Electromagnetic Science

vol. 1, no. 2, 0020111, June 2023

[Link]

Review

Reconfigurable Intelligent Surfaces:

Design, Implementation, and Practical Demonstration

Yijun Feng , Qi Hu, Kai Qu, Weixu Yang , Yilin Zheng, and Ke Chen
School of Electronic Science and Engineering, Nanjing University, Nanjing 210023, China

Corresponding authors: Yijun Feng, Email: yjfeng@[Link]; Ke Chen, Email: [Link]@[Link].

Received November 16, 2022; Accepted April 6, 2023; Published Online June 16, 2023.
Copyright © 2023 The Author(s). This is a gold open access article under a Creative Commons Attribution License (CC BY 4.0).

Abstract — Metasurfaces, ultrathin two-dimensional version of metamaterials, have attracted tremendous attention due to their exotic capabilities to
freely manipulate electromagnetic waves. By incorporating various tunable materials or elements into metasurface designs, reconfigurable metasurfaces
and related metadevices with functionalities controlled by external stimuli can be realized, opening a new avenue to achieving dynamic manipulation of
electromagnetic waves. Recently, based on the tunable metasurface concept, reconfigurable intelligent surfaces (RISs) have received significant atten-
tion and have been regarded as a promising emerging technology for future wireless communication due to their potential to enhance the capacity and
coverage of wireless networks by smartly reconfiguring the wireless propagation environment. Here, in this article, we first focus on technical issues of
RIS system implementation by reviewing the existing research contributions, paying special attention to designs in the microwave regime. Then, we
showcase our recent attempts to practically demonstrate RIS systems in real-world applications, including deploying reflective RIS systems in indoor
scenarios to enhance the wireless network coverage and utilizing intelligent omni-metasurfaces to improve both indoor and through-wall wireless com-
munication quality. Finally, we give our own perspectives on possible future directions and existing challenges for RISs toward a truly commercial intel-
ligent technology platform.
Keywords — Reconfigurable intelligent surface, Metasurface, Wireless communication.
Citation — Yijun Feng, Qi Hu, Kai Qu, et al., “Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration,” Elec-
tromagnetic Science, vol. 1, no. 2, article no. 0020111, 2023. doi: 10.23919/emsci.2022.0011.

I. Introduction beam scanning for target detection. Therefore, tunable, re-

configurable, and programmable metasurfaces have been
Controlling electromagnetic (EM) waves has been of inter- proposed, which will soon evolve into intelligent metasur-
est for a long time and is widely used in current science and faces integrated with sensors, feedback networks, and artifi-
technology. For example, exploring flexible control of the cial intelligence (AI), responding smartly and automatical-
propagation of wireless signals in a real three-dimensional ly to environments to perform different tasks with less or
(3D) space to form smart radio environments with pro- even no human intervention. A schematic of the abovemen-
grammable signal processing capabilities and channel mod- tioned concept progression is illustrated in Figure 1, focus-
eling can lead to innovative communication systems in the ing mainly on the evolution from metamaterials to metasur-
future. EM wave propagation relies on the host medium in faces and then to tunable metasurfaces and intelligent meta-
which the wave propagates and the specific boundaries surfaces.
formed by the interface between two regions. Considering In the early stages, metamaterial research mainly fo-
these points, researchers have proposed the concept of a cused on control and realization of effective medium pa-
ELECTROMAGNETIC SCIENCE

metamaterial, a kind of artificially engineered material rameters that cannot be found in naturally occurring materi-
composed of subwavelength structures, to provide a new als. Such exotic EM parameters stem from the interactions
platform for engineering materials exhibiting abnormal EM between waves and subwavelength structures, typically
parameters such as zero permittivity, zero permeability, and termed the EM resonances of meta-atoms, leading to many
a negative index. More recently, metasurfaces, the two-di- unconventional physical phenomena and novel devices,
mensional (2D) version of metamaterials, have emerged as such as negative refraction [1], invisibility cloaks [2], su-
a tool for wavefront engineering, with the advantages of perlenses [3], and perfect absorption [4]. In a metamaterial,
low loss, low profile, low cost, etc. However, the static EM all the meta-atoms are usually formed with the same struc-
responses of passive metasurfaces still limit them when dy- ture consisting of metallic/dielectric materials, which are
namic tuning of EM waves is required, for example, in then periodically arranged in 3D space. Although metama-
 0020111-2 Electromagnetic Science, vol.1, no.2

Magnetic Thermal Optical

Progression from
metamaterial to RIS system
Electrical Mechanical Diodes Phase
transition

Metamaterial 2D gradient
metasurface Reconfigurable
metasurface
RIS system

FPGA
e refraction
ak Negativ
Invisible clo
r
generato Al
Vortices
ologram
Meta- lens H
rable
configu
- ste ering Re meta-lens
am
ic Be tenna
Dynam er array an
pola r iz oller
less or Contr
Wire ication Sens
Versatile v er age m u n
Co nsion com
functionalities tion- te
C o mputatelligent ex
a g e
m enable d i n
mic i sensin
g
Dyna

Figure 1 Development of metamaterials, metasurfaces, reconfigurable metasurfaces, and reconfigurable intelligent metasurfaces (RISs).

terials have introduced unprecedented capabilities in con- ization [25], [26], frequency [27]–[29] or direction [5], [30],
trolling the EM parameters and wave propagation, several [31] multiplexing, a single metasurface can integrate sever-
challenges still exist, including bulky volumetric occupa- al wave functionalities; however, these functionalities are
tion, high cost, high insertion loss, and fabrication complex- still static operations that cannot provide dynamic control.
ity, all of which undoubtedly limit their further develop- The capability to dynamically tune the wave functionalities
ments and applications. Metasurfaces composed of planar will significantly improve the potential use of metasurfaces
subwavelength-sized structures with the designable EM re- in real-world applications, which has gradually given rise to
sponses have been proposed to overcome these issues the concept of reconfigurable metasurfaces, including
[5]–[12]. Metasurfaces can be viewed as the 2D version of switchable, tunable, and programmable metasurfaces in
metamaterials that are typically arranged into a planar sur- recent years, to achieve dynamic control of EM waves
face with specific spatially varying distributions of various nearly across the whole spectrum from the low microwave
meta-atoms, so they have the advantages of low profile, low region to the terahertz and optical regions [8], [32]–[37].
cost, low loss, and easy fabrication. Moreover, unlike con- Generally, reconfigurable metasurfaces should contain tun-
ventional metamaterials, whose wavefront tailoring is real- able materials or components that external stimuli can ac-
ized by phase accumulation of the wave propagating inside tively control. The main idea for realizing reconfigurable
the media, metasurfaces introduce abrupt phase changes metasurfaces is the codesign of tunable components and
and thus field discontinuities across the interface between resonant structures, where a change in external tunings can
two media through engineering of the surface impedances result in a change in the tunable components and, eventual-
of the constituent meta-atoms. Hence, by designing meta- ly, the overall EM behavior of the meta-atoms. Then, recon-
atoms with a proper spatial variation, metasurfaces exhibit figurable output functionalities are achieved by applying
unconventional control of EM wave propagation and scat- various control signals to form inhomogeneous metasur-
tering and enable many inconceivable physical effects, such faces with specific phase profiles. Driven by the promising
as anomalous refraction [12], anomalous reflection [13], perspective of tunable metasurfaces, many efforts have
ELECTROMAGNETIC SCIENCE

and high-resolution holographic imaging [14]–[16]. Meta- been devoted to exploring tunable components, including
surfaces have also unlocked many intriguing functional de- diodes [32], [38]–[42], graphene [43], [44], liquid crystals
vices and applications, such as skin cloaks [17], ultrathin [45]–[48], and phase-change materials [49], that can be
metalenses [8], [18]–[21], special beam generation [22], [23], actively tuned by external stimuli of thermal effects [49],
and metasurface antennas [24]. [50], electric voltage bias [32], [38]–[43], [51], mechanical
With the rapid development of metasurface research, actuation [52]–[54], etc. [55]–[58]. Reconfigurable meta-
passive metasurfaces have gradually shown limitations in surfaces of these kinds have been designed and implement-
practical applications because their wave-manipulation ed for different practical devices and applications, such as
functionalities are static and fixed once designed or fabri- tunable lenses/imagers [8], [41], [42], [59], [60], dynamic
cated. Using proper multiplexing techniques such as polar- beam shapers [61]–[63], reconfigurable antennas [64]–[67],
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-3

