Tapio et al.

J Wireless Com Network (2021) 2021:175

https://doi.org/10.1186/s13638-021-02048-5

REVIEW Open Access

Survey on reconfigurable intelligent surfaces

below 10 GHz
Visa Tapio1* , Ibrahim Hemadeh2, Alain Mourad2, Arman Shojaeifard2 and Markku Juntti1

*Correspondence:
[email protected] Abstract
1
Centre for Wireless Reconfigurable intelligent surface (RIS) is a programmable structure that can be used
Communications, University
of Oulu, Oulu, Finland to control the propagation of electromagnetic waves by changing the electric and
Full list of author information magnetic properties of the surface. By placing these surfaces in an environment, the
is available at the end of the properties of radio channels can be controlled. This opens up new opportunities to
article
improve the performance of wireless systems. In this paper, the basic operation of
antenna array and metasurface based RIS is described. While the current long term (6G)
research on RIS often prioritizes very high frequencies from tens to hundreds of GHz,
this paper puts emphasis rather on operating frequencies below 10 GHz which prom-
ise a much faster to market track for RIS applications. For this purpose, review of the
literature on the use of RIS in wireless communication applications operating below
10 GHz frequency band is provided.
Keywords: Reconfigurable intelligent surface, Metasurface, 5G

1 Introduction
Reconfigurable intelligent surface (RIS)—known also as intelligent reflecting surface
(IRS), and large intelligent surface (LIS)—is a programmable structure that can be used
to control the propagation of electromagnetic (EM) waves by changing the electric and
magnetic properties of the surface. In addition to the control of EM waves, RISs can be
used to sense the radio environment by integrating sensing capabilities into them. By
placing intelligent surfaces in the environment where wireless systems are operating, the
properties of the radio channels can be controlled at least partially.
The control of radio channels changes the conventional wireless system design para-
digm, in which the radio channel is seen as an uncontrollable entity that distorts the
transmitted signals. Therein the transmitter (TX) and the receiver (RX) are designed to
equalize the impact of the channel. Envisioned scenarios vary from cases where a single
RIS is placed on a wall to direct signals coming from a predetermined direction, e.g.,
base station, to the environment where almost all surfaces (walls, furniture, clothes, etc.)
are covered with a metasurface based RIS [1–4]. The introduction of intelligent surfaces
can improve the reliability and energy efficiency of wireless systems. They may allow
accurate localization of people and objects in environments where, for example, satel-
lite-based positioning systems do not work properly [5, 6]. In radio equipment design,

© The Author(s), 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits
use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original
author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third
party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the mate-
rial. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or
exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​
creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 2 of 18

intelligent surfaces realized with metasurfaces are anticipated to enable the realization of
low-complexity and energy efficient transceivers that require only a limited, ideally one,
active radio frequency (RF) chain [2, 3, 7].
Two approaches can be used to implement an RIS that control the impinging sig-
nals characteristics (e.g., reflection, refraction, absorption, focusing and polarization),
namely, conventional antenna arrays and metasurfaces. In a nutshell, metasurfaces are
electrically thin and dense two-dimensional arrays of structural elements possessing
desired properties granted by their constitutive elements. Elements are called meta-cells,
unit-cells or meta-atoms. The meta-cells size is much smaller than the signal’s wave-
length () typically, between /10 and /5 [8, 9]. In addition to the control of EM propa-
gation, metasurfaces can be used to realize complex operations such as data modulation
and mathematical operations and they have been demonstrated to be capable of storing
EM pulses for short time periods in the order of 15 ns [10, 11].
The RIS is mainly considered to be a beyond 5G technology operating at frequencies
from tens of gigahertzs (GHz) to terahertzs (THz). At these frequency bands, the signal
propagation is heavily attenuated and can be blocked completely by the obstacles in the
propagation environment. Hence, the channels between a multi-antenna TX and RX are
modeled as sparse channels and the direct link between the TX and RX is often assumed
to be completely blocked. In these cases, the ability to form beams with moderate sized
surfaces enables communication between the TX and RX.

2 Research method
This paper is a state-of-the art review on the application of the RIS below 10 GHz fre-
quency. At these frequencies the channel between the TX and RX is not in general com-
pletely blocked by the obstacles and the channel is not modeled as a sparse channel. This
means that the role of the RIS is not to enable the communication between the TX and
RX as at higher frequencies, but to increase the capacity, although in some cases the cov-
erage extension by directing the transmitted signal behind an obstacle such as a building
in urban environment or a hill in sub-urban/rural environment is applicable also at sub-
10 GHz frequencies. The purpose of this review is to explore the RIS implementation
options for the existing 5G networks operating at the 5G NR Frequency Range 1 (FR1),
but the survey is extended to frequencies up to 10 GHz.

3 Reflecting antenna arrays in RIS applications

3.1 Operation principle
Electrically tunable reflecting antenna arrays can be used to dynamically adjust
their radiation patterns [12]. Hence, within the context of RIS, antenna arrays can be
employed as reflecting structures. Specifically, a large number of low-cost and passive
antenna elements are employed, which operate as reflectors with no need for RF chains
as in the conventional multiple-input multiple-output (MIMO) systems using multiple
transmit and multiple receive antennas. To expound, Fig. 1 describes the operation of
an antenna array-based RIS, which relies on microstrip patch antennas. Note here that
other antenna types can also be used as RIS reflectors.
The structure in Fig. 1 consists of a dielectric substrate with relative permittiv-
ity ǫr placed on a metallic ground plane. Metallic patch elements are placed on the
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 3 of 18

Fig. 1 Patch antenna array as a reflector

other side of the dielectric substrate. The inter-spacing distance between the metal
patches in the array of Fig. 1 is typically in the order of /2 (  = wavelength). A sin-
gle patch antenna element can be modeled as a cavity resonator. Such a model con-
sists of the patch, the dielectric directly below the patch, the ground plane as well
as the virtual vertical walls (vertical dashed lines) shown in Fig. 1 [13]. When elec-
tromagnetic energy, with frequency equal to the resonance frequency of the cav-
ity resonator, is coupled into the cavity, it starts to oscillate. This oscillating energy
is then radiated by the antenna element, i.e., the incident signal is reflected. If the
properties of all elements are the same, the reflection angle of the reflected waves is
equivalent to the incident angle of impinging waves. This is illustrated in Fig. 1 where
the incident signal s1 (f , θi,1 ) coming from direction θi,1 at frequency f is reflected to
the specular direction θr,1 as α1 s1 (f , θr,1 ), where complex coefficient α1 denotes the
effect of the patch antenna. Similarly, the other incoming signal s2 (f , θi,2 ) from direc-
tion θi,2 is reflected towards θr,2 as α2 s2 (f , θr,2 ), which is different from the reflection
direction of s1 (f , θi,1 ). If the desired reflection angle is different from θr,2 , say θr,d in
Fig. 1, then the phase responses of the individual elements should be changed so that
α2 s2 (f , θr,2 ) = sd (f , θr,d ). This could be achieved by keeping the phase response of one
element, say, the one on the left in Fig. 1, unchanged and tune the phase response
of the other elements so that the aggregated waves reflecting on all the elements
propagates towards the desired direction. The phase response of an antenna ele-
ment changes, when its resonance frequency is changed. Hence, it can be controlled
by connecting a varactor between the metallic patch and the ground plate [14]. By
varying the bias voltage of the varactor, the reflection angle of the reflected wave can
be steered electrically. More specifically, when the phase response of an element is
changed, the reflection angles of all impinging waves are affected, such as the reflec-
tions of both s1 (f , θi,1 ) and s2 (f , θi,2 ) as can be seen in Fig. 1. This implies that the
incident waves coming from different directions cannot, in general, be reflected into a
common direction.
The model described above does not take into consideration the coupling effect
between the neighboring elements, because the cavity resonator model cannot
be used to model the coupling, but it provides a basic description of the reflection
behavior in planar antenna arrays. To determine the phase shift required to steer the
reflected EM wave toward a desired direction, consider the planar array model shown
in Fig. 2. The black dots shown in the figure indicate the positions of the elements,
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 4 of 18

