolic
theory of diffraction gratings. To reflectors, while RISs are usually
make these devices tunable and designed to reflect plane waves,
fully realize their potential, it but this is not a principal differ-
is necessary to unify the two ence. Such RISs are equivalent
approaches and study recon- to focusing reflectors with an
figurable reflectors from a infinite focal distance.
unified point of view. Here, Realizations of anoma-
we offer a tutorial on reflec- lously reflecting metasur-
tarrays and metasurface faces are usually designed
reflectors, explaining their as phase gradient reflec-
common fundamental tors, which are made
properties that stem from of reactive impedance
the diffraction theory. boundaries with a lin-
This tutorial is suitable for early varying phase of
graduate and postgradu- the local reflection coef-
ate students and hopefully ficient. However, recent
will help to develop deep- research has shown that
er understanding of both such realizations have a
phased arrays and diffrac- fundamentally limited
tion gratings. efficiency, which degrades
when the desired perfor-
Digital Object Identifier 10.1109/MAP.2023.3236278 mance significantly deviates
Date of current version: 7 February 2023 from that of uniform mirrors
IMAGE LICENSED BY INGRAM PUBLISHING

IEEE ANTENNAS
licensed use limited & PROPAGATION
to: SOCIETY MAGAZINE
FOR APPLIED J U NENG
MICROWAVE ELEC E 2 0AND
2 3 RES. Downloaded on June 22,2023 at 08:38:55
1045-9243/23©2023IEEE 21apply.
UTC from IEEE Xplore. Restrictions
 and retroreflectors [11], [12], [13], reflector are excited to create the
[14], [15]. This degradation is attrib- desired reflected fields. However,
uted to the excitation of parasitic In the design first, one needs to know what cur-
propagating modes that scatter part rent distribution is necessary to be
of the incident power in unwant- of reflectarrays realized. To this end, we discuss
ed directions. Actually, similar and reflecting active arrays [23], assuming that
effects are known also for reflec- we can fix any desirable current
tarrays, which function well only metasurfaces, the main distribution over a planar surface.
if the reflected rays do not have to challenge is to ensure Our goal is to determine what cur-
be tilted much. When the deflec- rent distribution we should set to
tion angle is large (i.e., elements are that the proper currents create the desired propagating
close to the reflectarray edge), the on the reflector are waves. Here, it is enough to con-
specular reflection is not controlled sider sheets of electric surface cur-
and considered wasted, lowering excited to create the rents. Then, we are able to properly
the efficiency [16, Sec. 4.1.4]. The desired reflected fields. determine the needed current pro-
physical reason behind a deterio- file over the reflector by using
rating reflection efficiency with an this simple model, although these
increasing deflection angle is the wave impedance mismatch sheets create waves on both sides (a ground plane or a comple-
between the incident and deflected plane waves. Recognized mentary sheet of magnetic surface current can be introduced to
first in perfect anomalously refracting metasurfaces [13], [15], realize one-side excitation).
[17], this wave impedance mismatch for efficiency reduction is For simplicity, we consider infinite arrays, and our desired
analogous to the angle-dependent scan impedance mismatch reflected modes are plane waves. For infinite arrays, the most
in phased arrays. While it can be tolerated for conventional common design goal is to ensure that in the regions far enough
applications of reflectarrays, for the envisaged use of anomalous from the array, where all the evanescent fields can be neglected,
reflectors as RISs, this limitation can significantly compromise there is only one plane wave propagating in the desired direc-
usability. Indeed, most usage scenarios of RISs assume that the tion, corresponding to a delta function array factor. For finite
reflected waves can be sent in any direction. arrays, this goal is equivalent to the radiation pattern having
Recently, it was shown that advanced metasurfaces can con- only one main beam in the desired direction, without any grat-
trol reflection theoretically perfectly, without any spurious scat- ing lobes. In this sense, conclusions made for infinite arrays will
tering (except that caused by manufacturing imperfections and hold also for finite arrays.
dissipation losses), e.g., [14], [15], [18], [19], [20], [21], and [22].
Different design approaches have been developed (we summa- THE PERIOD OF THE RADIATING CURRENT DISTRIBUTION
rize and discuss them in the “Reflectarrays and Metamirrors” FOR A GIVEN RADIATION DIRECTION
section). Interestingly, all of them are based on the theory of dif- Let us suppose that the reflected field that we want to cre-
fraction gratings and do not use the conventional design meth- ate in the far zone is a set of propagating plane waves. We
ods and topologies that have been developed for phased arrays assume that this desired set of plane waves varies along the
and reflectarray antennas. planar radiating surface (the coordinate x) as a periodic func-
For the proper understanding and further development tion. This means that the tangential wavenumbers (along x)
of devices for full and efficient control of wave reflection, of all radiated harmonics are in rational relations. The very
it appears that it is necessary to analyze the basic principle important special case is when only one obliquely propagating
of inhomogeneous reflectors, looking at both metasurfaces plane wave is launched in angle i (referenced to the surface
(realized with subwavelength structures) and reflectarrays normal) with tangential wavenumber k t = k 0 sin i, where
(formed by repeating antenna elements at half-wavelength k 0 = 2r/m is the free space wavenumber and m is the wave-
intervals) from a unified point of view. While these two length. Then, the x-dependence of this field e -jk t x is a peri-
techniques are different, they have fundamental similarities: odic function, with period D = m/sin i. The case of launching
both can be considered diffraction gratings. In this basic tuto- aperiodically distributed fields can be, in principle, treated as
rial article, from the diffraction grating theory, we explain the a limiting case of the infinite period. Later, we also discuss
fundamental principles behind any device that creates plane possibilities to launch a single plane wave with aperiodic cur-
waves propagating in a certain direction. In the final section, rent distributions.
we summarize and classify the currently known methods to It is clear that the required current distribution should be
design and realize anomalous reflectors and discuss current in a phase synchronism with all the desired free space modes.
research challenges. Assuming that the radiating current distribution is periodic,
with period D, it can be expanded into spatial Fourier series
ACTIVE ARRAYS with tangential wavenumbers:
In the design of reflectarrays and reflecting metasurfaces, the
k tn = 2rn , n = 0, !1, !2, f .(1)
main challenge is to ensure that the proper currents on the D