and on-demand polarizers [68]–[72]. In addition, reflectar- and before the terminals receive them. Recent studies show
ray and transmitarray antennas can perform similar dynam- that RISs can be used to build a dynamically controllable
ic wave functionalities due to the tunable components (e.g., smart wireless environment by simply placing them on
diodes) in the constituent elements [73]–[79]. The reconfig- large scatter (e.g., buildings) surfaces. RISs provide real-
urable surface in an antenna system is designed to dynami- time software-defined EM responses that can fully use the
cally reshape the incident wave from a particular feed propagation effect of EM waves and thus maximize the per-
antenna placed near the surface. Overall, reconfigurable formance of entire communication networks in terms of ex-
surfaces pose many challenges in the design and fabrica- tending the signal coverage and improving the quality of
tion process because appropriate strategies to make struc- received signals [87], [88], [90]–[111]. They also enable
tures with a tunable performance in a given frequency metasurface-assisted wireless communication by modulat-
domain should be found, while an additional bias network ing the transmitted information directly based on the EM
or control system would further increase these difficulties, performance of the metasurface [112]–[116]. As a long-
especially for the cases of individually addressable ele- term prospect, the RIS-based platform is envisioned to per-
ments, high-frequency operation, and surfaces with large- form joint tasks as much as possible, such as communica-
scale elements. tion, sensing, localization, computing, signal processing,
More recently, reconfigurable metasurfaces have fur- and automatic learning, with low cost, high throughput, and
ther evolved into reconfigurable intelligent metasurfaces high reliability. Overall, RISs gradually evolved from meta-
(RISs) and RIS-based systems, providing a smart paradigm materials, metasurfaces and reconfigurable metasurfaces, so
for EM wave manipulation [52], [59], [60], [80]–[106]. A they have the advantages of the previous artificially engi-
typical RIS comprises a reconfigurable metasurface, a soft- neered materials. Nevertheless, the more advanced the ca-
ware and hardware control system, and sensing or feedback pabilities or more versatile the wave functionalities de-
components, which are often integrated with algorithms to manded are, the more challenges RISs need to overcome,
enable self-adaptivity via their reprogrammable wave func- together with the pressing demands from the application
tionalities, with less or even no human intervention. From side, because one should consider the codesign of several
the viewpoint of output wave functionalities, an RIS is sim- parts far beyond material engineering.
ilar to a reconfigurable metasurface, both of which can ex- This article reviews recent research advances in this
hibit dynamic reflection/transmission EM responses to form field, focusing mainly on designs in the microwave regime
tunable inhomogeneous profiles at the metasurface aper- with an emphasis on our own work. After a comprehensive
tures. However, RISs make strides toward intelligent con- comparison of reconfigurable metasurface implementations
trol of EM waves: the sensing and feedback components act with electrical, thermal, mechanical, and other tuning mech-
as the “eyes ” to see the environment or meet the require- anisms, we highlight the experimental implementations of
ments, while onsite or cloud algorithms act as the “brains” RIS systems by exhibiting their practical uses in dynamic
to determine their output EM behavior. Hence, RISs have wave control, local wireless signal enhancement, etc. Final-
strong self-decision capabilities in which they can intelli- ly, we conclude this review, followed by our viewpoints on
gently perform a series of successive software/algorithm- the challenges in this emerging interdisciplinary research
defined tasks, such as smart beam shaping [85], adaptive area and perspectives on this field’s future developments.
retroreflection [52], and gesture recognition [59], [89]. Al-
though RIS research has been widely performed across II. Implementations of Reconfigurable Metasur-
nearly the whole spectrum, thus far, most of the reported faces
RISs are aimed at engineering microwaves for the follow- There are many approaches to realize reconfigurable meta-
ing reasons. First, the necessary physical basis for con- surfaces, which depend on the different working mecha-
structing such RISs is comparatively mature. Commercial nisms of tunable meta-atoms. By including stimuli-respon-
diodes can provide very fast switching or continuously tun- sive materials in their designs, meta-atoms may undergo
able speeds and are well compatible with the output volt- relatively large and rapid changes in their physical proper-
ages from widely used controllers, for example, field-pro- ties in response to external ambient stimuli [80], [117]–[123].
grammable gate arrays (FPGAs), while the standard print- In this case, when enacting external control based on the
ELECTROMAGNETIC SCIENCE

ed circuit board (PCB) process provides mature technology ambient conditions, such as biasing with an external cur-
for large-scale sample fabrication. The FPGA controller rent or voltage, applying an electric/magnetic field, heating
supports parallel communication and output voltages, en- or cooling with temperature changes, exerting mechanical
abling the phase profiles of massive meta-atoms to syn- pressure, or providing light illumination, the material prop-
chronously switch from one to another. Second, RISs have erties and the interaction between the meta-atom and the
emerged as a new technology for the next generation of EM wave will be tuned accordingly and therefore induce a
wireless communication and may solve problems in real- change in the metasurface functionality.
world applications. In the current wireless communication
scheme, microwave propagation cannot be controlled and 1. Different tuning mechanisms
customized after the waves are emitted from the sources In Figure 2, we briefly summarize the different tuning
 0020111-4 Electromagnetic Science, vol.1, no.2

mechanisms that have been applied in current implementa- Realizing local tuning through thermal, optical, and mag-
tions of reconfigurable metasurfaces as well as the corre- netic means for reconfigurable meta-atoms is quite difficult
sponding external stimuli used to actively control the dy- and complicated due to the additional large objects com-
namic modulation of metasurfaces. pared to the unit cell size that need to be involved in the
Both global and local tuning can be achieved through unit-cell design even in the microwave range. However,
these mechanisms to realize reconfigurable EM properties electrical control utilizing PIN diodes, varactors or micro-
of the whole metasurface or each individual meta-atom, re- electromechanical system (MEMS) switches, which have
spectively. In the local tuning scenario, more reconfig- comparably small sizes and simple biasing circuits, can be
urable functions, such as EM beam steering, dynamic fo- easily implemented in reconfigurable metasurfaces even
cusing, imaging and holography, can be achieved, but ex- with very many individually tunable unit cells and large
ternal stimuli should be applied locally to each meta-atom. physical sizes [83].

PIN diode, Varactor Tunable

Microwave

Ferroelectrics meta-atom
Ferromagnetics External active
Graphene control
MEMs,···
Graphene
Liquid crystal Voltage bias
Phase change Current bias
Terahertz

material (VO2, GST,···) Electric field

Semiconductor Magnetic field
(doped silicon,···) Temperature change
Superconductor,··· Mechanical deformation
Microfluids
Nonlinear material Light illumination
Optical

Phase change EM power

material (GST,···) ···
Reconfigurable
NEMs
metasurface
Microfluids,···

Figure 2 Different mechanisms to implement reconfigurable metasurfaces.

Table 1 ([39], [46], [48], [52], [53], [56], [124]–[158]) driven reconfigurable metasurface based on GST could
summarizes most of the stimuli-responsive mechanisms that achieve an eleven-fold change in the reflectance (absolute
have been used to realize tunable and reconfigurable meta- reflectance contrast reaching 80%) and quasi-continuous
surfaces in different operating regimes from megahertz to spectral tuning over 250 nm [136].
optical frequencies through various external stimuli. Recon- Photoconductive semiconductor materials, such as Si
figurable functionalities, such as tunable absorbers or polar- and GaAs, can be employed for implementing reconfig-
izers, dynamic beam forming or steering, dynamic lenses or urable metasurfaces by tuning their conductivity through
holograms, tunable scattering or retroreflection, informa- carrier photoexcitation under optical injection [137], [138].
tion modulators and wireless communication, can be The first reported optically controlled metasurface was a
achieved utilizing these approaches via electrical, thermal, patterned copper split-ring resonator (SRR) array on a
optical, magnetic or mechanical control. Among various ex- GaAs substrate, which demonstrated that under ultrafast
ternal stimuli, electrical control is largely used, as there are light pumping at 800 nm, the photoexcitation-induced con-
many electrically sensitive materials, including nematic liq- ductivity in GaAs can provide strong modulation of the
uid crystals (LCs), doped indium tin oxide (ITO) and alu- transmission amplitude of a terahertz wave [137]. In a
minum-doped zinc oxide [48], [124]–[127]. Semiconduc- diffractive array of GaAs semiconductor resonators that
tors, 2D materials (e.g., graphene), and transparent conduc- support both dipolar and quadrupolar Mie resonances,
tive oxides (TCOs), such as electric circuit elements, e.g., through optical injection of free carriers to spectrally shift
ELECTROMAGNETIC SCIENCE

PIN diodes and varactors, can also be loaded into meta- the multipoles and rebalance the multipole strengths, radia-
atoms whose EM characteristics (e.g., capacitance) can be tion into a diffraction order is enabled on an ultrafast
dramatically tuned by applying various voltages [39], timescale [138]. Graphene-based thin films can also be em-
[128]–[132]. Phase-change materials (PCMs), such as ger- ployed to realize absorption modulators for operation in the
manium-antimony-telluride (Ge2Sb2Te5, also called GST) THz regime with ultrafast optical control. Modulation of
and vanadium dioxide (VO2), which exhibit reversible absorption on the order of 40% within a few picoseconds
phase transitions with significant changes in their refractive has been achieved by applying an optical pump signal,
index, can be embedded in meta-atoms to realize tunable which modifies the conductivity of the graphene sheet
functions upon local temperature changes induced by elec- [139]. Ultrafast plasmon modulation in the near-infrared
trical heating [133]–[136]. For example, an electrically (NIR) to mid-infrared (MIR) range was successfully
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-5