Fig. 2 Coordinate system for a reflect array

given that the element spacing d is equal over the x and y planes. When an EM wave
impinges a planar array from the direction defined with the angle between the z-axis
and the propagation direction of the wave ( ϕ ) and the propagation direction in x − y
plane ( φ ), the signal components on the x − y plane are

XRIS (x, y) = f1 f2T hBS→RIS (ϕ, φ)xBS (ϕ, φ) (1)

where xBS (ϕ, φ) is the signal sent by a transmitter, e.g., base station (BS), and received by
the planar array from (ϕ, φ) direction, hBS→RIS (ϕ, φ) is the channel between the base sta-
tion and array seen by the signal xBS, and vectors f1 and f2 are

f1 =[1 ejκd sin u ejκ2d sin u · · · ejκ(N −1)d sin u ]T (2)

f2 =[1 ejκd sin v ejκ2d sin u · · · ejκ(N −1)d sin v ]T (3)

where sin v = sin ϕ sin φ and sin u = sin ϕ cos φ, and N and M are the numbers of ele-
ments in x and y directions, respectively. When (0,0,0) in Fig. 2 is selected as a reference
point, the relative phase shift at point (n, m) on the planar array is

ψ(n, m) = κ(nd sin ϕ cos φ + md sin ϕ sin φ) (4)

where κ = 2π/. In order to reflect the signal towards the (ϕr , φr ) direction, the required
relative phases of the elements are

ψr (n, m) = κ(nd sin ϕr cos φr + md sin ϕr sin φr ) (5)

This means that the phase of the element at point (n, m) must be changed by

ψr (n, m) =nd(sin ϕr cos φr − sin ϕ cos φ)

(6)
+ md(sin ϕr sin φr − sin ϕ sin φ)

The equations above assume that the phase shifts can be controlled continuously.
A recent theoretical study has shown that as the number of elements increases, the
number of required distinct phase values is reduced [15]. Furthermore, as the number
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 5 of 18

of elements approaches infinity, only two phase values are needed (1-bit control). For
instance, with 300 elements the phase control with 2-bits (4 phase values) reduces the
performance by 10% compared with the case where phase is controlled continuously,
i.e., not with discrete steps. Additionally, when the number of elements in the array is
further reduced to three, three bits are needed translating to eight phase shifts values,
given that the reduction in performance from continuous phase control case does not
exceed 10% [15].

3.2 Antenna array based RIS in wireless systems

In the RIS aided systems and algorithm development, the elements in the RIS are typi-
cally modeled to operate independently of each other, i.e., with no mutual coupling
between the elements of the RIS. This means that those theoretical results and algo-
rithms are applicable to antenna array-based reflectors. In addition, in many cases, the
reflectors are simply modeled as linear arrays. For performance studies, the array geom-
etry has a negligible impact on the reflection behavior of the RIS. However, the array
geometry needs to be considered to apply the algorithms for two-dimensional arrays.
Alternating optimization is used in [16] to find the optimal phase values for the RIS
elements and the transmit covariance matrix to maximize the channel capacity of an
antenna array based RIS assisted MIMO link. In the numerical examples in [16], a RIS
with 20 elements increased the achievable rate performance by 38.8% in a MIMO-
OFDM link under frequency-selective channel condition. The use of antenna array
based RIS in orthogonal and non-orthogonal multiple access schemes is studied in [17]
wherein the capacity and rate regions for both schemes are characterized. Numerical
examples show the performance gains achieved by deploying a RIS, e.g, in a two user
non-orthogonal multi access scheme the gain in the common average rate is 12 dB. In
order to achieve such high gains, a real-time RIS control link is assumed. It is also shown
that for practical systems, the phase control resolution in the RIS can be as low as two
bits (four phase values for the RIS elements).
The antenna array based model is used in [15] for determining the effect of the phase
control resolution on the RIS-assisted links. In [18], a uniform rectangular antenna array
model is used in an optimal joint beamforming design for a BS and RIS in a single-cell
multi-user network. The RIS in [18] can be switched between two modes. It can be used
either for channel estimation or as a steerable beamforming reflector. In a single-user
case, the results showed that the signal-to-noise ratio (SNR) increases quadratically with
the number of reflecting elements of the RIS. In the multi-user case, it was demonstrated
that interference suppression with an RIS can be jointly designed with the access point
(AP) transmit beamforming for improving the performance of all users in the system.
Furthermore, the wireless network performance can be improved in terms of energy
consumption, coverage as well as achievable rate as compared to the conventional sys-
tems without using the RIS [18].
In [19], Huang et. al design an antenna array-based RIS system for transmit power allo-
cation as well as for phase shift control single-cell multi-user systems, in which a multi-
antenna base station serves single antenna receivers. The methods proposed in [19] are
capable of improving the energy efficiency by three-fold in comparison to the conven-
tional multi-antenna amplify-and-forward relaying. In [20], RIS is used to enhance the
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 6 of 18

communications from a single antenna base station to single-antenna users located