licensed use22 J UJune

limited to: SOCIETY FOR APPLIED MICROWAVE ELEC ENG AND RES. Downloaded on N E 22,2023 IEEE ANTENNAS
2 0 2 3 at 08:38:55 UTC from & PROPAGATION
IEEE MAGAZINE
Xplore. Restrictions apply.
 In general, we should select the period D so that the tan- because D 2 m; see (4). The next task is to set such a current
gential wavenumbers of all desired plane waves will be found distribution so that the waves in all the unwanted directions
among this set of numbers (for some values of index n). Let us have zero amplitude.
consider the special case of launching only one plane wave at a Let us denote the current distribution as J (x), where x is
certain angle i. In this case, it is enough to properly set only one the coordinate along the antenna array plane. In our example of
Fourier harmonic of the current. The tangential wavenumber of exciting a single plane wave, the tangential fields of the desired
the desired plane wave is plane wave mode vary as e -jk 0 x sin i . As discussed in the preced-
ing, the current distribution on the antenna should be a periodic
k t = k 0 sin i = 2r sin i .(2) function with the same period, D = m/ sin i. But what specific
m
function with this period should we select? An appropriate cur-
We need to select D so that at least one harmonic of the rent distribution can be found by using the theory of waveguide
current distribution is in phase with the desired radiated wave. excitation, e.g., [24, Sec. 4.12]. The fields excited in a waveguide
Comparing (1) and (2), the condition reads n/D = sin i/m; that by a given distribution of external currents are found as an
is, D = nm/sin i. Since we want to send the energy in only one expansion over the waveguide eigenmodes. In our case, the
direction, it is reasonable to choose n as small as possible (that waveguide is free space, and its modes are plane waves that can,
is, D as small as possible) to minimize the number of diffraction in general, propagate at any angle i. The amplitudes of excited
maxima or “open channels” (the directions where the array can plane wave modes are proportional to the excitation integrals
radiate). Here, n = 0 is not a valid solution since in that case, (integrals of the current distribution and complex conjugate
k t = 0, so we select n = 1, which gives of the desired radiated field distribution, [24, eq. 4.108]). It is
enough to integrate over one period.
m
D= .(3) Continuing discussing the specific example of launching a
sin i
single plane wave along the i = 70° direction, where the period
In this case, the period of the radiating current is equal to the D . 1.0642m, in fact, allows three propagating plane wave
period of variations of the fields in the plane wave that we want modes, the current distribution J (x) should be such that
to create. Obviously, this is an expected result.
Very importantly, we note that the array period D is greater #0 D J(x) e jk x sin i dx
0
" maximum (5)
than or equal to m for any angle i. The limit D " m corresponds
to i " r/2, that is, to the endfire array. For small angles (radia- #0 D J(x) e -jk x sin i dx = 0,
0
#0 D J(x) dx = 0 .(6)
tion directions close to the normal), the period is very large. As a
specific numerical example, we consider arrays that create a single These are the coupling integrals for all three allowed propa-
plane wave in the i = 70° direction, for which D . 1.0642m. gating modes with the selected period of the array. These condi-
We have noted that it is desirable to select as small a period tions ensure that the current distribution effectively launches
D as possible. More specifically, for a given period D, all har- the desired wave and that it is orthogonal to all the other exist-
monics whose tangential wavenumber k tn satisfies the inequal- ing propagating harmonics. Importantly, we do not impose any
ity | k tn | = 2r | n|/D 1 k 0 = 2r/m are propagating modes. This restrictions on the amplitudes of evanescent fields in the vicin-
corresponds to ity of the array, as the goal is to create the desired field in the
far zone. However, note that in reflectarray and metasurface
| n| 1 D .(4) realizations, evanescent modes are to be controlled to optimize
m
far-zone fields, and we briefly discuss this in the “Reflectarrays
In our example case of i = 70°, we have | n| 1 1.0642. This and Metamirrors” section.
means that if we set any periodic current on the antenna (using Clearly, there are many possible solutions of J (x) satisfying
active sources) with this period, the radiation in the far zone (5) and (6). The simplest and most obvious one is the current
can go only to 0, 70°, and -70° directions (corresponding to with a constant amplitude and linear phase gradient. For exam-
n = 0, 1, and -1, respectively). All higher-order harmonics are ple, selecting J (x) = Ae -jk 0 x sin i (A is the complex amplitude),
evanescent. Note, however, that for scanning in other directions, we maximize the integral in (5) because the product of the two
the situation can be very different since we may need a rather exponentials is just unity. The integrals in (6) are zero because of
large D compared with m. double/single variations over the period. This is why the current
We see that from the diffraction theory point of view, any does not radiate into -70° and normal directions.
periodic antenna array that radiates in any direction except Let us consider another case with a small tilt angle, say,
the normal direction is a diffraction grating because its period i = 5°. We use the same simple theory and first select the
D is larger than the wavelength, and it radiates into an n ! 0 period such that the array radiates the plane wave along this
spatial harmonic. direction. Now the appropriate period is D = m/ sin i . 11.47m.
The plane wave Fourier harmonics will propagate when
THE OPTIMAL CURRENT DISTRIBUTION | n| 1 D/m = 11.47; see (4). Thus, our desired direction corre-
We see that any periodic reflectarray (radiating current dis- sponds to n = 1, but there are 22 more directions (11 on the left,
tribution) can reflect (radiate) in more than one direction 10 on the right side from the normal, and the normal direction)

licensed use IEEE ANTENNAS

limited & PROPAGATION
to: SOCIETY MAGAZINE
FOR APPLIED J U N ENG
MICROWAVE ELEC E 2 0AND 23apply.
2 3 RES. Downloaded on June 22,2023 at 08:38:55 UTC from IEEE Xplore. Restrictions
 where the waves can also propagate. The current distribution becomes J (x) = Ad (x - a) + Bd (x - b), where A and B are the
should be such that complex amplitudes of the two point sources (line sources in the
2D scenario). Substituting this current distribution into (6), we
#0 D J(x) e jk x sin i dx
0
" maximum (7) get only a trivial solution A = B = 0.
This means that we need at least three elements per period.
#0 D J(x) e jnk x sin i dx = 0,
0
n = 0, -1, !2, !3, f .(8) Let us assume

The current distribution with the ideal linear phase J (x) = Ad (x) + Bd ` x - D j + Cd ` x - 2D j (9)
3 3
gradient J (x) = Ae -jk 0 x sin i clearly satisfies all these rela-
tions. For approximate (e.g., step-wise) settings, finding and substitute it for (6); we get
solutions for (7) and (8) is a nontrivial task. On the other
B = - A ^1 + e j 3 h = Ae -j 3 ,
2r 2r 2r
hand, for small angles, the required linear phase variation C = -(A + B) = Ae j 3 (10)

is smooth, and the coupling to other propagating harmon-

ics is weak. Thus, reasonably smooth discretization of the to ensure that there is no radiation in the normal and -70°
linear phase current creates nearly perfect single-wave directions. We see that in this equal spacing configuration, the
fields in the far zone. three line sources have the same amplitude and a linear phase
drop (a total of 2r over one period). By substituting this solution
REALIZATION WITH SMALL RADIATING ELEMENTS into (5), we find that the result is 3A (nonzero), meaning that in
(LINEAR PHASE GRADIENT) the far zone, there is a perfect plane wave in the desired direc-
Next, we discuss the principle of diffraction grating (meta- tion. The positions of the small radiators can be varied, adding
grating) realizations [18], [25], [26], [27], [28]. Instead of a degrees of freedom in design. Here, we note that a recent paper
continuous current distribution, the diffraction grating real- [29] studied the excitation of plane waves by arrays of line cur-
izations excite the desired plane wave, with just a few small rents. In that paper, excitation conditions are also imposed on
radiating elements placed in each period. To understand how evanescent modes, resulting in the conclusion that only continu-
it works, we assume that the current distribution over the ous current distributions with a linear phase gradient can launch
antenna plane is a set of a few delta function sources. That is, a single plane wave.
we consider an array of small radiators with an isotropic ele- The preceding results are verified with numerical simula-
ment pattern. tions, as presented in Figure 1. When the three line sources (2D
Let us try to reach the goal for radiation in i = 70° by using point sources in the figure) are assigned according to (9) and
only two such small active radiators per period, with arbitrary (10), with A = 1 mA, indeed, there is a perfect plane wave gen-
selected positions x = a and b. The current distribution function erated in the desired direction i that is larger than the critical

PML

1.5 0.6 0.6

D

5λ 0 0 0

–1.5 –0.6 –0.6

PML

(a) (b) (c) (d)