Table 1 Summary of different approaches to realize tunable and reconfigurable metasurfaces

External stimuli Tuning mechanism Operation regimes Materials Function Frequency Ref.
Semiconductors Modulator 0.81/0.89 THz [124]
Carrier doping GHz to Visible Graphene Dynamic scattering 12.9−13.9 GHz [125]
TCO Beam steering/Lens 1.5−3 μm [126]
Beam steering 32.5 GHz [127]
Liquid crystals
Beam steering 0.67 THz [48]
Modulator 1.2−2 μm [133]
Phase transition GHz to Visible VO2
Dynamic imaging 0.42–0.48 THz [134]
Electrical Modulator 0.5−0.8 μm [135]
GST
Modulator/Beam manipulation 1.4−1.7 μm [136]
Tunable scattering 2.1−2.9 GHz [39]
PIN diodes Dynamic beam manipulation 6/9.8 GHz [128]
Dynamic beam manipulation 9.5−11.5 GHz [129]
Capacitance MHz to GHz
Dynamic metalens 6.7−7.1 GHz [130]
Varactors Beam steering 3.15 GHz [131]
Tunable absorber 4−5.3 GHz [132]
Modulator 0.56 THz [137]
Semiconductors
Dynamic beam manipulation 1−1.1 μm [138]
Carrier doping THz to Visible Graphene Tunable absorber 2.2 and 6.5 THz [139]
Modulator 1.4−2.2/3.5−5 μm [140]
TCO
Modulator/Polarizer 0.5−0.9 μm [141]
Optical
VO2 Modulator 0.41 and 0.8 THz [142]
Phase transition THz to Visible Dynamic lens/Hologram/Modulator 3.75 μm [143]
GST
Modulator 0.72−1.13 THz [144]
Dynamic cloak/Vortex beam 5.62−6.62 GHz [56]
Light-microwave transmitter GHz Photodiode & Varactors
Wireless communication 4−6.3 GHz [145]
GST Dynamic beam/Lens 3.1 μm [146]
Liquid crystals Dynamic beam steering 0.75−0.8 μm [46]
Tunable absorber 5−20 μm [147]
Thermal Phase transition THz to Visible VO2
Modulator 1.35−1.65 THz [148]
Superconductor Modulator 0.2−0.5 THz [149]
Semiconductor Modulator 1.4−1.5 μm [150]
Nonreciprocal shift 4.76 GHz [151]
Ferrite (YIG)
Tunable absorber 8−12 GHz [152]
Magnetic Magnetostatic field GHz to Visible
InAs Tunable absorber 0.1−5 THz [153]
PDMS/Fe Tunable reflector 0.7−1 μm [154]
Micromotor Retroreflector 3.7−4.6 GHz [52]
MEMS/NEMS Modulator 1.2 and 1.6 μm [155]
ELECTROMAGNETIC SCIENCE

Tunable lens 7.8−18 GHz [53]

Mechanical Mechanics GHz to Visible Structural deformation
Tunable lens 0.6 μm [156]
Retroreflector 15 GHz [157]
Microfluids
Tunable colors 0.5−0.6 μm [158]

demonstrated by intraband pumping of a metasurface com- an optical interrogation network can convert visible light il-
posed of ITO nanorods, providing a useful method of opti- lumination patterns to voltages and apply a bias to the meta-
cal tuning by tailoring the ITO metasurface geometry [140], surface elements, which could provide an interesting ap-
[141]. By combining a varactor with a photodiode, a digital proach to generate specific phase distributions for realiza-
metasurface based on electronic varactors integrated with tion of various wavefront tailoring devices and direct wire-
 0020111-6 Electromagnetic Science, vol.1, no.2

less communication [56], [145]. overall structure shape or periodicity of the meta-atoms,
Thermal control is another candidate for the ambient thereby realizing on-demand reconfigurability. For exam-
stimulus for reconfigurable metasurfaces, where PCMs with ple, a kirigami-based reconfigurable gradient metasurface
temperature-dependent permittivity are often employed in has been proposed and successfully applied in designing a
tunable meta-atom design. For example, VO2, as a com- metalens with a tunable focal length or chromatic aberra-
monly used PCM, could behave as an insulator at room tion-free focusing [53]. An origami metawall has been de-
temperature and change to a metallic state at higher temper- signed and demonstrated with reconfigurable functions of
atures due to the enhancement of the free carrier concentra- an absorber, a mirror, or a negative reflector depending on
tion. It has been utilized to design a tunable meta-atom to mechanical stretching or compression of the structure [161].
shift its resonance frequency and modulate the transmis- Most of the above discussed tuning mechanisms can
sion amplitude via direct substrate heating [147], [148]. be applied to different frequency bands from the micro-
Chalcogenide PCMs such as GST have also been em- wave to optical range. In particular, achieving reconfigur-
ployed in tunable metasurface design, where the GST layer able metasurfaces for applications from gigahertz to a few
can be partly switched from the amorphous to crystalline terahertz is relatively easy, and the control methods to ad-
state to generate intermediate refractive indices enabling dress each meta-atom to realize spatial variation of the EM
beam switching and bifocal lensing [146]. In addition, su- properties of the metasurface are more practical and can
perconductor materials are also employed to design tunable mostly be implemented with external electrical stimuli. In
metasurfaces, as their conductivity strongly depends on the next sections, we will focus on the designs and imple-
temperature, which may induce the transition from a super- mentations of microwave reconfigurable metasurfaces and
conducting state to a normal state. By designing an SRR ar- their practical applications.
ray with a high-temperature superconducting film [149] or a 2. Microwave reconfigurable metasurfaces
niobium nitride film [159], efficient metamaterial reso- In the microwave frequency regime, PIN diodes, varactors,
nance switching or transmission modulation can be real- and other voltage-driven circuit elements have been widely
ized by varying the ambient temperature. used to design tunable meta-atoms to achieve reconfigur-
Other tuning approaches, including magnetic and me- able EM functionalities, as they have compact sizes so can
chanical control, are summarized in Table 1. Magnetically be easily embedded in the passive metallic meta-atom struc-
responsive structures have been explored through ferrite- ture. Although PCMs such as liquid crystals or graphene
based metasurfaces to provide a magnetically controllable sheets have also been employed in microwave tunable
nonreciprocal Goos-Hänchen shift enabled by magnetic metasurfaces, they are comparably complicated in the real-
plasmonic gradient metasurfaces consisting of an array of ization of external control, especially for large-scale meta-
ferrite rods [151] or magnetically tunable perfect absorp- surfaces with individual meta-atom addressing [125], [162].
tion based on ferromagnetic resonance with a tunable work- Figure 3 shows a design example of a reconfigurable
ing frequency controlled by the applied magnetic field [152]. reflective metasurface for achieving 1-bit phase control for
Magnetic field-driven polymer microplate arrays have also dual-polarization incident EM waves [40]. The meta-atom
been proposed to modulate the optical properties by mag- shown in Figure 3(a) has a metallic resonant structure with
netically controlling the tilt angle of the microplates to re- central and rotational symmetry, which can respond to dual
versibly switch their surface reflectivity from a higher to polarization and can maximally suppress the cross-polariza-
lower state [154]. Dynamically reconfigurable metasur- tion level of the reflected waves. Four PIN diodes are sym-
faces can also be realized by mechanically changing the ge- metrically loaded into the meta-atom to realize a 1-bit ele-
ometries of the constituent meta-atoms or altering the dis- ment with the features of high efficiency, programmability,
tances between adjacent meta-atoms. For example, MEMSs and independent dual-polarization control, where both dual-
or nanoelectromechanical systems (NEMSs) are employed polarized reflection phases can be dynamically switched be-
to realize tunable metasurfaces either in microwave and ter- tween 0° and 180° under DC current biasing. The diodes
ahertz bands or in optical frequency ranges, respectively connecting the central patch to the four side patches (top
[155], [160]. A reconfigurable omnidirectional adaptive metallic layer) are independently biased through metallic
retroreflector has been realized by dynamically and contin- vias from the ground plane (middle metallic layer) and bot-
ELECTROMAGNETIC SCIENCE

uously controlling the reflection phase of constituent meta- tom layer. The ground plane acts as one electrode, and the
atoms through alteration of their orientation states, which biasing network (bottom metallic layer) under the ground
can be individually mechanically addressed with a tiny mi- plane acts as the other electrode, thus enabling each ele-
cromotor connected to each meta-atom [52]. Metasurfaces ment to work independently in both polarization channels.
fabricated on flexible or stretchable substrates can offer a Placing the ground plane between the resonant layer and the
direct way to enact reconfigurability upon mechanical de- biasing layer could remarkably weaken the deterioration ef-
formation of the structures [53], [156]. In this scenario, both fect of the dense metallic bias lines on the overall reflec-
origami- and kirigami-based structures could be useful so- tion response of the metasurface element, which is often
lutions to create deployable continuous-state tunable meta- used in tunable reflective metasurface design. The simulat-
surfaces in which a folding pattern enables a change of the ed reflection phases and amplitudes are shown in Figure
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-7

3(b), which illustrate that the phase difference is approxi- pends on the PIN diodes welded in the parallel direction un-
mately 180° between the two working states, and the reflec- der specific polarized waves and is not affected by the
tion amplitudes are all over 0.96 within the operation band- working states of PIN diodes in the orthogonal direction,
width with a center frequency of 7.45 GHz. More impor- enabling independent phase control of both polarization
tantly, the reflection response of the meta-atom only de- channels.