within the vicinity of the RIS. The flexibility of joint resource allocation is accomplished
by allowing the RIS reflection coefficients to be dynamically tuned over different time
slots to create time-varying channels to each user.
In [21], RIS is used as a reflector to enhance the communication link with the aid of
beamforming. Additionally, it employs spatial modulation for sending its private data
to the receiver. The data sent by RIS is used to control the reflecting elements using the
ON and OFF states, where the phase shifts of the elements in the ON state are used for
steering the reflected signal. The employed beam steering improves the average SNR at
the receiver by up to 9 dB, when average 50% of elements are in active state (RIS size
32 × 32 elements). Furthermore, a two-step approach is utilized for retrieving the
information transmitted from both the transmitter and RIS. The proposed techniques
provide a trade-off between the attainable beamforming gain of the reflected signal and
the achievable information rate of RIS data.
In [22], the steering of the reflected signal from an intelligent surface is used to
enhance the signals for transmission pairs, while canceling the interference from other
transmitters. This study also reports measured results as well as the theoretical analysis
of the potential gain of the proposed system. The intelligent surface used relies on an
(6 × 8)-element array of controllable patch elements, which are controlled with the aid
of varactors. In the measurements, the RIS is used for suppressing the interference. It
is shown that the maximum achieved interference suppression reaches around 30 dB,
however, in a very narrow bandwidth. For instance, the −20 dB bandwidth is less than 5
MHz and the −10 dB bandwidth is less than 10 MHz. The results show that the control-
lable surface can be used for interference mitigation.
The array design in [20] is based on a reflector array described in [12]. More specifi-
cally, the reflect array designed and implemented in [12] consists of 70 elements that
operate at 5.8 GHz. The array is capable of steering the incident beam angle between
− 50◦ to +50◦ from the array normal with 5 ◦ steps. The total signal loss in the system
is 7.4 dB, which is caused by the feed efficiency. Particularly, the loss due to element
absorption, phase error and element factor are reported to be 1.8 dB, 0.8 dB and 0.8 dB,
respectively.
A RIS prototype with 256 elements has been published in [23]. The phase shifts of the
RIS elements in [23] are controlled with two-bits and it achieves a 23 dBi antenna gain at
2.3 GHz. The physical size of the RIS prototype is 0.8 m × 0.8 m. The −1 dB bandwidth
is 350 MHz and the beam can be scanned from 0 ◦ to 60◦ from the broadside of the RIS.
Many of the articles considering the algorithms to control the RIS use the MIMO
type processing [18, 19, 21], where the channel is first estimated, and the beam steering
and the radio resource management is based on the estimated channel state informa-
tion. A different approach is proposed in [24], where no channel estimation is needed to
control the reflection. Therein the reflecting surface used 3200-elements antenna array
with an area of array 6 m2. The control of the array is implemented with a majority vot-
ing algorithm. In the training phase, the array is randomly configured to states where
some of the elements reflect the incident signal and others do not. The receiver reports
the received power at different stages of the training to the network. After K measure-
ments, the median of the received power is calculated. Specifically, if the received power
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 7 of 18

is above the median more often when element i has been active than when it has been
switched off, the element i is voted to be on. Otherwise it is voted to be off. The training
of the surface requires no changes to the signal processing in the transceivers. Further-
more, the received power level is reported with the received signal strength indicator
(RSSI), which is commonly available in essentially all wireless communication systems.
The RSSI information is reported to the control unit of the reflecting surface. In the test
scenarios, the receiver is kept at the same position close to the surface, while the posi-
tion of the transmitter is altered. The SNR gain in these experiments spans from 3.8 to
20 dB with a median of 9.5 dB.

4 Metasurfaces in RIS applications

4.1 Metasurface
In antenna array based RIS, the structure of the radio surface is fixed, which imposes
a limited control on the RIS. However, metasurfaces, which are electrically thin and
dense two-dimensional arrays of structural elements possessing the desired proper-
ties granted by their constitutive elements, allow a more flexible control on the oper-
ation of RIS. More specifically, the metasurface can be designed as a cluster of basic
elements (colored elements in the top part of Fig. 3), where the operation of a single
element is designed based on the requirements set on the operation of RIS. Alterna-
tively, it can be designed as a collection of sub-areas, as depicted in the bottom part
of Fig. 3. Hence, the granularity of the control spans from hundreds or thousands
controllable elements (depending on the metasurface size) to the case where a whole
RIS is a single area with one controllable state, e.g., surface reflects or absorbs the
incoming wave. The signal reflected by a meta-surface can also be described with
(1)–(6) assuming that the elements of the metasurface have been fixed, or in the case
of controllable metasurface, after the state of the controllable elements have been
changed. However, in a metasurface, the dimensions and spacing between the ele-
mentary elements (called unit cells, meta cells, meta atoms, etc.) are much smaller
than the wavelength of the EM wave to be controlled (typically between /10 and
/5 ) [8, 9]. This means that in the design phase, a metasurface cannot be modeled as
an antenna array, but instead the analysis and design of metasurfaces can be based
on the concept of polarizability.
By distributing unit cells with different polarizations, the reflection and refraction
can be realized to other directions according to the generalized Snell’s law defined
by (7). The polarizability of a unit cell can be controlled by the cell’s size and shape
as well as by its inter-spacing distance from neighboring unit cells. This approach is
used in the design of fixed metasurfaces. A more detailed description of the connec-
tion between the polarizabilities of unit cells and the desired directions of refrac-
tion and can be found in [25]. A controllable metasurface can be implemented with
identical unit cells distributed evenly on the surface, where their polarizability can
be controlled. This can be realized by connecting the meta cells (metal patches
in Fig. 5) to each other or to the ground with a varactor in the same manner as in
antenna array-based RIS.
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 8 of 18

Meta-cells in metasurface
...

...

A2 A5 A7
A1 A6
Sub-areas in metsurface

A4
A9
A3
A8
A10

A12
A11 A13 A14
Fig. 3 Metasurface as a cluster of basic elements or as a collection of sub-areas

4.2 Reflection and refraction in metasurfaces

The Snell’s law relates the incident angle with the angle of refraction, i.e., the change in
the EM propagation direction (see Fig. 4b), at the boundary of two dielectric materials
and it can be written as

υ2 sin θt = υ1 sin θi (7)

where, θi is the incident angle and θt is the angle of refraction, while υ1 and υ2 are the
refractive indices of the materials where the EM wave is arriving to the boundary
between the two materials (material 1) and the material where the EM is propagating
after the refraction (material 2), respectively. In the case of reflection, the reflection
angle (θr ) from the same boundary is the same as the incident angle, i.e., θr = θi . The
above states that the angles of the reflection and refraction depend only on the material
properties at both sides of the boundary between the materials and the incident angle
of the EM wave. The boundary between the two materials does not allow to steer the
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 9 of 18

Fig. 4 a Metasurface in x-y-z coordinate systems, b reflection, c refraction

EM waves to any other directions. However, when an appropriately designed thin sheet
is placed between the two materials, it is possible to control the behavior of the EM
waves above and below the sheet more freely [25–29]. The engineered thin sheets with
designed reflective, refractive and absorption properties are called metasurfaces.
A metasurface is shown in Fig. 4a in the x − y − z coordinate systems. It can be
designed to reflect the impinging EM wave. The reflecting wave then propagates in
the same medium where it propagated before the reflection, but the propagation
direction has changed. The EM wave arriving from the direction defined by the x
and z coordinates is shown in Fig. 4b. The relation between the reflection and inci-
dence angles in the case of metasurfaces can be written as

∂φr (x, y, 0)
υ1 (sin θ1 − cos θr ) = (8)
∂x

where φr (x, y, 0) is the phase of the reflection coefficient at point (x, y, 0). If the propaga-
tion direction of EM is defined in the y − x plane, the partial derivative in (8) is calcu-
lated with the respect to y. Other directions can be calculated as the superposition of
these two.
When metasurfaces are used to implement an RIS, the design goal in addition to
control the direction of reflection, is to maximize the reflected power to the desired
direction.
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 10 of 18

Dielectric substrate on
Metasurface boundary
ground plane
...
Metal patches

Metal patches

... ... ...