FIGURE 1. The radiation from three discrete radiators, with the current distribution given by (9) and (10). Note that the
supercell shown here has a lateral shift along the x-axis. (a) The simulation setup, where the currents are flowing in the out-
of-screen direction. (b)–(d) The scattered electric field patterns for angles i = 70, 40, and 25°, respectively. PML: perfectly
matched layer.

licensed use24 J UJune

limited to: SOCIETY FOR APPLIED MICROWAVE ELEC ENG AND RES. Downloaded on N E 22,2023 IEEE ANTENNAS
2 0 2 3 at 08:38:55 UTC from & PROPAGATION
IEEE MAGAZINE
Xplore. Restrictions apply.
 angle 30° in Figure 1(b) and (c). The critical angles correspond phase profile in the step-wise fashion. In this case, we can write
to D/m ! N (i = 30° gives D/m = 2), which opens up more the current distribution as
propagating modes. However, when i # 30°, the scattered field
is not a single plane wave, as more channels are open, as demon- Floor ` xN j
J (N, x) = Ae jU (N, x), with U (N, x) = - 2r D - dz
strated in Figure 1(d). The working angles i 2 30° correspond N
(12)
to | n| 1 D/m 1 2, which means that three point sources are
enough for generating perfect plane waves when only three where A is the complex amplitude, N is the segment number in
modes | n| 1 2 are propagating modes. each period D, U (N, x) is the phase function corresponding to
For the case of many propagating modes, the design compli- the discretized phase profile, dz is a phase shift factor, and the
cates. For the preceding example of radiation in the direction of function Floor(x) gives the greatest integer smaller than or equal
i = 5°, we need to satisfy a system of 23 equations. However, to the argument x.
we can use a simple approach based on (9) and (10). Similarly, With such a segmented current distribution, we can calcu-
for N discrete sources that are evenly distributed over D, we can late the integral in (8), and the results are listed in Table 1 for
assume that the current distribution follows A = 1 and dz = 0. The value will indicate whether there is a
plane wave propagating to each port n. As one can see, when
Ae -j2r N d ` x - Dm j (11)
N-1
J (N, x) = /
m N = 3, the integration is nonzero for ports n = - 5, - 2, 1, 4,
m=0
N and 7, meaning that if these ports correspond to propagating
modes (depending on angle i), there will be plane waves propa-
i.e., with the same amplitude and a linear phase drop of 2r over gating to those ports. As the number of segments N increases,
D. Associated with a uniform linear array of isotropic radiators more ports are suppressed, corresponding to zero integrals, and
with a linear phase, the radiation pattern of the source current only the port n = 1 is excited. Indeed, as N increases, the seg-
(11) corresponds to the array factor in array theory [30]. Then, mented current distribution is closer to the analytical uniform
the radiation will be perfect if we select enough discrete points one with a linear phase. We further note that a change of dz in
N, ensuring that integration in (8) satisfies the orthogonality (12) does not change the integration results at all, meaning that
conditions for all n that correspond to unwanted propagating the phase can be arbitrarily shifted.
channels. In general, if the periodicity allows propagation of Actually, the point sources and uniform current segments
2m + 1 plane waves, one needs m + 2 discrete sources to may be replaced by suitable array elements, say, with currents
excite only one of these plane waves. This is also true for equally of square shape or cosine shape. We check whether the radia-
spaced segments, and it can be concluded from Table 1 in the tion remains perfect for the same element number per period.
next section. Instead of (9) or (12), we assume

REALIZATION WITH RADIATING SEGMENTS J (x) = AF (x) + BF ` x - D j + CF ` x - 2 D j (13)

3 3
(LINEAR PHASE GRADIENT)
In actual realizations of antenna arrays, for example, using patch where F (x) is the normalized current distribution function at
antennas, the radiating currents are not point sources; rather, each array element and A, B, and C are the complex amplitudes
they are small radiating elements. To analyze this case, we con- of the currents at the array elements. The radiation pattern of
sider the discretization of the current distribution into segments. one array element with the current distribution F (x) is the ele-
In each segment, the current is uniform, while for different seg- ment pattern. For example, we can assume a square shape or a
ments, currents have different phases, still following the linear sine shape

TABLE 1. THE ABSOLUTE VALUE OF INTEGRALS (8) FOR THE CURRENT DISTRIBUTION (12),
WITH PORT NUMBER n AND SEGMENT NUMBER N.
n
N
−7 −6 −5 −4 −3 −2 −1 0 1 2 3 4 5 6 7
2 0.09D 0 0.13D 0 0.21D 0 0.64D 0 0.64D 0 0.21D 0 0.13D 0 0.09D
3 0 0 0.16D 0 0 0.42D 0 0 0.83D 0 0 0.21D 0 0 0.12D
4 0.13D 0 0 0 0.3D 0 0 0 0.91D 0 0 0 0.18D 0 0
5 0 0 0 0.24D 0 0 0 0 0.94D 0 0 0 0 0.16D 0
6 0 0 0.2D 0 0 0 0 0 0.96D 0 0 0 0 0 0.14D
7 0 0.17D 0 0 0 0 0 0 0.97D 0 0 0 0 0 0
8 0.14D 0 0 0 0 0 0 0 0.97D 0 0 0 0 0 0
9 0 0 0 0 0 0 0 0 0.98D 0 0 0 0 0 0

licensed use IEEE ANTENNAS

limited & PROPAGATION
to: SOCIETY MAGAZINE
FOR APPLIED J U N ENG
MICROWAVE ELEC E 2 0AND 25apply.
2 3 RES. Downloaded on June 22,2023 at 08:38:55 UTC from IEEE Xplore. Restrictions
 sin c rb m
1 if 0 1 x 1 w sin (rx/w) if 0 1 x 1 w
F (x ) = ' , F (x ) = '  D sin ` ra j jrb
0 otherwise 0 otherwise B= A e-j (r - r a/D), C = A D e D .
(14)
sin ; E sin ; E
r (b - a ) r (b - a)
D D (16)

where w is the width of the array element and is smaller than

D/3. Then, following the same approach as before, i.e., solving It is clear that when the radiating elements are not equal-
(6) with the current in (13) and (14), we find that the condition ly spaced, the required phase distribution does not follow
for radiation only to the desired angle (n = 1) is when the com- the linear phase gradient law. In addition, the amplitudes
plex amplitudes are B = Ae -j2r/3 and C = Ae j2r/3, which are of the currents are also changed. For example, when we
independent of w. We note that this solution is exactly the same set a = D/6 and b = 2D/3, as in Figure 2(a), the required
as the solution (10) for point source realization in the “Realiza- complex amplitudes of the second and third sources read
tion With Small Radiating Elements (Linear Phase Gradient)” B = (- 3/4 - j 3 /4) A and C = (-1/4 + j 3 /4) A, and with
section, clearly indicating that the point sources and uniform this setting, we can still get the desired single-wave radiation,
current sections can be replaced by other suitable radiating as illustrated in Figure 2(a).
elements. This agrees with the principle of pattern multiplica- If we set b = 2a in (16), we find the required com-
tion in arrays [30]. The limitation is that the element pattern of plex a mpl it udes a s B = 2A cos (ra D) e -j (r - r a/D) a nd
these elements should not have a null in the desired direction C = Ae j2 ra D . If we further make a much smaller than D so
of radiation. that the three sources are clustered, the phase quickly varies
among the three sources; i.e., the first and third sources have
GENERAL PERIODIC CURRENT small phases of zero and 2ra/D (a % D), while the middle
DISTRIBUTIONS (METAGRATINGS) source has a phase that is close to - r. Moreover, the ampli-
It is important to note that the radiating elements of periodic tude of the middle source is close to two, while the amplitudes
arrays do not have to be equally spaced, and the phase shifts of the other two are one. As examples, in Figure 2(b) and (c),
between the elements do not have to follow the linear phase we perform two sets of simulations, with b = 2a = D/4 and
advance law. Indeed, instead of (9), let us assume b = 2a = D/10, respectively. The results show that these clus-
tered sources can still generate the desired radiation without
J (x) = Ad (x) + Bd (x - a) + Cd (x - b) (15) any unwanted scatterings. Anomalous reflectors based on peri-
odic arrays of a few small scatterers in each period are often
where 0 1 a and b 1 D. By substituting it into (6) to elimi- called metagratings [18].
nate the wave radiation in the normal and - i directions We note that it is straightforward to analyze similarly the
(n = 0 and n = -1), we find that the solution of B and C, in case of small tilt angles with more discrete point sources. Obvi-
terms of A, reads ously, the realization of such fast variations of the reflection