(a) h (b)
360 1.0
0/0
0/1
180

Amplitude
1/0

Phase (°)
1/1 0.8
p
0

y
−180 0.6
a
7.2 7.4 7.6 7.8
x PIN diodes
z Frequency (GHz)

(c) (d)
x-polar. y-polar.
0 0
−30 30 −30 30
Normalized gain

−60 60 −60 60

−90 90 −90 90
−20 dB 0 dB −20 dB 0 dB

Figure 3 One-bit reconfigurable metasurface for dual polarization [40]. (a) Schematic of the dual-polarized coding metasurface elements; (b) Simulated
reflection phases and amplitudes. The numbers before and after the slashes (/) represent the coding states along the z - and y -directions. Measured radia-
tion patterns for beam scanning from −40° to 40° with a 10° step at 7.45 GHz in the (c) z -polarization channel and (d) y -polarization channel.

As an application of the 1-bit programmable metasur- bias voltage, which have been widely investigated for real-
face, a reconfigurable reflectarray (RRA) with independent izing continuously tunable metasurfaces, such as the mi-
dual-polarized beam steering has been demonstrated. The crowave absorbers with a tunable working bandwidth or
RRA is entirely driven by the external digital signals from tunable absorption [132], [164] or reconfigurable metalens-
the FPGA-based hardware system. By addressing each es with a tunable focal length or a dynamic focal point [130].
meta-atom with an independent DC voltage to achieve dif- The typical capacitance range for commercially available
ferent spatial phase distributions, wide-angle beam scan- varactor diodes can be a few pico-Farad, e.g., 0.5–3.5 pF in
ning is obtained in dual-polarized channels, with low cross- the GHz frequency band under a bias voltage ranging from
polarization levels and acceptable side lobe levels, as 0 V to 20 V. By embedding varactors in meta-atoms, con-
shown in Figures 3(c) and (d). The measured results show tinuous phase manipulation can be obtained via voltage
that wide-angle beam scanning up to ± 40° can be obtained control that may cover the full range of 360°, which can be
at approximately 7.45 GHz, with peak gains of 21.13 dBi easily utilized to achieve 1- and 2-bit or even multibit phase
and 20.89 dBi in the dual channels, corresponding to aper- states for more complicated EM wavefront or beam form-
ture efficiencies of 16.35% and 15.47%, respectively. Such ing control.
a dual-polarized RRA may provide a new method for multi- Although utilizing PIN diodes or varactor diodes is the
ELECTROMAGNETIC SCIENCE

beam antenna design, which will be conducive to the real- most widely used approach to realize microwave reconfig-
ization of the large-scale multiple-input multiple-output urable metasurfaces with fast switching between different
(MIMO) antennas for 5G technology [163]. In addition, the phase states, the insertion loss caused by electronic devices
dual-polarization working mode will highly enhance the can lead to deterioration of the operation efficiency. Anoth-
storage densities and information capacities of the systems, er issue is that the operation of these diode-embedded meta-
offering potential capabilities of fast information communi- surfaces always requires a power supply, which may in-
cation and detection. crease energy consumption, especially for large-scale meta-
Unlike PIN diodes that can only be switched between surfaces with many elements. To overcome these chal-
two states, varactor diodes can be adjusted to achieve a con- lenges, an alternative strategy of controlling the tunable ele-
tinuously varying capacitance upon changing the reverse ment via mechanical devices, such as micromotors [52],
 0020111-8 Electromagnetic Science, vol.1, no.2

[165], piezoelectric actuators [166], and spring devices [167], metallic ground and the dielectric substrate, can be continu-
is gradually emerging for microwave reconfigurable meta- ously modulated by the stepping motor. The entire metasur-
surfaces. A higher phase resolution can be effectively face shares a fixed dielectric substrate printed with periodic
achieved by continuously performing rotational operations metallic patches, as illustrated in the fabricated prototype
[52], [168], [169] on the tunable element or changing the shown in Figure 4(b). Notably, this mechanically con-
height/thickness [166], [168], [170] in the structure (chang- trolled RIS possesses both high efficiency and continuous
ing its resonance properties or optical path difference) via phase modulation with a full range of over 360°, as shown
suitable mechanical devices. in Figure 4(c). For practical implementation, the stepping
As a simple example, we show a typical design of a re- motor of each unit cell is operated by an independent DC
flective metasurface mechanically controlled by tuning the voltage and controlled by an FPGA (shown in Figure 4(d)).
thickness of an air layer. As shown in Figure 4(a), the unit The greatest advantage of this mechanically controlled re-
cell is designed in the S-band with a center frequency of 2.55 configurable metasurface is that it only needs to be pow-
GHz. For each unit cell, 4 × 4 square metallic patches even- ered when switching between different tuning states for
ly distributed on a dielectric substrate serve as the resonant beam shaping requirements, which is potentially suitable
structure. A movable metallic ground controlled by a step- for some application scenarios that require high energy con-
ping motor is set behind the dielectric substrate. Hence, the servation but not high reconfiguration frequency, such as
air thickness h, i.e., the distance between the movable indoor signal enhancement and wireless power transfer.

(a) (b)

h
Unit cell

Stepping
motor
y

z x

y
z
(c) x (d)
360
1.0
Reflection amplitude

300
Reflection phase (°)

0.8
240
0.6
180
120 0.4

60 0.2
z y
0 0 x
0 2 4 6 8 10 12 14
h (mm)

Figure 4 Mechanically controlled reconfigurable metasurface with a tunable air thickness. (a) Proposed structure of the unit cell; (b) Photograph of the
fabricated prototype; (c) Simulated amplitude and phase responses versus the air thickness; (d) Schematic of the operation mode of the mechanically con-
trolled metasurface.

III. Practical RIS Systems and Demonstrations in unit (MCU) or FPGA to directly program the states (reflec-
ELECTROMAGNETIC SCIENCE

Wireless Communication tion or transmission phase, magnitude, polarization) of each

meta-atom with different spatial coding sequences in a real-
An RIS is a 2D metasurface structure that is reconfigurable
time control manner. By judiciously designing and imple-
in terms of its functionality when interacting with propagat-
ing EM waves. Practical RIS systems are mainly composed menting the RIS system, the incident EM wave signals can
of reconfigurable metasurfaces integrated with tunable or be manipulated through either a reflection [83] or transmis-
switchable meta-atoms to realize passive EM wave manipu- sion [130] channel, or even both channels [171], with high
lation with controllable wavefront, intensity, polarization design freedom in the wave propagation direction, polariza-
and beam direction. Various dynamic EM wave responses tion, and intensity. For example, an RIS can focus the inci-
can be enacted by reconfigurable metasurfaces through an dent EM power to predesigned areas to change the wave in-
intelligent control system with a microprogrammed control tensity distribution and increase certain local signal cover-
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-9