Vias connecng
patches to ground Dielectric substrate Ground plane
plate
...
...

Mushroom structure top view Mushroom structure side view

Fig. 5 Top and side view of mushroom structure

4.3 Controllable metasurfaces
Often used structure to implement a configurable of tunable metasurface is the mush-
room structure portrayed in Fig. 5. It consists of rectangular small metal patches placed
close to each other on a thin dielectric substrate. At the other side of the substrate is
the ground plane. The patches of Fig. 5 are connected to the ground plane by vias going
through the dielectric substrate.
Metasurfaces and similar electrically thin structures have been used for a long time
to control the propagation of EM waves. In communication applications, these designs
have been used especially in antenna designs. Most of the metasurfaces are fixed. They
are designed to steer the EM wave coming from a known direction to the predetermined
directions. A comprehensive review of metasurfaces can be found in [3]. It covers the
applications and technologies of metasurfaces from microwave to optical frequencies.
In metasurface based RIS applications, the properties of the metasurface must be con-
trollable. The incident field must be directed to directions where the receiving nodes
are in order to improve the performance of a communication or positioning system. A
review of the state-of-the-art on controllable metasurfaces is presented in this section.
The implementations described below represent the typical published designs. The fre-
quency range of interest in this survey is 2–10 GHz.
The programmability or tunability of metasurface based RISs requires the integration
of active elements with the unit cells of the metasurfaces. Controlling the microwave
frequencies can be realized with the aid of varactors. A common implementation tech-
nique of varactors uses p-i-n diodes as in [30]. At higher frequencies (terahertz to optical
frequencies) tunability can be realized with, e.g., transparent conductive oxides, ferro-
electric materials, liquid crystal, graphene and phase change materials [31]. Graphene
has also been used to implement a tunable absorber working at 3.41–4.55 GHz in [32].
Since the considered application is RIS and the frequency range in interest in this work
is below 10 GHz, these are not considered here.
A metasurface for electrically controlled beam steering was introduced in [30]. The
structure of the metasurface consists of small square metal patches on a dielectric sub-
strate. Patches are connected to the ground plate using a via. This kind of structure
is commonly called a mushroom structure and it is illustrated in Fig. 5. Each patch is
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 11 of 18

connected to its adjacent patches by a varactor. By changing their control voltages, the
resonant frequency of the unit cells (patches with vias) can be varied. By adjusting the
resonant frequency of the unit cells, the phase of the reflection coefficient at different
points on the metasurface can be controlled. This allows to create the required phase
gradient to steer the reflected EM wave. The reflection can be controlled in two dimen-
sions (both azimuth and elevation planes). The designed metasurface in [30], is operat-
ing at 3.5 to 5.0 GHz frequency and is able to steer the incident beam over ±40◦ from the
surface normal. The reported main beam width and gain are 15◦ and 14 dBi, respectively.
Maximum sidelobes are −10 dB below the maximum gain of the radiation pattern and
the bandwidth is about 8% at 4.5 GHz (≈ 360 MHz).
The control of the reflection in [30] is based on analog signal processing, the reflection
direction is changed by changing the bias voltages of the varactors. The reflection from
a metasurface can be controlled also digitally. A digitally controllable structure based
on the similar mushroom structure as in [30] has been considered in [33]. The varactor
is connected between the metal patch and the ground plate in [33] and not between the
patches as in [30]. The varactor bias line design allows the beam steering only in one
dimension. The control is implemented with a field programmable array (FPGA), var-
actors can be programmed to be in two states, “ON” or “OFF”. The implementation in
[33] is designed for 10 GHz center frequency. The maximum diode switching speed is 5
MHz. The control voltage is used to change the operating frequency around the center
frequency as well as the reflection direction, i.e., beams pointing into different directions
are at different frequencies. The range of the beam steering control is between 45◦ and
50◦.
Metasurface controlled by an FPGA is considered also in [34]. By changing the state
of the diode between “0” and “1”, the reflection coefficient of the unit cell is changed
between 0 ◦ and 180◦ . By properly selecting the control bit pattern on the surface, the
reflection can be directed to desired directions. The prototype consists of 30 × 30 unit
cells.
An FPGA controlled reflective metasurface is described also in [35]. The control of the
surface is implemented by switching the diode “ON” or “OFF” as in [33, 34]. The imple-
mented metasurface consists of an array of 20 × 20 unit cells. The operating frequency
is 3.5 GHz. The same unit cell structure is used also in [36]. Instead of controlling the
bias voltage digitally, the control is realized with light. By changing the intensity of the
light illuminating the surface, the bias voltages of the varactors can be switched “ON”
or “OFF” with photodiodes. The control of the metasurface by switching the varactor
diodes “ON” or “OFF” is used also in [37]. The metasurface can be used for generating
multi-ple beams or twin-beam scanning in the low frequency band around 6 GHz while
dynamic beam-scanning in high frequency band at 9.8 GHz.
In [38], the varactors are used to tune the operating frequency of the metasurface
between 4.1. and 6.6 GHz. Varactors are used only for the frequency tuning, not to
direct the reflection.
A metasurface based on a more complex unit cell is described in [39]. The structure
consists of a parallel LC resonant circuit and small metallic loops. Varactors integrated
into the unit cell allow the dynamic tuning of the surface impedance of the surface. The
design in [39] is used to control the refraction, i.e., the EM wave traveling through the
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 12 of 18