PML

D
θ 0.3 0.2 0.04

5λ 0 0 0

–0.3 –0.2 –0.04

0 a b 0ab 0ab

PML

(a) (b) (c)

FIGURE 2. The radiation from N = 3 discrete current sources that are arbitrarily positioned in one period (with a lateral shift
along x). Each subfigure shows the simulation setup and radiated electric field pattern for i = 40°. The assumed values are
(a) a = D/6 and b = 2D/3, (b) b = 2a = D/4, and (c) b = 2a = D/10. The complex amplitudes B and C are obtained from (16),
with A = 1mA.

licensed use26 J UJune

limited to: SOCIETY FOR APPLIED MICROWAVE ELEC ENG AND RES. Downloaded on N E 22,2023 IEEE ANTENNAS
2 0 2 3 at 08:38:55 UTC from & PROPAGATION
IEEE MAGAZINE
Xplore. Restrictions apply.
 phase in reflectarrays is challenging. Thus, for the case when element realizations all have the linear phase profile, while for
many plane waves can propagate, it is preferred to use equal the nonequally spaced element design, the phase is not a linear
spacing and a linear phase gradient. 2r phase dropping over the period, and the amplitude is not a
In fact, these conclusions hold true not only for small point- constant in Figure 3(c).
like radiating elements but also for arrays of radiating segments.
For example, if we assume the current distribution function as DIFFRACTION GRATING APPROACH VERSUS
PHASED-ARRAY APPROACH
J (x) = AF (x) + BF (x - a) + CF (x - b) (17) In the preceding theory, we started from stating that current
distributions for launching a single plane wave in a desired
similar to (15) for the point sources, the required amplitudes are direction should be periodic functions, with the period related
also found to be given by (16) and independent of the segment to the period of the radiated plane wave. One reason for that
width w. For infinite arrays, the performance should not depend assumption is that the excitation current and desired field must
on the current distribution function and element pattern if the be in a phase synchronism. The other reason is that by fixing the
amplitudes A, B, and C are set properly and the element pattern period as defined by the desired reflection angle or angles, one
has no null in the desired radiation direction. makes sure that the reflector can create a plane wave exactly in
The assumption of a real-valued function F (x) corresponds the desired direction. The design goal is then to properly dis-
to the case of small resonant antennas as array elements. This is tribute the reflected power among the allowed “open channels,”
the case when the current distribution over a unit element is a minimizing scattering into all directions except the desired
standing wave, and the current phase is uniform over the unit cell. one. From the practical point of view, the design of periodic
Also, arrays of slots in metal sheets can be modeled in this way. structures reduces to the design of only one period, which is a
However, it is possible to assume that the current distribution is great simplification.
a complex-valued function, and the conclusions will not change. However, the current distribution does not have to be a peri-
It is obvious that these models assume that the current odic function. Conditions (5) and (6) as well as (7) and (8) can be
distribution over each array element is fixed and independent considered conditions on the spatial Fourier transforms of the
from the excitation of this element and all the other elements current distributions if we assume that J (x) is not necessarily
in the array. In the case of reflectarrays and metasurfaces, this periodic and extend the integration over the whole x-axis. Actu-
is equivalent to the assumption that the distribution of near ally, Floquet periodic functions of the form
fields in the vicinity of the array does not depend on the inci- 3
/ A m e -j`k sin i + j
2rm x
dence angle. J (x) = 0
d (18)
m = -3
To give a clear overview of the different models of the dif-
fraction grating approach, in Figure 3, we summarize the can satisfy the required conditions for launching only one
different current distribution functions discussed in the preced- plane wave in the i direction. Formula (18) corresponds to an
ing. As we can see in Figure 3(a) and (b), the equally spaced ideal current profile e -jk 0 x sin i that is modulated by an arbitrary

1 1 2 Case 1 Case 2
|J |

|J |

|J |

Ideal Current
0 Three Radiators 0 0

0 D 0 D 0 D

Square Shape
Five Segments
0 0 Sin Shape 0 Case 3
∠J

∠J

∠J

–π –π –π
–2π –2π –2π
0 D 0 D 0 D
x x x
(a) (b) (c)

FIGURE 3. The required current distribution functions (amplitude and phase) for the diffraction grating approach. (a) The
ideal current J (x) = Ae -jk0 x sin i in the “The Optimal Current Distribution” section, the required currents for three radiators
is given in (9) and (10) for Figure 1 [or, equivalently, (11), with N = 3] in the “Realization With Small Radiating Elements
(Linear Phase Gradient)” section, and that for five segments with the current in (12), with N = 5 and dz = 0 in the
“Realization With Radiating Segments (Linear Phase Gradient)” section. (b) Currents with the form of (13) and (14), with
patches showing the square shape and sine shape F (x). (c) The nonequally spaced element realizations in the “General
Periodic Current Distributions (Metagratings)” section, using (15)–(17). Cases 1 and 2 correspond to the ones in Figure 2(a)
and (b). Case 3 is the “patch” realization in (17), with the sine shape current in (14) and a = 2D/5, b = 3D/4, and w = D/10.
Here, A = 1 is used in all figures.

licensed use IEEE ANTENNAS

limited & PROPAGATION
to: SOCIETY MAGAZINE
FOR APPLIED J U N ENG
MICROWAVE ELEC E 2 0AND 27apply.
2 3 RES. Downloaded on June 22,2023 at 08:38:55 UTC from IEEE Xplore. Restrictions
 periodic function F (x) with period d: J (x) = F (x) e -jk 0 x sin i . wave propagation directions, it is possible to excite only one
Expanding F (x) into a Fourier series, we arrive at (18). Really, propagating plane wave by properly phasing the currents at
the m = 0 term is the ideal current profile e -jk 0 x sin i (the uni- the phased-array elements.
form amplitude and linear phase gradient) that launches a single If the radiation direction is fixed to angle i and no scan
plane wave in the i direction. Indeed, this function is obvi- is required, by choosing m = - 1 in (20), the period d of
ously orthogonal to all other plane waves as eigenmodes of free the general periodic structure can be set for that specific
space because angle as