age. Owing to their capability for proactive EM wave ma- of the resulting channel cross-talk and reduction of the in-
nipulation, RIS systems have been recognized as a key formation capacity in wireless communication [102].
technology to modify the wireless communication environ- The working frequency bandwidth of an RIS for pas-
ment and have become a hotspot of research in wireless sive beam forming of wireless signals via reflection or re-
communications to achieve smart and reconfigurable wire- fraction is dominated by the reconfiguration mechanism of
less channel/radio propagation environments for next-gen- the meta-atoms interacting with the EM waves. In most cas-
eration wireless communication systems [87], [88], [90], es, tunable EM resonance is employed to enact phase shifts
[92], [172]. By deploying RIS systems and smartly coordi- or magnitude variations to reconfigure the wireless signals
nating their wave interaction properties, the wireless chan- output from the RIS through external electrical biasing.
nels between transmitters and receivers can be flexibly re- Such EM resonance may possess a strong dispersive nature
configured to circumvent obstacles and fulfil an enhanced and limit the signal frequency to a single working band.
received signal at the terminal devices, which may provide However, by judiciously designing meta-atoms with multi-
a fundamental way to tackle wireless channel fading and in- mode resonance through interleaved or multilayered struc-
terference issues and potentially improve the wireless com- tures, broadband or multiband RISs can be achieved. For
munication capacity and reliability. RISs possess many ad- example, a 1-bit high-efficiency programmable metasur-
vantages, such as easy implementation and deployment, face with a bilayer structure has been reported to achieve
spectral efficiency enhancement through the construction of completely independent functions with real-time reconfig-
a software-defined wireless environment, greater energy ef- urability in both the C-band and X-band [128].
ficiency and environmental friendliness than conventional The switching speed between different phase or mag-
relaying systems, and compatibility with the standards and nitude states in each element of the reconfigurable metasur-
hardware of existing wireless networks. face is also an important characteristic that is dominated by
1. Practical considerations and requirements of RIS sys- the response time of the tunable element used in the meta-
tems atom. PIN diodes and varactors are the most commonly
used devices in the microwave region for reconfigurable
For applications in wireless communication scenarios, the meta-atom phase or amplitude, which have much faster dy-
main practical consideration is the hardware implementa- namic responses than other methods, such as utilization of
tion of the RIS system, including both the reconfigurable mechanical actuators, liquid crystals, or PCMs [186]. A
metasurface and its electronic controller. Many issues influ- switching time for beam forming within several tens of
ence the performance of RISs, such as the maximum num- nanoseconds can be obtained by employing microwave PIN
ber of meta-atoms in the metasurface with all integrated diodes, which is sufficient for RIS-assisted wireless net-
bias circuitry for individual addressing, the number of works [92]. For the application of an RIS-based direct-sig-
quantization levels of the RIS phase or magnitude respons- nal-modulation transmitter, increasing the switching speed
es, and the percentage of the scattering environment that through innovative designs is required to achieve a much
can be coated by an RIS. The deployment of an RIS system better data rate.
is often limited by the tradeoffs caused by the hardware In addition to the EM properties and architecture de-
constraints based on analyzing their effect on the channel sign of RIS systems, both the software of efficient algo-
modification [173], the complexity and power consumption rithms for channel estimation and signal processing and the
of the control system, and performance metrics such as the hardware for controlling signals are key design aspects that
outage probability [174] and ergodic capacity [175]. need to be considered and realized with low power con-
Table 2 ([40], [59], [82], [85], [115], [116], [131], [171], sumption and low complexity. To enable the flexible de-
[175]–[185]) summarizes the recent experimental demon- ployment of RISs, the inherent limitations of wired lines
strations and prototypes of RIS systems. Various tunable and power lines for controlling them cannot be underesti-
functions have been demonstrated with RIS systems, such mated. In most demonstrated RIS systems, computer-con-
as the dynamic EM beam steering and focusing used in trolled FPGAs or microcontrollers are involved to provide
RIS-assisted wireless communication to adaptively modify highly reliable and easily implementable signal and data
the propagation channel and improve the performance and processing. However, many issues, including signal integri-
the signal modulation through the phase, magnitude or po-
ELECTROMAGNETIC SCIENCE

ty, power consumption, and remote control, still need to be

larization of the output EM wave that may enable a direct- properly resolved.
signal-modulation transmitter for a new architecture of
wireless communication. Most reconfigurable metasurfaces 2. One-bit reflective RIS systems
utilize PIN diodes as tunable or switching elements in meta- By deploying RIS systems in the environment, e.g., coated
atoms to achieve one- or two-bit phase modulation for easy on the walls of buildings or rooms, they can bring addition-
and robust control with simple digital control circuitry. Al- al phase shifts to the reflected signals, and by jointly opti-
though multibit phase quantization may provide more free- mizing the phase shifts of all scattering elements, namely,
dom in EM wavefront manipulation, recent theoretical stud- passive beamforming, the signal reflections can be coher-
ies indicate that a one-bit coding RIS system may yield sat- ently focused at the intended receiver and nulled in other di-
isfactory performance in most practical scenarios in terms rections. Thus, a smart radio environment can be estab-
 0020111-10 Electromagnetic Science, vol.1, no.2

Table 2 Summary of experimental demonstrations and prototypes of RIS systems

Modulation/
Phase/Polarization/
Controller Dimensions Switching Frequency range Realized tunable functions Ref.
Control
speed
2-bit phase, single-pol.,
DC voltage source 6 × 6 array – 5 GHz Beam scanning over a 100° × 100° window [176]
varactors
1-bit phase, single-pol., Establishment of a robust link between
Microcontroller 14 × 16 array – 60 GHz [177]
relay switches transceivers by beam steering
2-bit phase, single-pol., Real-time reprogrammable digital-
FPGA 24 × 32 array – 3.2 GHz [59]
PIN diodes metasurface imager
2-bit phase, single-pol., Self-adaptive adjustment of EM radiation
FPGA 30 × 30 array – 9 GHz [85]
PIN diodes beams, single- and multi-beam steering
2-bit phase, single-pol., Harmonic manipulation and BFSK wireless
FPGA 16 × 16 array 78.125 kHz 3.6 GHz [115]
varactors communication system
1-bit phase, single-pol., Backscatter wireless communication with
FPGA 768 elements 500 kHz 2.4 GHz [178]
PIN diodes commodity Wi-Fi signals
2-bit phase, dual-pol., Reflect antennas for beam scanning at
FPGA 16 × 16 array – 2.3 GHz, 28.5 GHz [175]
PIN diodes 2.3 GHz & 28.5 GHz
1-bit amplitude, single- Improvement of the signal strength and
Raspberry Pi
3200 elements pol., 1 MHz 2.4 GHz channel capacity for indoor wireless [179]
controller
RF switch communication
1-bit phase, dual-pol., Beamforming both in the near field and far
FPGA 20 × 20 array 100 kHz 27.5–29.5 GHz [82]
PIN diodes field
2-bit phase, single-pol., Improvement of the quality of indoor wireless
FPGA 576 elements – 2.55 GHz [180]
PIN diodes communication
2-bit phase, single-pol., Space- and frequency-division multiplexing
FPGA 16 × 8 array 2.5 MHz 9.5 GHz [181]
PIN diodes for dual-channel wireless communication
3-bit phase, single-pol.,
DC voltage 10 × 10 array – 3.15 GHz Angle-insensitive beam steering [131]
varactors
Dynamic lens switching between single focus
Voltage bias 50 × 50 cm2 PZT actuators – 27.2–28.2 GHz [182]
and dual focus
1-bit phase, single-pol., Broadband manipulation of harmonics for
FPGA 56 × 20 array 1.875 MHz 22–33 GHz [116]
PIN diodes 256 QAM wireless communication
1-bit phase, single-pol., Dynamic beam scanning via reflection,
FPGA 20 × 20 array – 6.23 GHz [171]
PIN diodes transmission, or both
1-bit phase, dual-pol., Dual-polarized wide-angle beam-scanning up
FPGA 20 × 20 array – 6.9–7.7 GHz [40]
PIN diodes to ± 40°
2-bit phase, single-pol., Transmissive RIS with a high data rate, 2D
FPGA 16 × 16 array – 25–28.6 GHz [183]
PIN diodes beamforming up to ± 60°
3-bit phase, single-pol., Improvement of indoor wireless
Microcontroller 120 × 120 cm2 – 3.5 GHz [184]
varactors communication
1-bit phase, dual-pol., Improvement of indoor 5G millimeter-wave
Microcontroller 112 × 112 mm2 ~4s 32–36 GHz [185]
gear system wireless communication

lished that can assist information sensing, analog comput- each element needs to be low cost. The meta-atoms in an
ing, and wireless communication [187]. With optimal con- RIS need to have consistent phase and amplitude responses
trol of the signal transmitter and terminal receiver, RIS-as- under external bias control and moderate angle stability un-
ELECTROMAGNETIC SCIENCE

sisted wireless systems will become more flexible to sup- der oblique incidence so that a single RIS can be used to
port diverse user requirements, e.g., enhanced data rate, ex- serve different terminals at different locations.
tended coverage, minimized power consumption, and more In the following, we will give a comprehensive design
secure transmissions [87], [188], [189]. example of a practical reflective RIS system and showcase
An RIS system needs to have sufficient physical size its real application in enhancing the signal coverage of a
to compensate for the link budget deficit that results from wireless network. For practical considerations, a reflective-
its nearly passive implementation without any power ampli- type reconfigurable metasurface is chosen in the RIS sys-
fication. In general, at least hundreds of elements must be tem due to its achievable high efficiency. Both 1-bit and 2-
equipped in an RIS system to offer a competitive gain. To bit phase states are realized by loading PIN diodes into the
realize affordable RIS implementation and deployment, EM resonant structure in the meta-atoms [180]. As illus-
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-11

trated in Figure 5(a), a simple H-shaped metallic structure induced currents are distributed in the top metallic layer,
backed by another metallic strip forms a capacitively cou- which ensures a low-loss reflection mode for the meta-
pled EM resonator under illumination of a y-polarized atom. The geometric structure is optimized to function as a
wave. A slot is cut in the middle strip and connected to a 1-bit phase switchable reflector at a center frequency of ap-
PIN diode, and a third layer of metallic sheet is used as a proximately 2.55 GHz suitable for the 5G wireless network.
ground plane to block the EM wave for reflection operation. Figures 5(b) and (c) show the simulated reflection phases
By switching on and off the PIN diode through DC voltage and magnitudes of the designed meta-atom. By applying di-
biasing, the resonance will be perturbed with a 180° reflec- verse bias voltages, a reflection phase change of approxi-
tion phase change, enabling a 1-bit reflective meta-atom. mately 180° and a high magnitude efficiency of over 80%
Such a design of a double-layer resonator ensures that most are successfully achieved at 2.55 GHz.