surface, and not the reflection. The implemented prototype operates at 3 GHz frequency.
The transmittivity of the surface can be tuned from 0 to 1 when the phase of the EM
wave is kept unaltered, or the phase of the transmission can be tuned by 360◦ while
the transmittivity is kept constant. In [39], the control of the refraction angle itself is
not demonstrated. But if different areas (unit cells) of the surface are programmed to
induce different phase shifts of the EM wave, the refraction direction can, in principle,
be controlled.
In [40], unit cells are realized with two twisted square split ring resonators to imple-
ment a tunable chiral metasurface. The surface is designed to act as a lens capable to
steer the EM wave propagating through it and its operating frequency range is 8–10
GHz. About 80% of the incident power is transmitted through the surface. The tuning of
the surface is done with varactors.
The application discussed in [41] is an antenna array design. The results illustrate the
ability of a metasurface to steer the EM wave propagating through it (microwave lens).
The scanning range of an antenna array is increased from [− 36◦ , 38◦ ] to [− 56◦ , 60◦ ]
with the usage of the metasurface design but the gain in main lobe direction is reduced
due to losses in the metasurface. The transmittivity of the surface is above 0.8 from 9
GHz to 10 GHz. The unit cell in [41] is based on a sandwich structure, where the top and
bottom layers are co-centric combinations of square ring and circular patches.
The design in [42] is used to implement an absorber around 3.6 GHz frequency. It is
not directly applicable to be used in RIS applications, but is included here because of its
control circuitry design. In most of the tunable/controllable metasurface designs below
10 GHz, the tuning is implemented with lumped components (varactors). The tuning is
accomplished by changing the bias voltage of the varactors. The unit cell designed with
metal-oxide-semiconductor field-effect transistors in [42] enables the control of not only
the capacitance but both the real and imaginary part (resistance and reactance) of the
loading impedance.
The tunability of the metasurface in [43] is proposed to be implemented using mixed
signal integrated circuit embedded in the unit cell. The IC connects two metal patches
of the unit cell to enable the control of unit cell resistance and reactance. This offers the
local control over the surface resulting in the control of the reflection properties. The
operation of the concept is verified by a design operating at 5 GHz frequency. The per-
formance of the design is studied with simulations. The IC proposed for the tuning it not
designed but its operation is modeled as a continuously tunable, lumped complex tun-
able impedance.

4.4 Metasurface based RIS in wireless systems

The scenario used in [9] and [44] is an indoor wireless communication environment
in which the interior walls of the building are fully covered with metasurfaces. Users
inside the building require different services simultaneously, some of the users want to
have optimal connectivity, others may require protection from eavesdropping and some
users are interested in wireless power transfer. In addition, interference and unauthor-
ized connection from outside the building are blocked by the system. The main parts
of [9] and [44] describe a general framework for software interface architecture for the
control of the envisioned system. The metasurfaces are assumed to be almost ideal, i.e., it
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 13 of 18

is assumed that metasurfaces do not exhibit any major limitations regarding the electro-
magnetic functionalities that they can support.
In [45, 46], interior wall of a building is assumed to be covered by a surface whose EM
properties can be switched between two states. In one state, the EM wave can propagate
through the wall, in the other state the wall attenuates the EM wave propagating from
one room to the next through the wall. This can be used to decrease the interference
between local networks operating in adjacent rooms or to deny access to the local net-
work from outside the room. The system model in [45, 46] is much simpler than the one
in [44], but it is also more realistic in the sense that it does not require ideal configurable
surface. In the simulation case, it is assumed that when the signal is allowed to propagate
through the wall it is attenuated by 2 dB and when the wall is configured to attenuate the
signal, the attenuation is 25 dB.
According a Press Release [47], NTT DOCOMO has demonstrated the control of EM
propagation through a window using a metasurface operating at 28 GHz. Although this
test was conducted at a frequency range above our focus, it has been included, because
it is the first test of metasurface based RIS application published by one of the major
operators.
A different kind of application of a metasurface in wireless communications is demon-
strated in [48]. The reflecting metasurface is used for data modulation. In the proposed
transmitter architecture, a feed antenna transmits a carrier wave to the metasurface. The
reflection coefficient of the metasurface is controlled with the transmitted baseband data
signal resulting in a modulated signal at the radio frequency. The center frequency in the
test has been 4 GHz and QPSK is used for the data modulation. The achieved data rate
is 2.048 Mbps with a single carrier modulation. The benefit of using the metasurface as
a transmitter is that the up-conversion based on mixers used in conventional transmit-
ters is omitted. The disadvantage is the power loss. In [48], the conventional transmitter
requires 5 dB less power than the one based on the metasurface modulator.

5 Results and discussion

The potential use cases and the benefits of the RIS at frequencies below 10 GHz are still
not fully known. The results from prototypes reviewed in this survey indicate that a RIS
can be used to control the EM propagation also at this frequency range. An EM wave
is attenuated, when it reflects from floors, walls, and ceilings. The reflection attenua-
tion of typical construction materials is in the order of 7–10 dB. This implies that there
are typically few dominant directions where the signal arrives from at any point. This is
also visible in typical measurement-based channel models, e.g., in 5G channel models
[49]. Although the dominant channel direction below 10 GHz is not as clear as at higher
frequencies (tens or hundreds GHz), a RIS has potential to extend the coverage and
increase the channel capacity. However, channel measurements for RIS assisted systems
operating at the sub-gigahertz frequency range are needed to verify the potential gains.
Intelligent radio surfaces below 10 GHz frequency can be implemented with antenna
arrays or metasurfaces. The electromagnetic theories and models for the analysis and
design for both options are well known and commercial simulation tools for modeling
are available. The exact theoretical model for a large complex structure can, in principle,
be formed, but in practical work the design is done with design and simulation tools.
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 14 of 18

This is evident also in publications reporting the designed, simulated, and implemented
metasurfaces. Design with simulation tools can be efficient, but the development of
control algorithms for metasurface based RIS necessitates models for the metasurface
operation. The properties of the metasurfaces, e.g. the surface impedance, depends on
the polarization and the angle-of-arrival (both azimuth and elevation) of the impinging
electromagnetic wave. Further, the unit-cells of the metasurface are mutually coupled to
their neighbouring cells. Accurate models for these effects are not yet available for con-
trol algorithm development.
Unit cells used for the reconfigurable metasurfaces below 10 GHz frequency are
mostly based on simple geometrical shapes such as rectangular patches, H-shape, and
rectangular rings. The sizes of the implemented surfaces are in the order of 30 × 30 unit
cells (about 30 cm × 30 cm). The materials used are the same as those used in printed
circuit boards at microwave frequencies. More exotic materials and complex unit cell
structures are used at higher frequencies, terahertz and optical frequencies. The reflec-
tivity and transmittivity of the prototypes are below 0.9 (10% or more loss). The tuning is
implemented typically with varactors. In published antenna array model based algorithm
designs, the elements of the arrays are assumed to operate independently (no mutual
coupling between elements) and the phase control is assumed to be implemented in the
same manner as in the antenna arrays. While in the conventional antenna array cases the
phase of the transmitted (or received) signal can be controlled with phase shifting cir-
cuits, in reflect arrays the phase shift must be done by changing the physical properties
of the elements, namely, the resonance frequency. If the number of elements in a reflect
array is large, the control of RIS can be realized by switching the phase response of the
elements between distinct values, i.e., the phase control does not need to be continuous.
The unit-cells are close to each other in metasurfaces, which means that they cannot be
assumed to operate independently. In addition, the number of controllable elements can
be much higher than in the reflect array based designs. In the published demonstrations
of controllable metasurfaces, the operation of unit-cells has had two states. In some
cases, the allowable unit-cell states are ON or OFF. In other cases, the unit-cell reflection
angle has had two distinct values. The details of the algorithms have not been published.
In addition to the theoretical works, some tests and demonstrations of metasurface
based intelligent radio surfaces in wireless systems have been published. The most com-
plete demonstrations are probably the ones performed by NTT DOCOMO in Japan.
Unfortunately, only press releases with restricted information content on these were
found. Other published tests show that it is possible to improve the performance of a
wireless system by using the controllable reflecting surfaces, but the available informa-
tion is insufficient to assess the gain in practical scenarios with complete wireless sys-
tems based on the published results so far.
Commercial products based on the metasurfaces are mainly transmit/receive anten-
nas. Some of the companies have commercial-off-the-shelf antenna products, but these
companies are mainly design houses offering design services for metasurface based
products. The feasibility of using fixed reflecting surfaces to improve the coverage and
service quality of wireless communication has been already demonstrated, especially by
the experiments by DOCOMO in Japan. Theoretical studies have already showed that
surfaces with controllable properties can further enhance the performance of wireless
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 15 of 18