#-33 e -jk x sin i e jk x sin il dx = 0

0 0
for all il ! i. (19) d 1 m/ (1 + sin i) .(21)

Thus, if we ensure that all other members of the series (18) This is recognized as the condition for avoiding grat-
have this property, currents of the form (18) will excite only one ing lobes in arrays [30]. For instance, for our example of
propagating plane wave. To ensure this property, we demand i = 70°, we need d 1 0.5155m. This is also verified from
that k 0 sin i + 2rm d 2 k 0 for all m ! 0. In this case, the the results of Figure 1:
tangential components of the wave vectors of all harmonics with ■■ For the minimum required radiators N = 3 in each super-
m ! 0 are larger than k 0 sin il for all il , and the orthogonality cell, the critical angle is i = 30°, which gives the critical
condition (19) is satisfied. It is convenient to rewrite this condi- supercell period D = m/ sin i = 2m. This requires that the
tion as critical distance between radiators is d = D/3 = 2m/3,
which is consistent with (21), for general periodic struc-
sin i + m m 2 1 .(20) tures. For angles greater than this critical angle, (21) is
d
always valid.
Then, the condition for the required repeating period d of ■■ For N = 5, the critical angle is i = 14.48°, which gives
the current elements can be analyzed. For i " 0, the condition the critical supercell period D = m/ sin i = 4m. Thus, the
is satisfied if d 1 m (the most “dangerous” term is, obviously, the critical distance is d = D/5 = 4m/5, which is consistent
term m = - 1) . For i " r/2, the condition is true if d 1 m/2. with (21).
That is why for scanning phased arrays, the period d is conven- The periodic “modulation function” F (x) can be a com-
tionally chosen to be equal to m/2: this choice ensures that for plex-valued function. An important special case is the form
any scan angle, no parasitic diffraction lobes will be created for f (xl ) e jk 0 xl sin i, defined in the region - d/2 1 xl 1 d/2. Periodi-
infinite scanning phased arrays. cally repeating this modulation, we have the current in the Mth
The case when the current distribution is not a peri- period (where x = Md + xl ) as
odic function is equivalent to the limit case when the
period is infinity. As we see, when the period approaches J (x) = F (x) e -jk 0 x sin i = f (xl ) e jk 0 (x - Md) sin i e -jk 0 x sin i
infinity, although there are infinitely many allowed plane = f (xl ) e -jMdk 0 sin i. (22)

I (x ) I (x + D) = I (x )
…… ……
x
λ
D= D Metasurface
sin θ
(a)

…… I1 I2 I3 I1 I2 I3 ……
x
λ Metagrating
D= D
sin θ
(b)

I (x ′) I (x ′)e jϕ I (x ′)e j 2ϕ I (x ′)e j3ϕ

…… ……
x
λ Phased Array
d= d
2
(c)

FIGURE 4. The (a) metasurface design approach (effectively continuous periodic current distribution), (b) metagrating design
approach (a few small scatterers in each period), and (c) phased-array design approach (in general, aperiodic current distribution).

licensed use28 J UJune

limited to: SOCIETY FOR APPLIED MICROWAVE ELEC ENG AND RES. Downloaded on N E 22,2023 IEEE ANTENNAS
2 0 2 3 at 08:38:55 UTC from & PROPAGATION
IEEE MAGAZINE
Xplore. Restrictions apply.
 This is the phased-array design approach. The structure is with n being the unit index, as shown in the figure. Importantly,
formed by an array of unit cells with size d (the array elements). we note that this current distribution is, in general, not a peri-
The unit cells have the same current distribution function f (xl ), odic function. The periodicity condition reads
but the phase of their current linearly varies from cell to cell.
nr sin i = 2rm (24)
The continuous linear phase profile is replaced by a discrete
one. The most common assumption is the real-valued function where m is the number of unit cells that will have the same
f (xl ), which corresponds to arrays of, e.g., resonant dipoles, phase as cell number zero. Obviously, this condition can be sat-
resonant patches, and horn antennas. isfied only if sin i is a rational number.
We see that the phased-array approach is quite different Figure 5 gives an example of the active phased array for gen-
from the conventional approach to the design of diffraction grat- erating radiation toward i = 70°. Due to the nonperiodicity of
ings and metasurfaces for reflection control. Here, we start from the current distribution, we consider a finite phased array made
fixing the unit cell size d, usually at d = m/2. Assuming the of N = 40 “patches” with width w = m/4 and spacing d = m/2.
same current distribution function over each unit cell (usually The current of this finite phased array is set to follow (22), with
the resonant mode of an antenna array element), we adjust the f (xl ) being a cosine shape [the inset of Figure 5(a)]. It indeed
phases of the unit cells to create constructive interference in the radiates in the desired direction, as shown by the electric field
desired direction. pattern in Figure 5(b) and the normalized far field (the black
Figure 4 illustrates and compares these two approaches. Fig- solid curve) in Figure 5(c). Due to the finiteness of the array,
ure 4(a) and (b) correspond to the diffraction grating approach there are side lobes but no other main beams. When increas-
for designing anomalous reflectors. In this case, we select a ing the radiating aperture size by adding more “patches,” the
proper period D = m/ sin i and then find a suitable current pattern tends to a delta function. This is illustrated by the red
distribution (the solution is not unique). For the case of the dis- dashed curve in Figure 5(c) for N = 100.
cretized point radiators in Figure 4(b), for launching the wave to If we choose the element spacing in the phased array equal
70°, where we have three propagating Floquet modes, we need to D/3 = (1/3) m/sin i, the phase array becomes periodic, and
three discrete sources in each period. it is equivalent to the diffraction grating picture for i 2 30°.
In Figure 4(c), which describes the phased-array Hence, for active arrays, the difference between diffraction
approach, we have an array of identical “patches” spaced by grating and phased-array approaches is of a practical nature in
m/2. We assume that all the elements have the same ampli- essence. In diffraction gratings, a periodic physical configura-
tude pattern (a cosine shape is shown) and that the phase is tion stipulated by the incidence and reflection angles is chosen
uniform over each unit cell. Then, we feed this array with a so that the design domain is reduced to a single supercell. In
linearly varying phase source so that the phases vary from phased arrays, the cell size (i.e., the element spacing) is deter-
unit to unit as mined a priori (usually m/2) based on the required beam scan-
ning capability. The functionality of creating one propagating
wave is performed by element excitations following a spatially
U (n) = k 0 x (n) sin i = 2r n m sin i = rn sin i (23)
m 2 linear phase profile.