(a) (b)
0

Reflective amplitude (dB)

PIN on
−3
PIN off
z
y −6
x

−9

PIN diode 2.1 2.4 2.7 3.0

Frequency (GHz)

(d)

First layer Second layer

(c)
100
PIN on
0
Reflective phase (°)

PIN off

−100

−200

−300

−400
2.1 2.4 2.7 3.0
Frequency (GHz)

Figure 5 RIS realized with a 1-bit reflective metasurface. (a) Schematic of the meta-atom; Simulated (b) reflection amplitudes and (c) phases of the meta-
atom; (d) Realized RIS system.

A single metasurface is designed with 14 × 14 meta- nals in a weak coverage area, the RIS system can be operat-
atoms with a size of 470 × 470 mm2, and several metasur- ed to redistribute the EM wave channels from the transmit-
face sheets can be easily assembled together to realize dif- ter to the receiver. To this aim, the RIS system controller
ferent sized RIS boards required for practical deployment, uses signal intensity measurements from the local end-
ELECTROMAGNETIC SCIENCE

as shown in Figure 5(d). To control the PIN diode in each points to maximize the signal strength at a particular loca-
meta-atom, we connect an off-surface control unit to the tion, which is realized through an optimization procedure
backside of the RIS board via ribbon cables with on-sur- employing the strengthen elitist genetic algorithm (SEGA)
face biasing lines, which altogether comprise the entire con- [190]. In the optimization process, different coding se-
trol circuit network. The phase state of each meta-atom is quences are evaluated to create dynamic irregular beam
controlled through the bias network with FPGAs by enact- shaping via the RIS to increase the signal intensity at the re-
ing different binary coding sequences in a wireless remote ceiver based on the feedback from the measured power lev-
manner through regular Wi-Fi signals, facilitating movable els of the endpoints.
deployment of the RIS boards. We have evaluated the RIS system in various indoor
To demonstrate the ability to enhance the wireless sig- field tests. As an example, as shown in Figure 6, the RIS
 0020111-12 Electromagnetic Science, vol.1, no.2

system is deployed in an indoor scenario including a corri- (a maximum of 33 dB) for the weak coverage area (in the
dor, interior walls, windows, and a large metallic door (blue left corner). We also consider the scenario in which the two
rectangle in Figure 6. In the test, a signal transmitter work- RIS boards are located at different places, such as face to
ing at 2.55 GHz is located at one end of the long corridor, face at the two sides of the corner area, as shown in Figure
and due to blocking of the EM waves by the metal door, 6(c). Although the RIS deployment is changed, similar
weak coverage of the wireless signal is observed in the low- wireless signal coverage to that in Figure 6(b) can be ob-
er left corner at the other end of the corridor (denoted by the tained with the beam forming ability through the two RIS
yellow dashed circle in Figure 6(a)). The wireless power in- boards, indicating that the influences of the RIS deploy-
tensity distribution in the corner is measured at 20 different ment positions can be largely eliminated by optimization of
locations using an omnidirectional dipole antenna, and the the irregular beam shaping. We also note that although the
resulting signal coverage maps with and without deploy- original signal strengths are quite nonuniform, with a pow-
ment of the RIS system are compared, as shown in Figures er level difference of more than 30 dB (as shown in Figure
6(a)–(c). To enhance the signal coverage, two RIS boards, 6(a)), by enacting RIS-assisted beam forming, a much more
each with 2 × 3 metasurface sheets assembled in a movable uniform signal power distribution can be obtained in the
frame, are deployed along the wall at the other end of the weak coverage area through adaptive coding pattern opti-
corridor, and different allocations of the RIS boards are mization (as shown in Figures 6(b) and (c)). Since higher
studied. As shown in Figure 6(b), when two RIS boards are received signals can lead to better signal-to-noise ratio per-
deployed together at the other side approximately 7.78 m formance and reduce communication failure, the RIS sys-
from the emitter, the signal intensity can be significantly in- tem is considered a promising technology able to improve
creased, with an average enhancement of more than 20 dB the quality of next-generation wireless communication.

(a) (b) (c)

dBm Signal emitter dBm Signal emitter dBm Signal emitter
−40 −40 −40

−70 −70 −70

778 cm

778 cm
778 cm

Metal Metal Metal

−100 door −100 door −100 door

RIS 2

0.8 m

Without RIS RIS 1 & 2 RIS 1

Figure 6 Indoor scenario and signal coverage enhancement with the deployment of a practical RIS system. Wireless signal intensity distributions (a) with-
out and (b) and (c) with the RIS system (denoted by the green bars).

3. Full-space intelligent omni-metasurfaces and transmissive RISs, they may still suffer from the re-
Most of the currently utilized RIS systems only support the striction that they cannot serve user terminals located on
reflection mode, implying that only backward scattering is both sides of metasurfaces. To surmount this obstacle, om-
modulated and access points on the opposite side of RISs ni-metasurfaces with full-space wireless coverage have
are outside of wireless coverage. In addition to assisting in- been proposed to simultaneously serve the terminals on
door wireless communication via reflective RISs, improv- both sides of the RISs [105]. Although there have been a
ing outdoor-to-indoor or adjacent room communication by few attempts to assist wireless communication with omni-
enhancing through-wall signals is similarly crucial. In such metasurfaces [179], [182], they only support dynamic con-
ELECTROMAGNETIC SCIENCE

scenarios, RIS systems operating in transmission mode trol of either the reflection or transmission operation mode,
must provide flexible control of forward scattering, thus ex- which may further limit their scattering manipulation abili-
tending wireless coverage. For example, a passive metalens ty enacted by phase modulation.
has been proposed to focus incoming waves to the other Here, we present an intelligent omni-metasurface
side of a wall with high efficiency, which can reconnect a equipped with dynamic transmission/reflection phase con-
network channel that is otherwise broken due to the weak trol in different modes to further explore the possibilities of
signal strength [191]. Moreover, through-wall imaging us- RISs in wireless communications. As illustrated in Figures
ing a dynamic metasurface antenna can be constructed by 7(a)–(c), the proposed omni-metasurface can be flexibly
forming diverse radiation patterns scattered by the wall [111]. switched among reflection mode, transmission mode, and
Despite the significant progress made for reflective duplex mode (simultaneous reflection and transmission) by
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-13

modulating the loaded tunable components through an ex- ing profiles optimized by genetic algorithms, the proposed
ternal hardware controller, thus customizing it for user ter- approach can overcome coverage holes and enhance local
minals located on the same side, opposite sides and both signals in both backward and forward half-spaces, resulting
sides of the surface. Moreover, by leveraging spatial cod- in full-space wireless coverage.

(a) (b) (c)

Room 1 Room 2 Room 1 Room 2 Room 1 Room 2

User User User User

y y y
User x User User User
Source z Source x z x z
Source
(d) (e) (f)
1.0 600
Transmission amplitude
1.0 600

Transmission phase (°)

Reflection amplitude

Reflection phase (°)

T0
0.8 T1 400 400
T0 0.8
Phase T1 200 200
control layer x y 0.6 0.6
z 0 0
0.4 0.4
−200 R0
R1 −200
Mode control 0.2 −400 0.2 R0
−400
R1
layer 0 −600 0 −600
5.2 5.7 6.2 6.7 7.2 5.2 5.7 6.2 6.7 7.2
Frequency (GHz) Frequency (GHz)
(g) (h) dBm (i) dB
−63.7 16.9
x x x
z z z
Signal
source −77.3 8.5

Omni-metasurface
−91.0 0

Figure 7 Proposed omni-metasurface for assisting full-space wireless communication. (a) Reflection mode; (b) Transmission mode; (c) Duplex mode; (d)
Perspective view of the meta-atom; Simulated (e) transmission and (f) reflection performance of the proposed meta-atom; Measured signal strength (g)
without and (h) with the omni-metasurface; (i) Measured signal strength enhancement.