Table 1 Comparison of metasurface and reflectarray based RIS

Relectarray Metasurface

Operation well-known, relatively easy to model Difficult to model controllable surfaces

Simple to manufacture Potentially more advanced functions
Simple control Mutual coupling of meta-cells complicates the control
Readily available technology Reconfigurable metasurfaces still in research phase

systems. New, improved services, such as accurate positioning and improved infor-
mation security, have been envisioned and some initial published studies support this
assumption. However, to fully understand the benefits the intelligent radio surfaces can
bring, more extensive investigations on the possible use cases needs to be done.
Controllable surfaces can be implemented with conventional reflect arrays, but meta-
surface based structures can offer more extensive options for the control. The opera-
tion of fixed metasurfaces is known and some controllable or tunable metasurfaces have
been designed and implemented. Published controllable metasurfaces are relatively
small and their electromagnetic properties (reflection efficiency, beam width, steering
range and accuracy, etc.) have not yet been thoroughly studied. The research on the con-
trol of the intelligent surfaces, i.e., system architecture and algorithms, is in early stages.
Algorithms published so far apply techniques used in the MIMO systems. While this
approach can provide good results with relatively small antenna array based surfaces,
the efficient control of large reflect arrays may require new approaches. The large num-
bers of controllable elements and their strong mutual coupling in metasurfaces require
deviation from the MIMO processing-based approach. The efficient control of metasur-
face based intelligent surfaces is largely an open problem. The comparison of a reflectar-
ray and a metasurface as a RIS implementation technology is summarized in Table 1.

Abbreviations
AP: Access point; dB: Decibel; dBi: Decibel isotropic; EM: Electromagnetic; FPGA: Field programmable gate array; FR1: 5G
NR frequency range 1; GHz: Gigahertz; IC: Integrated circuit; IRS: Intelligent reflecting surface; LC: inductor-capacitor; LIS:
Large intelligent surface; MHz: Megahertz; NR: New radio; QPSK: Quadrature phase shift keying; RIS: Reconfigurable intel-
ligent surface; MIMO: Multiple-input multiple-output; RF: Radio frequency; RX: Receiver; THz: Terahertz; TX: Transmitter.

Acknowledgements
Not applicable.

Author contributions
All authors have contributed to the literature search and to the writing of the survey. All authors read and approved the
final manuscript.

Authors’ information
Visa Tapio received his M.Sc. (EE) and Lic.Sc. (EE) degrees from University of Oulu, Oulu, Finland in 1995 and 2004,
respectively. V. Tapio has been with University of Oulu since 1994. His research interests include the wireless transceiver
development, especially the relation between the physical characterization of transceivers at radio frequencies and
digital signal processing algorithm.

Ibrahin Hemadeh is an R&D staff engineer at Interdigital. He received his MSc degree (Hons.) in wireless communications
and his PhD degree in electronics and electrical engineering from The University of Southampton, UK, in 2012 and 2017,
respectively. In 2017, he joined the Southampton Next Generation Wireless Group, The University of Southampton, as
a Post-Doctoral Researcher. In 2018, he joined the 5G Innovation Centre (5GIC) at the University of Surrey. His research
interests include reconfigurable intelligent surfaces, millimeter-wave communications, multi-functional multiple input
multiple output (MIMO), multi-dimensional (time-space and frequency) transceiver designs, channel coding, as well as
multi-user MIMO.

Alain Mourad is a Director Engineering R&D at InterDigital Labs in London (UK) leading research on 5G and beyond. Prior
to InterDigital, Alain was a Principal Engineer at Samsung Electronics R&D and previously a Senior Engineer at Mitsubishi
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 16 of 18

Electric R&D Centre Europe where he was active in the specification of wireless standards (3GPP, IEEE802, DVB, ATSC).
Alain is a prolific inventor with over 50 issued patents. He received Samsung Electronics Inventor Awards in 2012 and
2013, InterDigital Awards in 2016, 2018 and 2020, and 2018 Global Telecoms Awards “Highly Commended” for “Advanc-
ing the road to 5G”.

Arman Shojaeifard has +7 years of work experience in research and development of air-interface technologies and
architectures for 4G LTE/LTE-A, 5G NR, and beyond. He received the BEng degree in Information Systems Engineering
from Imperial College London in 2008. He then graduated from King’s College London in 2009, with an MSc degree in
Signal Processing (with Distinction), and in 2013, with a PhD degree in Wireless Communications. He is currently a Staff
Engineer at InterDigital where he leads the CELTIC-NEXT European collaborative project on AIMM (AI-enabled Massive
MIMO). He was previously a Wireless Research Manager and 3GPP RAN1 WG Delegate at British Telecoms. Prior to that,
he held post-doctoral research positions at University College London and University of Manchester. He is currently an
Editor for IEEE Wireless Communications Letters and IET Electronics Letters.

Markku Juntti received his Dr.Sc. (EE) degree from University of Oulu, Oulu, Finland in 1997. Dr. Juntti was with University
of Oulu in 1992–98. In academic year 1994–95, he was a Visiting Scholar at Rice University, Houston, Texas. In 1999–2000,
he was a Senior Specialist with Nokia Networks, Oulu, Finland. Dr. Juntti has been a professor of communications engi-
neering since 2000 at University of Oulu, Centre for Wireless Communications (CWC), where he also serves as Head of
CWC – Radio Technologies Research Unit. His research interests include signal processing for wireless networks as well as
communication and information theory. Dr. Juntti is also an Adjunct Professor at Rice University.
Funding
The research by V. Tapio and M. Juntti has been funded in part by InterDigital Corporation and the Academy of Finland
6Genesis Flagship (grant 318927).

Availability of data and materials

Data sharing is not applicable to this article as no datasets were generated during the current study.