PML
Re[J ] Im[J ] 90
1 1 120 60
J/A

0
–1 0.8
Normalized Electric Field

0 5 10 15 20 25 30 35 40 0.6 150 30
x /d 0.4
0 40 d 0.2
(a) 0 180 0
0.2
0.4
θ = 70° 0.6 210 330
3
0.8
0 1 240 300
270
–3 N = 40 N = 100
(b) (c)

FIGURE 5. The radiation from a finite phased array consisting of N = 40 “patches,” with w = m/4 and d = m/2. (a) The
simulation setup. The inset shows the current distribution. (b) The simulated electric field pattern for the N = 40 phased
array. (c) The normalized far field (electric field). The black solid (red dashed) line corresponds to the phased array
consisting of N = 40 (N = 100) “patches.”

licensed use IEEE ANTENNAS

limited & PROPAGATION
to: SOCIETY MAGAZINE
FOR APPLIED J U N ENG
MICROWAVE ELEC E 2 0AND 29apply.
2 3 RES. Downloaded on June 22,2023 at 08:38:55 UTC from IEEE Xplore. Restrictions
 The fundamental difference waves (and fields scattered from all
between the two design approaches other array elements). The usual
is in realizing scanning capability. In diffraction gratings, design aim is to synthesize the
For diffraction gratings, the peri- induced current distribution J (x)
od of the array should be adjusted
a periodic physical to create a constant amplitude lin-
for each desired reflection angle, configuration stipulated ear phase profile. However, unlike
which, in practice, requires enough in the active phased-array case, we
small subwavelength unit cells. For
by the incidence and do not have full control over the
phased arrays, the period is fixed, reflection angles is magnitude and phase of J (x) by
which, in principle, allows the using a passive surface subject for
use of m/2-sized unit cells. In this
chosen so that the the illumination of incident waves.
case, the required current distribu- design domain is The coupling of array elements and
tion over the plane is, in general, interference of incident and reflect-
aperiodic. This issue does not cre-
reduced to a single ed waves need to be carefully taken
ate practical problems for active supercell. into account.
phased arrays, which are all of a Let us first consider the design
transmitting type. However, the approach based on the diffraction
realization of the required linear phase profile for wide-angle grating theory. That is, we fix the array period as defined by the
scanning by adjusting the reactive bulk loads of passive reflec- incidence angle and desired reflection direction and then design
tarray antenna elements is rather difficult. a surface to perform the required operation. In the following
Let us stress that there is a very important simplifying discussion, we consider the same example functionality as in the
assumption in the root of the phased-array antenna theory. examples in the preceding: the reflection of a normally incident
Namely, it is assumed that the current distribution over one plane wave into a plane wave in an arbitrary direction. We can
array element does not depend on the currents on the other ele- consider two cases:
ments. The current density on the first element can be written 1) If we specify the scattered field as a single plane wave, for
as I 0 f (x), where f (x) = 0 at x 1 0, with x 2 d, and I 0 is the example, tilted at 70°, and calculate the required surface
current amplitude fixed by external sources. Only under this impedance Z s of the reflector, an alternately active/lossy sur-
assumption, the current of the Mth element is I M f (x - Md), face results [11], [13]. This is not desirable or even possible
where f is the same function. This assumption is an approxima- from the realization perspective.
tion. It holds very well for active resonant antenna elements 2) If we specify the periodic surface reactance to give a local
(when the current distribution is approximately fixed by the reflection coefficient profile with a unit magnitude and
resonant mode) and arrays of horn antennas (where the cur- linear phase gradient, passivity is guaranteed. However, the
rent, or aperture field, is very close to the fundamental mode reflected field ends up including parasitic propagating waves.
of the horn). We stress that the current amplitudes I M depend In our example, reflected power also goes into the - 70 and
on external voltages applied to all elements. By knowing the 0° directions, e.g., [14]. Other (and more than two) parasitic
impedance matrix of the array, we can always find such voltages scattering directions open up for other combinations of the
to realize the desired distribution of current amplitudes over the incidence and anomalous reflection angles.
antenna plane. Neither case is satisfactory. Researchers work on finding
Finally, let us discuss the role of diffraction grating theory other solutions that use only passive (as small a dissipation as
in understanding phased arrays. Although the current and possible) array elements. The only possibility is to allow the
field distribution over a reflectarray aperture is not periodic, excitation of other field modes beside the only plane wave in the
the array is “geometrically periodic,” as it is formed by identical desired direction. If all these other modes are evanescent Flo-
antennas separated by a m/2 distance. Moreover, the current quet harmonics, they do not propagate into the far zone, and the
distribution is Floquet periodic, defined by a periodic modula- perfect performance of the device is not compromised. There
tion function F (x) . Another aspect is that an array without peri- can be several ways to approach this task:
odicity of loads and induced current can be treated as a periodic 1) Model the reflector as a boundary defined by its surface
array with an infinite period (D = 3) . In this case, there are (input) impedance. Allowing the excitation of surface-bound
infinitely many propagating “Floquet harmonics” of the diffrac- modes, design the surface impedance so that only the desired
tion gratings theory. Generally speaking, the diffraction grating reflected propagating mode exists, enforcing Re [Z s (x)] = 0
method has limited applicability; however, some general conclu- as an additional condition. Finding Z s (x) for creating only
sions remain valid. one propagating plane wave is simple, but guaranteeing pas-
sivity is not straightforward. This approach has been used, for
REFLECTARRAYS AND METAMIRRORS example, in [19] and [22]. Importantly, actual realizations of
In contrast to active antenna arrays, where each antenna ele- the optimized surface impedance still require the locally
ment is fed by an external controllable source, currents on ele- periodic approximation, possibly leading to performance
ments of reflectarrays and metasurfaces are excited by incident degradation. For this reason, the method developed in [10]