As illustrated in Figure 7(d), the key procedure to To experimentally verify the full-space wave manipu-
achieve an intelligent omni-metasurface is simplified by lation abilities of the design, a prototype consisting of 20 ×
separately designing two distinct meta-structures capable of 20 meta-atoms with a total size of 408 × 418 mm2 is fabri-
phase modulation and transmission-reflection control and cated, which can be utilized to provide backward or for-
then combining them to form the required meta-atom [171]. ward wireless coverage by setting all the meta-atoms to re-
The phase-control layer is designed with a 180° phase dif- flection or transmission mode. Moreover, it can be utilized
ference for transmission waves by simultaneously switch- to extend the provided wireless coverage into full space by
ing on or off the embedded PIN diodes, and the correspond- adopting a multiplexing technique and setting the meta-
ing operation state is encoded as “0” or “1”. Meanwhile, the atoms to both reflection and transmission modes.
transmission-reflection mode-control layer is designed with As an exemplary demonstration, the omni-metasur-
high-efficiency reflection or high-efficiency transmission face is utilized to assist adjacent room communication by
when the inserted PIN diodes are switched to the “OFF” or enhancing through-wall and reflected signals in a real-
“ON” state, and the corresponding operation state is encod- world scenario with windows, doors, and interior walls. The
ed as “R” or “T”, respectively. Therefore, a 1-bit phase con- omni-metasurface is placed on a wall with a propagation
ELECTROMAGNETIC SCIENCE

figuration in transmission operation mode can be directly loss of approximately −4.5 dB. An Alford-loop-type anten-
acquired by switching diodes of the phase-control layer on na with an omnidirectional radiation pattern for horizontal
or off and setting the mode-control layer to transmission polarization is adopted to mimic a user terminal. Mean-
mode. In contrast, a 1-bit phase configuration in reflection while, a horn antenna is adopted as the signal source and
mode can be implemented by altering diodes of the phase- placed at a distance of 1.16 m from the metasurface. To in-
control layer on or off and setting the mode-control layer to vestigate the ability to reshape wireless signals propagating
reflection mode. The corresponding simulated results for in full space, we evaluate the signals received at 8 locations
the four operation states are shown in Figures 7(e) and (f), in the forward room and 6 locations in the backward room.
from which we can observe a 180° phase difference for Every receiver location in the forward room is separated by
both transmitted and reflected waves at 6.23 GHz. a distance of 1.36 m in the x-direction and 1.65 m in the z-
 0020111-14 Electromagnetic Science, vol.1, no.2

direction, and those in the backward room are separated by wave functionalities such as local signal enhancement and
a distance of 1.2 m in both the x- and z-directions. dynamic beam shaping in reflection or full-space opera-
For each receiver location, we search for the spatial tions. Finally, we would like to mention some challenges in
coding pattern that will maximize the received signal pow- this emerging field and promising research directions based
er. Specifically, based on a genetic algorithm and the real- on our own perspectives.
time measured feedback signal strength, we configure test The concept of RISs has gradually evolved from meta-
states of diodes loaded into the metasurface and monitor the materials and metasurfaces, so they have the advantages of
changes in the received signal in real time. The power of these previous concepts but pose additional challenges for
the signal source is set to 0 dBm. The measured results the design of bias networks and external controllers. At the
without and with the omni-metasurface in the full space are physical level, RIS meta-atoms that have excellent EM per-
plotted in Figures 7(g) and (h), respectively. The received formance with low-cost and low-complexity bias networks
signal at every position can be enhanced. For a clear view, are desired. However, the control range of the phases and
the measured enhancement of the signal strength is dis- their quantization levels contradict the design of the bias
played in Figure 7(i). In the backward room, the reflected network. For example, if varactors are used for RIS design
signal can be enhanced by a maximum of 16.9 dB at posi- to achieve continuously tunable phase responses in the mi-
tion A and an average of 8.7 dB at all positions. In the for- crowave region, then a continuous change in the bias volt-
ward room, the through-wall signal can be improved by up age should be adopted, but this will increase the complexi-
to 9.9 dB at most at position B and 4.7 dB on average at all ty of the external voltage controller because a high-speed
positions. To simultaneously activate the reflection and digital-to-analog converter and a voltage amplifier should
transmission functions, a field or spatial multiplexing be added to each bias line. Suppose that several PIN diodes
method is used in the design, assisted by an optimization al- are adopted to achieve a multibit RIS. In this case, each
gorithm, that is, part of the meta-atoms are switched to re- diode in a meta-atom should be independently addressed by
flection mode while the others are switched to transmission the voltage controller through an independent bias line, in-
mode based on the algorithm calculation. The reflected or creasing the design complexity. These problems will be
transmitted beam width may be narrowed by the RISs, more apparent for the high-frequency region, such as the 5G
which will reduce the signal coverage. To solve this issue, a and 6G millimeter bands, where the physical space in the
dynamic sensing technique (e.g., beam scanning) may be meta-atoms becomes more compacted. In addition, recon-
used to find the users’ locations and then output on-de- figurable reflectarrays and transmitarrays can also perform
mand solutions to enhance the wireless signal intensity. No- tunable wave functionalities in many cases, although with a
tably, signal enhancement is not only relevant to transmit- larger element size comparable to half of the wavelength. In
ters and receivers but also dependent on the propagation en- general, reconfigurable reflectarrays and transmitarrays are
vironment. Since stronger received signals can lead to bet- used for tuning the incidence wave from a feed antenna,
ter signal-to-noise ratio performance and reduce communi- while RISs are mostly used for reshaping the wave in the
cation failure, the proposed omni-metasurface may be able far-field region of the antenna system or that reflected from
to improve both indoor and outdoor-to-indoor communica- the environment. However, the relevant concepts, especial-
tion quality. Further evolution of the omni-metasurface con- ly from the viewpoint of constituent elements, have be-
cept into a self-learning version by utilizing sensors is envi- come vague in recent years, and the idea that the surfaces
sioned. In this way, such surfaces can actively adapt to dif- used in RIS systems should have tunability in shaping the
ferent requirements without human interference. Mean- EM wave propagation environment may become prevalent,
while, more theoretical physics-based models are encour- regardless of their element types as metasurfaces, reflectar-
aged to guide the optimization of RIS-assisted wireless rays, or even frequency-selective surfaces based on the
communication. original definitions. Due to restrictions on paper length, we
were not able to include all the branches of reconfigurable
IV. Summary and Outlooks
artificial structures and materials, for example, reflectar-
In summary, we provide an overview of the RIS concept, rays, transmitarrays, and the active frequency-selective sur-
focusing on the design in the microwave regime. The devel- faces [73]–[79], [192], [193].
ELECTROMAGNETIC SCIENCE

opment of this field originated from metamaterials and High-speed control and response of RIS elements are
metasurfaces. The design methods are similar between RISs necessary for fast modulation of the EM environment and
and reconfigurable metasurfaces at the meta-atom level, wave propagation. Although stand-alone diodes have a very
both involving co-optimization of the resonant structure and fast response time approaching several nanoseconds, the
the actively controllable component. We summarize the upper limit is restricted by the overall circuit loop, includ-
various control methods for reconfigurable metasurfaces, ing the resonant structures, bias lines inside and outside the
including current and voltage bias, mechanical actuation, meta-atoms, and external controllers. Hence, the switching
and thermal driving. As practical demonstrations, we show or control speed will be much slower in experiments. In ad-
our recent works indicating that RISs can dynamically re- dition, realizing high-efficiency RISs comparable to totally
shape the EM propagation environment to achieve versatile passive structures is still a great challenge because tunable
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-15