Declarations
Competing interests
The authors declare that they have no competing interests.

Author details
1
Centre for Wireless Communications, University of Oulu, Oulu, Finland. 2 InterDigital Europe, Ltd, London, UK.

Received: 8 April 2021 Accepted: 1 September 2021

References
1. M. Di Renzo, A. Zappone, M. Debbah, M.-S. Alouini, C. Yuen, J. de Rosny, S. Tretyakov, Smart radio environments
empowered by reconfigurable intelligent surfaces: How it works, state of research, and the road ahead. IEEE J. Sel.
Areas Commun. 38(11), 2450–2525 (2020). https://​doi.​org/​10.​1109/​JSAC.​2020.​30072​11
2. E. Basar, M. Di Renzo, J. De Rosny, M. Debbah, M. Alouini, R. Zhang, Wireless communications through reconfigurable
intelligent surfaces. IEEE Access 7, 116753–116773 (2019). https://​doi.​org/​10.​1109/​ACCESS.​2019.​29351​92
3. Q. Wu, R. Zhang, Towards smart and reconfigurable environment: intelligent reflecting surface aided wireless net-
work. IEEE Commun. Mag. 58(1), 106–112 (2020). https://​doi.​org/​10.​1109/​MCOM.​001.​19001​07
4. N. Rajatheva, White paper on broadband connectivity in 6G. Technical report, Finnish 6G Flagship Program (2020).
http://​julti​ka.​oulu.​fi/​files/​isbn9​78952​62267​98.​pdf
5. H. Wymeersch, J. He, B. Denis, A. Clemente, M. Juntti, Radio localization and mapping with reconfigurable intelligent
surfaces: challenges, opportunities, and research directions. IEEE Veh. Technol. Mag. 15(4), 52–61 (2020). https://​doi.​
org/​10.​1109/​MVT.​2020.​30236​82
6. C. de Lima, et al. 6G white paper on localization and sensing. Technical report, Finnish 6G Flagship Program (2020).
http://​julti​ka.​oulu.​fi/​files/​isbn9​78952​62267​43.​pdf
7. M. Di Renzo, et. al. Smart radio environments empowered by reconfigurable AI meta-surfaces: an idea whose time
has come. J. Wirel. Com Network 2019, 129 (2019) (2019). https://​doi.​org/​10.​1186/​s13638-​019-​1438-9
8. S.B. Glybovski, S.A. Tretyakov, P.A. Belov, Y.S. Kivshar, C.R. Simovski, Metasurfaces: from microwaves to visible. Phys.
Rep. 634, 1–72 (2016). https://​doi.​org/​10.​1016/j.​physr​ep.​2016.​04.​004
9. C. Liaskos, S. Nie, A. Tsioliaridou, A. Pitsillides, S. Ioannidis, I. Akyildiz, A novel communication paradigm for high
capacity and security via programmable indoor wireless environments in next generation wireless systems. Ad Hoc
Netw. 87, 1–16 (2019). https://​doi.​org/​10.​1016/j.​adhoc.​2018.​11.​001
10. T. Nakanishi, T. Otani, Y. Tamayama, M. Kitano, Storage of electromagnetic waves in a metamaterial that mimics elec-
tromagnetically induced transparency. Phys. Rev. B 87, 116110 (2013). https://​doi.​org/​10.​1103/​PhysR​evB.​87.​161110
11. A. Silva, F. Monticone, G. Castaldi, V. Galdi, A. Alù, N. Engheta, Performing mathematical operations with metamateri-
als. Science 343(6167), 160–163 (2014). https://​doi.​org/​10.​1126/​scien​ce.​12428​18
12. S.V. Hum, M. Okoniewski, R.J. Davies, Modeling and design of electronically tunable reflectarrays. IEEE Trans. Anten-
nas Propag. 55(8), 2200–2210 (2007). https://​doi.​org/​10.​1109/​TAP.​2007.​902002
13. C.A. Balanis, Antenna Theory: Analysis and Design, 2nd edn. (Wiley, 1997)
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 17 of 18