licensed use30 J UJune

limited to: SOCIETY FOR APPLIED MICROWAVE ELEC ENG AND RES. Downloaded on N E 22,2023 IEEE ANTENNAS
2 0 2 3 at 08:38:55 UTC from & PROPAGATION
IEEE MAGAZINE
Xplore. Restrictions apply.
 imposes an additional condition of slow variation of the sur- to reach the goal, and we may need to use subwavelength ele-
face impedance, which can make the approximate design of ments, similar to the metasurface scenario.
surface topology more accurate. The main difficulty in the phased-array (reflectarray)
2) Start from a certain topology (usually a patterned metal sheet approach is that the array of antennas (including the loads) is
on a grounded dielectric substrate) and optimize the evanes- no longer a periodic structure. This makes the number of ports
cent fields so that the effective sheet impedance of the thin for connecting load impedances infinite, and all the loads are,
patterned metal layer is lossless at every point. This is the in general, different. In contrast, the periodic metasurface
approach used in [21] and [31] for scalar metasurfaces and approach requires the design and optimization of only one
[32] for tensor metasurfaces. In this approach, the imped- period, and usually only a few parameters need to be optimized.
ance model is used only for a single patterned sheet, not for We stress that this is not a problem in a conventional practical
the whole metasurface structure. In addition, the locally phased-array antenna in either transmission or receiving mode
periodic approximation is also used in this method but only because it corresponds to a transmissive device. A transmissive
for the design of the reactive sheet instead of the whole meta- device does not have the interference problem between the inci-
surface body. dent and scattered waves, which is the reason for the active/lossy
3) Considering arrays of thin metal strips on a grounded sub- nature of the required surface impedance of an ideal boundary
strate and using the periodic Green’s function for a grounded anomalously performing reflection.
dielectric substrate, find the loading impedances of the strips
analytically. Translate the required loading element into a CONCLUSIONS
gap capacitor, as is done, e.g., in [27] and [28]. This approach In recent literature on metasurfaces for the control of reflected
is a variation of the metagrating design method. This way, it waves, there have been many publications on the suggested use
is possible to treat anomalous transmission as well as reflec- of reconfigurable anomalous reflectors for the engineering and
tion [33]. optimization of the propagation environment. The vast majority
4) Design the current distribution so that only the desired plane of these works are based on the locally periodic approximation of
wave is created by direct optimization of the metasurface the array response. That is, a locally defined reflection coefficient
structure (not relying on sheet/surface impedance boundary is used for calculating the reflected fields, and it is assumed that
conditions and not requiring reactive input impedance at any the fields can be somehow controlled by changing the param-
surface except the ground plane). Passivity is guaranteed by eters of the unit cell at each position. This is the conventional
microscopically considering passive inclusions, but creating reflectarray antenna design assumption. It is known that this
only one plane wave is not as easy as in 1). This approach is local design gives acceptable performance for moderate tilt
used in [14] and [34]. angles (moderate deviations from the reflection law for uniform
5) Find a surface where the input impedance is purely reactive mirrors). However, for envisaged applications, it is required that
for the desired set of waves, and design the local surface reac- beams can be directed in any desired direction, which needs
tance X s (x) [20], [35]. This is the power flow-conformal solu- more advanced designs. The majority of the current research
tion, which does not need optimizations of the spatial in this field uses variations of periodic metasurfaces (diffraction
dispersion properties of the reflector. grating design methods). Here, impressive results have been
Let us next discuss the design approach based on the achieved in designing and realizing anomalous reflectors for
phased-array theory. That is, we fix the unit cell size, usually to fixed sets of incidence and reflection angles. The main problem
m/2. We position geometrically identical passive antenna ele- is the realization of electrical reconfigurability: in this design
ments in each cell and find such loads connected to the ele- approach, the period of the array is defined by the incidence and
ments so that the induced current distribution has the required reflection angles, which demands the change of the array period
linear phase profile [with a periodic modulation, as in (22)]. by adjusting the tunable components of small unit cells.
Under the simplifying assumption that the current distribu- On the other hand, the phased-array approach uses a set of
tion over each antenna element is the same for any incident field periodically arranged antennas, and the geometrical period is
distribution and any loads connected to all the elements, we can fixed for all scan angles. However, the distribution of control-
use the impedance matrix method to find the load impedances lable loads is not periodic in this case, which requires the global
for realizing the desired phase distribution. Unfortunately, we optimization of the whole array for each scan angle. Moreover,
run into the same problem as in the diffraction grating design: it is not known if the optimization of loads of a conventional
these loads have active/lossy behavior [35]. Thus, one needs to m/2 -spaced array can lead to acceptable performance for the
impose an additional constraint on the load impedances (zero requirement of an extremely wide scan angle range. Approaches
real parts) and design some optimization procedure for finding for optimizing the reactive loads of reflectarray antennas can be
reactive loads that approximate the required amplitude and found, e.g., in [35] and [36], but it appears that more research on
current distribution in the best possible way. Basically, we end this design approach is needed.
up with the same problem of engineering near fields so that the
power is properly channeled from “virtually lossy” to “virtually ACKNOWLEDGMENT
active” antenna elements (engineering spatial dispersion). It is This work was supported, in part, by the National Natural
not clear if just one control element per m/2 will be enough Science Foundation of China, under grant 12274339, the