14. R. Garg, I.I.P. Bhartia, Microstrip Antenna Design Handbook (Artech House, Inc., 2001)
15. H. Zhang, B. Di, L. Song, Z. Han, Reconfigurable intelligent surfaces assisted communications with limited phase
shifts: how many phase shifts are enough? IEEE Trans. Veh. Technol. 69(4), 4498–4502 (2020). https://​doi.​org/​10.​
1109/​T VT.​2020.​29730​73
16. S. Zhang, R. Zhang, Capacity characterization for intelligent reflecting surface aided MIMO communication. IEEE J.
Sel. Areas Commun. 38(8), 1823–1838 (2020). https://​doi.​org/​10.​1109/​JSAC.​2020.​30008​14
17. X. Mu, Y. Liu, L. Guo, J. Lin, N. Al-Dhahir, Capacity and optimal resource allocation for IRS-assisted multi-user com-
munication systems. IEEE Trans. Commun. 69(6), 3771–3786 (2021). https://​doi.​org/​10.​1109/​TCOMM.​2021.​30626​51
18. Q. Wu, R. Zhang, Intelligent reflecting surface enhanced wireless network via joint active and passive beamforming.
IEEE Trans. Wireless Commun. 18(11), 5394–5409 (2019). https://​doi.​org/​10.​1109/​T WC.​2019.​29360​25
19. C. Huang, A. Zappone, G.C. Alexandropoulos, M. Debbah, C. Yuen, Reconfigurable intelligent surfaces for energy
efficiency in wireless communication. IEEE Trans. Wireless Commun. 18(8), 4157–4170 (2019). https://​doi.​org/​10.​
1109/​T WC.​2019.​29226​09
20. Y. Yang, S. Zhang, R. Zhang, IRS-enhanced OFDMA: Joint resource allocation and passive beamforming optimization.
IEEE Wireless Commun. Lett. 9(6), 760–764 (2020). https://​doi.​org/​10.​1109/​LWC.​2020.​29683​03
21. W. Yan, X. Yuan, X. Kuai, Passive beamforming and information transfer via large intelligent surface. IEEE Wireless
Commun. Lett. 9(4), 533–537 (2020). https://​doi.​org/​10.​1109/​LWC.​2019.​29616​70
22. X. Tan, Z. Sun, J.M. Jornet, D. Pados, Increasing indoor spectrum sharing capacity using smart reflect-array. In: 2016
IEEE International Conference on Communications (ICC), pp. 1–6 (2016). https://​doi.​org/​10.​1109/​ICC.​2016.​75109​62
23. L. Dai, B. Wang, M. Wang, X. Yang, J. Tan, S. Bi, S. Xu, F. Yang, Z. Chen, M.D. Renzo, C.-B. Chae, L. Hanzo, Reconfigurable
intelligent surface-based wireless communications: antenna design, prototyping, and experimental results. IEEE
Access 8, 45913–45923 (2020). https://​doi.​org/​10.​1109/​ACCESS.​2020.​29777​72
24. V. Arun, H. Balakrishhnan, Rfocus: beamforming using thousands of passive antennas. In: USENIX Symposium on
Networked Systems Design and Implementation (NSDI ’20) (2020)
25. V.S. Asadchy, M. Albooyeh, S.N. Tcvetkova, A. Díaz-Rubio, Y. Ra’di, S.A. Tretyakov, Perfect control of reflection and
refraction using spatially dispersive metasurfaces. Phys. Rev. B 94, 075142 (2016). https://​doi.​org/​10.​1103/​PhysR​evB.​
94.​075142
26. Y. Vahabzadeh, N. Chamanara, K. Achouri, C. Caloz, Computational analysis of metasurfaces. IEEE J. Multiscale Mul-
tiphysics Comput. Tech. 3, 37–49 (2018). https://​doi.​org/​10.​1109/​JMMCT.​2018.​28298​71
27. E.F. Kuester, M.A. Mohamed, M. Piket-May, C.L. Holloway, Averaged transition conditions for electromagnetic fields at
a metafilm. IEEE Trans. Antennas Propag. 51(10), 2641–2651 (2003). https://​doi.​org/​10.​1109/​TAP.​2003.​817560
28. S. Tretyakov, Analtical Modeling in Applied Electromagnetics (Artech House, Inc., 2003)
29. K. Achouri et al., Synthesis of electromagnetic metasurfaces: principles and illustrations. EPJ Appl. Metamat. 2, 12
(2015). https://​doi.​org/​10.​1051/​epjam/​20150​16
30. D.F. Sievenpiper, J.H. Schaffner, H.J. Song, R.Y. Loo, G. Tangonan, Two-dimensional beam steering using an electrically
tunable impedance surface. IEEE Trans. Antennas Propag. 51(10), 2713–2722 (2003). https://​doi.​org/​10.​1109/​TAP.​
2003.​817558
31. L. Bao, T.J. Cui, Tunable, reconfigurable and programmable metamaterials. Microw. Opt. Technol. Lett. 62(1), 9–32
(2020)
32. J. Zhang et al., Two-dimensional beam steering using an electrically tunable impedance surface. ACS Photonics 7(1),
265–271 (2020)
33. J. Zhang et al., Space–time-coding digital metasurfaces. Nat. Commun. 9, 1–11 (2018)
34. T.J. Cui et al., Coding metamaterials, digital metamaterials and programmable metamaterials. Light Sci. Appl. 3, 1–9
(2014)
35. S. Tian, H. Liu, L. Li, Design of 1-bit digital reconfigurable reflective metasurface for beam-scanning. Appl. Sci. 7(9),
1–8 (2017)
36. X.G. Zhang et al., Light-controllable digital coding metasurfaces. Appl. Sci. 5(11), (2018)
37. N. Zhang, K. Chen, Y. Zheng, Q. Hu, K. Qu, J. Zhao, J. Wang, Y. Feng, Programmable coding metasurface for dual-band
independent real-time beam control. IEEE Journal on Emerging and Selected Topics in Circuits and Systems 10(1),
20–28 (2020). https://​doi.​org/​10.​1109/​JETCAS.​2020.​29733​10
38. H.-X.Xu et al., Tunable microwave metasurfaces for high-performance operations: dispersion compensation and
dynamical switch. Sci. Rep. 6, 38255 (2016). https://​doi.​org/​10.​1038/​srep3​8255
39. B. Zhu et al., Tunable microwave metasurfaces for high-performance operations: dispersion compensation and
dynamical switch. Sci. Rep. 4, 4971 (2014). https://​doi.​org/​10.​1038/​srep0​4971
40. K. Chen, G. Ding, J. Zhao, Y. Feng, Dynamic control of microwave with tunable metamaterial and metasurface. In:
2018 IEEE Asia-Pacific Conference on Antennas and Propagation (APCAP), pp. 285–286 (2018). https://​doi.​org/​10.​
1109/​APCAP.​2018.​85380​97
41. Y. Lv, X. Ding, B. Wang, D.E. Anagnostou, Scanning range expansion of planar phased arrays using metasurfaces. IEEE
Trans. Antennas Propag. 68(3), 1402–1410 (2020). https://​doi.​org/​10.​1109/​TAP.​2020.​29672​76
42. K.M. Kossifos, M.A. Antoniades, J. Georgiou, Integrated-circuit enabled adaptive metasurface absorber with inde-
pendent tuning of orthogonal polarization planes. IEEE Access 8, 50227–50235 (2020). https://​doi.​org/​10.​1109/​
ACCESS.​2020.​29778​52
43. F. Liu, O. Tsilipakos, A. Pitilakis, A.C. Tasolamprou, M.S. Mirmoosa, N.V. Kantartzis, D.-H. Kwon, J. Georgiou, K. Kossifos,
M.A. Antoniades, M. Kafesaki, C.M. Soukoulis, S.A. Tretyakov, Intelligent metasurfaces with continuously tunable local
surface impedance for multiple reconfigurable functions. Phys. Rev. Appl. 11, 044024 (2019). https://​doi.​org/​10.​
1103/​PhysR​evApp​lied.​11.​044024
44. C. Liaskos, S. Nie, A. Tsioliaridou, A. Pitsillides, S. Ioannidis, I. Akyildiz, A new wireless communication paradigm
through software-controlled metasurfaces. IEEE Commun. Mag. 56(9), 162–169 (2018). https://​doi.​org/​10.​1109/​
MCOM.​2018.​17006​59
45. L. Subrt, P. Pechac, Controlling propagation environments using intelligent walls. In: 2012 6th European Conference
on Antennas and Propagation (EUCAP), pp. 1–5 (2012). https://​doi.​org/​10.​1109/​EuCAP.​2012.​62065​17
Tapio et al. J Wireless Com Network (2021) 2021:175 Page 18 of 18

46. L. Subrt, P. Pechac, Intelligent walls as autonomous parts of smart indoor environments. IET Commun. 6(8),
1004–1010 (2012). https://​doi.​org/​10.​1049/​iet-​com.​2010.​0544
47. DOCOMO, N.: Docomo conducts world’s first successful trial of transparent dynamic metasurface. Technical report,
NTT DOCOMO, INC. www.​nttdo​como.​co.​jp/​engli​sh/​info/​media_​center/​pr/​2020/​ 0117_​00.​html
48. W. Tang, X. Li, J.Y. Dai, S. Jin, Y. Zeng, Q. Cheng, T.J. Cui, Wireless communications with programmable metasurface:
transceiver design and experimental results. China Commun. 16(5), 46–61 (2019)
49. 5G: Study on channel model for frequencies from 0.5 to 100 GHz, ETSI TR 138 901 v14.3.0. Technical report, ETSI
(2018)