licensed use IEEE ANTENNAS

limited & PROPAGATION
to: SOCIETY MAGAZINE
FOR APPLIED J U N ENG
MICROWAVE ELEC E 2 0AND 31apply.
2 3 RES. Downloaded on June 22,2023 at 08:38:55 UTC from IEEE Xplore. Restrictions
 European Integrated Training Network Meta Wireless and the [13] V. S. Asadchy, M. Albooyeh, S. N. Tcvetkova, A. Díaz-Rubio, Y. Ra’di, and
S. Tretyakov, “Perfect control of reflection and refraction using spatially disper-
U.S. Army Research Office, under grant W911NF-19-2-0244. sive metasurfaces,” Phys. Rev. B, vol. 94, no. 7, Aug. 2016, Art. no. 075142, doi:
Fu Liu also acknowledges financial support from Xi’an Jiaotong 10.1103/PhysRevB.94.075142.
University and Shaanxi Province, China. [14] A. Díaz-Rubio, V. Asadchy, A. Elsakka, and S. Tretyakov, “From the gen-
eralized reflection law to the realization of perfect anomalous reflectors,” Sci.
Adv., vol. 3, no. 8, Aug. 2017, Art. no. e1602714, doi: 10.1126/sciadv.1602714.
AUTHOR INFORMATION [15] V. S. Asadchy, A. Wickberg, A. Díaz-Rubio, and M. Wegener, “Eliminating
Fu Liu ([email protected]) is with the School of Electronic Sci- scattering loss in anomalously reflecting optical metasurfaces,” ACS Photon.,
vol. 4, no. 5, pp. 1264–1270, Apr. 2017, doi: 10.1021/acsphotonics.7b00213.
ence and Engineering, Faculty of Electronic and Information [16] J. Huang and J. A. Encinar, Reflectarray Antennas. Hoboken, NJ, USA:
Engineering, Xi’an Jiaotong University, Xi’an 710049, China. His Wiley, 2008.
research interests include complex artificial electromagnetic [17] J. P. S. Wong, A. Epstein, and G. V. Eleftheriades, “Reflectionless wide-
angle refracting metasurfaces,” IEEE Antennas Wireless Propag. Lett., vol. 15,
materials (metamaterials and metasurfaces), transformation pp. 1293–1296, 2016, doi: 10.1109/LAWP.2015.2505629.
optics, and wireless power transfer. He is a Member of IEEE. [18] Y. Ra’di, D. L. Sounas, and A. Alù, “Metagratings: Beyond the limits of
Do-Hoon Kwon ([email protected]) is with the Depart- graded metasurfaces for wave front control,” Phys. Rev. Lett., vol. 119, no. 6,
2018, Art. no. 067404, doi: 10.1103/PhysRevLett.119.067404.
ment of Electrical and Computer Engineering, University of [19] D.-H. Kwon, “Lossless scalar metasurfaces for anomalous reflection based
Massachusetts Amherst, Amherst, MA 01003 USA. His research on efficient surface field optimization,” IEEE Antennas Wireless Propag. Lett.,
interests include antenna and array bandwidth properties, vol. 17, no. 7, pp. 1149–1152, Jul. 2018, doi: 10.1109/LAWP.2018.2836299.
[20] A. Díaz-Rubio, J. Li, C. Shen, S. Cummer, and S. Tretyakov, “Power flow-
microwave metamaterials and metasurfaces, and cloaking. He is conformal metamirrors for engineering wave reflections,” Sci. Adv., vol. 5, no. 2,
a Senior Member of IEEE. Feb. 2019, Art. no. eaau7288, doi: 10.1126/sciadv.aau7288.
Sergei Tretyakov ([email protected]) is with the [21] X. Wang, A. Díaz-Rubio, and S. A. Tretyakov, “Independent control of
multiple channels in metasurface devices,” Phys. Rev. Appl., vol. 14, no. 2, Aug.
Department of Electronics and Nanoengineering, Aalto Universi- 2020, Art. no. 024089, doi: 10.1103/PhysRevApplied.14.024089.
ty, 02150 Espoo, Finland. His research interests include electro- [22] J. Budhu and A. Grbic, “Perfectly reflecting metasurface reflectarrays:
magnetic field theory, complex media electromagnetics, metama- Mutual coupling modeling between unique elements through homogeniza-
tion,” IEEE Trans. Antennas Propag., vol. 69, no. 1, pp. 122–134, Jan. 2021, doi:
terials, and microwave engineering. He is a Fellow of IEEE. 10.1109/TAP.2020.3001450.
[23] R. J. Mailloux, Phased Array Antenna Handbook, 3rd ed. Norwood, MA,
USA: Artech House, 2017.
REFERENCES [24] R. E. Collin, Foundations for Microwave Engineering. New York, NY, USA:
[1] C. Liaskos, S. Nie, A. Tsioliaridou, A. Pitsillides, S. Ioannidis, and I. Akyildiz,
IEEE Press, 2001.
“A new wireless communication paradigm through software-controlled meta-
[25] V. Popov, F. Boust, and S. N. Burokur, “Controlling diffraction patterns with
surfaces,” IEEE Commun. Mag., vol. 56, no. 9, pp. 162–169, Sep. 2018, doi:
metagratings,” Phys. Rev. Appl., vol. 10, no. 1, May 2018, Art. no. 011002, doi:
10.1109/MCOM.2018.1700659.
10.1103/PhysRevApplied.10.011002.
[2] M. Di Renzo et al., “Smart radio environments empowered by reconfigurable
[26] A. Epstein and O. Rabinovich, “Unveiling the properties of metagratings
ai meta-surfaces: An idea whose time has come,” EURASIP J. Wireless Commun.
via a detailed analytical model for synthesis and analysis,” Phys. Rev. Appl.,
Netw., vol. 2019, no. 1, pp. 1–20, May 2019, doi: 10.1186/s13638-019-1438-9.
vol. 8, no. 5, Aug. 2017, Art. no. 054037, doi: 10.1103/PhysRevApplied.8.054037.
[3] E. Basar, M. Di Renzo, J. De Rosny, M. Debbah, M.-S. Alouini, and
[27] O. Rabinovich and A. Epstein, “Analytical design of printed circuit
R. Zhang, “Wireless communications through reconfigurable intelligent sur-
board (PCB) metagratings for perfect anomalous reflection,” IEEE Trans.
faces,” IEEE Access, vol. 7, pp. 116,753–116,773, Aug. 2019, doi: 10.1109/
Antennas Propag., vol. 66, no. 8, pp. 4086–4095, Aug. 2018, doi: 10.1109/
ACCESS.2019.2935192.
TAP.2018.2836379.
[4] W. Tang et al., “The future of wireless? Subject editor spotlight on program-
[28] O. Rabinovich, I. Kaplon, J. Reis, and A. Epstein, “Experimental demon-
mable metasurfaces,” IET Electron. Lett., vol. 55, no. 7, pp. 360–361, Apr. 2019,
stration and in-depth investigation of analytically designed anomalous reflection
doi: 10.1049/el.2019.0471.
metagratings,” Phys. Rev. B, vol. 99, no. 12, Mar. 2019, Art. no. 125101, doi:
[5] Q. Wu and R. Zhang, “Towards smart and reconfigurable environment:
10.1103/PhysRevB.99.125101.
Intelligent reflecting surface aided wireless network,” IEEE Commun. Mag.,
[29] T. B. Hansen, “Array of line sources that produces preselected plane
vol. 58, no. 1, pp. 106–112, Jan. 2020, doi: 10.1109/MCOM.001.1900107.
waves,” Wave Mot., vol. 106, no. 20, Jun. 2021, Art. no. 102791, doi: 10.1016/j.
[6] Ö. Özdogan, E. Björnson, and E. G. Larsson, “Intelligent reflecting surfaces:
wavemoti.2021.102791.
Physics, propagation, and pathloss modeling,” IEEE Wireless Commun. Lett.,
[30] C. A. Balanis, Antenna Theory: Analysis and Design, 4th ed. Hoboken, NJ,
vol. 9, no. 5, pp. 581–585, May 2020, doi: 10.1109/LWC.2019.2960779.
USA: Wiley, 2016.
[7] C. Huang, A. Zappone, G. C. Alexandropoulos, M. Debbah, and C. Yuen,
[31] D.-H. Kwon, “Planar metasurface design for wide-angle refraction using
“Reconfigurable intelligent surfaces for energy efficiency in wireless commu-
interface field optimization,” IEEE Antennas Wireless Propag. Lett., vol. 20,
nication,” IEEE Trans. Wireless Commun., vol. 18, no. 8, pp. 4157–4170, Aug.
no. 4, pp. 428–432, Apr. 2021, doi: 10.1109/LAWP.2021.3050194.
2019, doi: 10.1109/TWC.2019.2922609.
[32] C. Yepes, M. Faenzi, S. Maci, and E. Martini, “Perfect non-specular reflec-
[8] Q. Wu and R. Zhang, “Intelligent reflecting surface enhanced wireless net-
tion with polarization control by using a locally passive metasurface sheet on
work via joint active and passive beamforming,” IEEE Trans. Wireless Commun.,
a grounded dielectric slab,” Appl. Phys. Lett., vol. 118, no. 23, Jun. 2021, Art.
vol. 18, no. 11, pp. 5394–5409, Nov. 2019, doi: 10.1109/TWC.2019.2936025.
no. 231601, doi: 10.1063/5.0048970.
[9] M. Di Renzo et al., “Smart radio environments empowered by reconfigu-
[33] O. Rabinovich and A. Epstein, “Arbitrary diffraction engineering with mul-
rable intelligent surfaces: How it works, state of research, and the road ahead,”
tilayered multielement metagratings,” IEEE Trans. Antennas Propag., vol. 68,
IEEE J. Sel. Areas Commun., vol. 38, no. 11, pp. 2450–2525, Nov. 2020, doi:
no. 3, pp. 1553–1568, Mar. 2020, doi: 10.1109/TAP.2019.2950813.
10.1109/JSAC.2020.3007211.
[34] A. M. H. Wong and G. V. Eleftheriades, “Perfect anomaloua reflection with
[10] M. Di Renzo, F. H. Danufane, and S. Tretyakov, “Communication models
a bipartite Huygens’ metasurface,” Phys. Rev. X, vol. 8, no. 1, Feb. 2018. Art.
for reconfigurable intelligent surfaces: From surface electromagnetics to wire-
no. 011036, doi: 10.1103/PhysRevX.8.011036.
less networks optimization,” Proc. IEEE, vol. 110, no. 9, pp. 1164–1209, Aug.
[35] O. Tereshin, V. Sedov, and A. Chaplin, Synthesis of Antennas on Wave-
2022, doi: 10.1109/JPROC.2022.3195536.
Delay Structures (in Russian). Moscow, Russia: Svyaz, 1980.
[11] N. M. Estakhri and A. Alù, “Wave-front transformation with gradient meta-
[36] G. Xu, V. G. Ataloglou, S. V. Hum, and G. V. Eleftheriades, “Extreme
surfaces,” Phys. Rev. X, vol. 6, no. 4, Oct. 2016, Art. no. 041008.
beam-forming with impedance metasurfaces featuring embedded sources and
[12] A. Epstein and G. V. Eleftheriades, “Synthesis of passive lossless metasur-
auxiliary surface wave optimization,” IEEE Access, vol. 10, pp. 28,760–28,684,
faces using auxiliary fields for reflectionless beam splitting and perfect reflec-
Mar. 2022, doi: 10.1109/ACCESS.2022.3157291.
tion,” Phys. Rev. Lett., vol. 117, no. 25, Dec. 2016, Art. no. 256103, doi: 10.1103/

PhysRevLett.117.256103.

licensed use32 J UJune

limited to: SOCIETY FOR APPLIED MICROWAVE ELEC ENG AND RES. Downloaded on N E 22,2023 IEEE ANTENNAS
2 0 2 3 at 08:38:55 UTC from & PROPAGATION
IEEE MAGAZINE
Xplore. Restrictions apply.