materials or components will bring additional loss that can- should be fabricated at a low cost, and the external control
not be ignored. Such energy loss would further cause un- signals should be realized with low power consumption and
equal amplitude responses among different operating states low complexity. In addition, RISs may work jointly with
of the meta-atoms. Therefore, new mechanisms involving multiantenna base stations and receiving terminals, increas-
coupling effects between materials and EM waves and new ing the optimization algorithm complexity. Outdoor appli-
materials with low-loss characteristics should be explored cations will even impose additional requirements to with-
to solve the above problems. stand possibly harsh environments, such as wind, rain,
For system-level implementations, most of the report- snow, sunlight, temperature fluctuations, and dust. This al-
ed RIS research is focused on tunable components without so requires that the RIS, control system, and EM properties
power amplification functions. In other words, the output should be stable for months or years. Reshaping of the EM
wave energy is always less than the input. Due to the “mul- propagation environment in a particular communication
tiplicative fading” effect introduced by RISs, achieving no- band for practical use with minimum influence on other
ticeable capacity gains is almost impossible for RISs with- wireless bands is desired. However, RISs have difficulty
out amplification in some cases where the direct link be- discriminating or even cannot discriminate the frequency
tween the emitter and receiver is not weak, but recent stud- symbols of signals with high accuracy because RISs do not
ies have shown that active RISs with energy amplification have digital or RF chains to process incident signals.
can possibly overcome this effect [194]. The nonreciprocal RISs capable of dynamically reshaping EM waves are
and power-sensitive properties of chip amplifiers may influ- the physical basis for further use in wireless communica-
ence the modulation signals and information transmission, tion, while studies beyond the physical layer are a key step
which should be seriously considered for practical applica- to make RISs truly applicable. This review mainly focuses
tions [195]–[197]. on the design methods and practical demonstrations of EM
Most of the existing metasurfaces are designed for re- wave enhancement. However, from the viewpoint of wire-
flection operations, so they may have the disadvantages of a less communication theory, there are still many challenges
high profile and a blockage effect from the feeding/receiv- and problems, such as energy-efficient communication
ing antennas. Moreover, they cannot process transmitted channel sensing and estimation, practical protocols for in-
waves when transmission manipulation is necessary, for ex- formation exchange, and real-time allocation and optimiza-
ample, in an indoor environment with several subrooms. tion of different RISs to serve multiple data streams in dy-
Although recent studies have proposed several ways to im- namic and heterogeneous networks. More concrete discus-
plement omni-metasurfaces to tune EM waves in both re- sions from the viewpoint of wireless communication theory
flection and transmission modes, there are still some prob- can be found in some recent reviews [90], [199], [200].
lems that could be further resolved, such as the efficiency, Overall, the two communities need more synergic research
bandwidth, and independently addressable properties. and dialog to push this work further for real-world applica-
The heuristic use of AI-empowered RIS applications tions. Nevertheless, the rapid development of this field may
has combined computation-enabled AI and physical meta- somehow overcome these issues in the near future, making
structures, accelerating RIS development toward smart plat- RIS technologies truly intelligent platforms.
forms. Most AI-empowered RISs are developed for one or
Acknowledgements
several kinds of problems based on a vast amount of train-
ing data, so how to extend the generalization of AI-empow- This work was supported by the National Natural Science
ered RISs so that they can adapt well to unexpected tasks Foundation of China (Grant Nos. 62271243, 62071215,
may be further explored. In addition, the metasurface and 91963128, and 61731010), the National Key Research and
the training data are separated at the physical level in most Development Program of China (Grant No. 2017YFA07
current strategies to execute deep-learning methods. Data 00201), and the Joint Fund of Ministry of Education for
training and learning directly executed by the metasurface Equipment Pre-research (Grant No. 8091B032112). We
itself may maximally alleviate the human intervention and would like to acknowledge the support from the Priority
accelerate the response time. Academic Program Development of Jiangsu Higher Educa-
RISs provide a new paradigm in wireless communica- tion Institutions (PAPD), Fundamental Research Funds for
ELECTROMAGNETIC SCIENCE

tions, but the realization of RISs is still in its early stages the Central Universities, and Jiangsu Provincial Key Labo-
and far from widespread use in the real world [198]. Stan- ratory of Advanced Manipulating Technique of Electro-
dardization work should be done at a regional level for in- magnetic Wave.
dustrial and commercial services. Although extensive re-
search and activities are dedicated to RISs globally, stan-
References
dardization work is still in progress. RISs should be large [1] R. A. Shelby, D. R. Smith, and S. Schultz, “Experimental verifica-
enough to capture incident waves and thus offer a competi- tion of a negative index of refraction,” Science, vol. 292, no. 5514,
pp. 77–79, 2001.
tive gain, especially for outdoor environments. Therefore, [2] D. Schurig, J. J. Mock, B. J. Justice, et al., “Metamaterial electro-
to enable flexible deployment of an RIS, which may con- magnetic cloak at microwave frequencies,” Science, vol. 314, no.
tain hundreds or thousands of elements, the RIS elements 5801, pp. 977–980, 2006.
 0020111-16 Electromagnetic Science, vol.1, no.2

[3] J. B. Pendry, “Negative refraction makes a perfect lens,” Physical high-gain reflector antenna,” IEEE Transactions on Antennas and
Review Letters, vol. 85, no. 18, pp. 3966–3969, 2000. Propagation, vol. 68, no. 3, pp. 1426–1435, 2020.
[4] N. I. Landy, S. Sajuyigbe, J. J. Mock, et al., “Perfect metamaterial [25] S. Liu, T. J. Cui, Q. Xu, et al., “Anisotropic coding metamaterials
absorber,” Physical Review Letters, vol. 100, no. 20, article no. and their powerful manipulation of differently polarized terahertz
207402, 2008. waves,” Light: Science & Applications, vol. 5, no. 5, article no.
[5] K. Chen, G. W. Ding, G. W. Hu, et al., “Directional Janus meta- e16076, 2016.
surface,” Advanced Materials, vol. 32, no. 2, article no. 1906352, [26] L. Cui, W. J. Wang, G. W. Ding, et al., “Polarization-dependent
2020. bi-functional metasurface for directive radiation and diffusion-like
[6] Q. He, S. L. Sun, S. Y. Xiao, et al., “High-efficiency metasurfaces: scattering,” AIP Advances, vol. 7, no. 11, article no. 115214, 2017.
Principles, realizations, and applications,” Advanced Optical Mate- [27] M. Khorasaninejad, W. T. Chen, A. Y. Zhu, et al., “Multispectral
rials, vol. 6, no. 19, article no. 1800415, 2018. chiral imaging with a metalens,” Nano Letters, vol. 16, no. 7, pp.
[7] O. Quevedo-Teruel, H. S. Chen, A. Díaz-Rubio, et al., “Roadmap 4595–4600, 2016.
on metasurfaces,” Journal of Optics, vol. 21, no. 7, article no. [28] B. Sima, K. Chen, X. Y. Luo, et al., “Combining frequency-selec-
073002, 2019. tive scattering and specular reflection through phase-dispersion tai-
[8] K. Chen and Y. J. Feng, “A review of recent progress on direction- loring of a metasurface,” Physical Review Applied, vol. 10, no. 6,
al metasurfaces: Concept, design, and application,” Journal of article no. 064043, 2018.
Physics D: Applied Physics, vol. 55, no. 38, article no. 383001, [29] X. Y. Luo, W. L. Guo, K. Qu, et al., “Quad-channel independent
2022. wavefront encoding with dual-band multitasking metasurface,”
[9] N. F. Yu and F. Capasso, “Flat optics with designer metasurfaces,” Optics Express, vol. 29, no. 10, pp. 15678–15688, 2021.
Nature Materials, vol. 13, no. 2, pp. 139–150, 2014. [30] C. W. Wan, C. J. Dai, S. Wan, et al., “Dual-encryption freedom
[10] S. B. Glybovski, S. A. Tretyakov, P. A. Belov, et al., “Metasur- via a monolayer-nanotextured Janus metasurface in the broadband
faces: From microwaves to visible,” Physics Reports, vol. 634, pp. visible,” Optics Express, vol. 29, no. 21, pp. 33954–33961, 2021.
1–72, 2016. [31] H. X. Xu, C. H. Wang, G. W. Hu, et al., “Spin-encoded wave-
[11] T. J. Cui, S. Liu, and L. Zhang, “Information metamaterials and length-direction multitasking Janus metasurfaces,” Advanced Opti-
metasurfaces,” Journal of Materials Chemistry C, vol. 5, no. 15, cal Materials, vol. 9, no. 11, article no. 2100190, 2021.
pp. 3644–3668, 2017. [32] T. J. Cui, M. Q. Qi, X. Wan, et al., “Coding metamaterials, digital
[12] N. F. Yu, P. Genevet, M. A. Kats, et al., “Light propagation with metamaterials and programmable metamaterials,” Light: Science
phase discontinuities: Generalized laws of reflection and refrac- & Applications, vol. 3, no. 10, article no. e218, 2014.
tion,” Science, vol. 334, no. 6054, pp. 333–337, 2011. [33] Z. Q. Miao, Q. Wu, X. Li, et al., “Widely tunable terahertz phase
[13] S. L. Sun, Q. He, S. Y. Xiao, et al., “Gradient-index meta-surfaces modulation with gate-controlled graphene metasurfaces,” Physical
as a bridge linking propagating waves and surface waves,” Nature Review X, vol. 5, no. 4, article no. 041027, 2015.
Materials, vol. 11, no. 5, pp. 426–431, 2012. [34] S. Lee, W. T. Kim, J. H. Kang, et al., “Single-layer metasurfaces
[14] E. Arbabi, S. M. Kamali, A. Arbabi, et al., “Full-stokes imaging as spectrally tunable terahertz half- and quarter-waveplates,” ACS
polarimetry using dielectric metasurfaces,” ACS Photonics, vol. 5, Applied Materials & Interfaces, vol. 11, no. 8, pp. 7655–7660,
no. 8, pp. 3132–3140, 2018. 2019.
[15] X. J. Ni, A. V. Kildishev, and V. M. Shalaev, “Metasurface holo- [35] J. Y. Guo, T. Wang, H. Zhao, et al., “Reconfigurable terahertz
grams for visible light,” Nature Communications, vol. 4, no. 1, ar- metasurface pure phase holograms,” Advanced Optical Materials,
ticle no. 2807, 2013. vol. 7, no. 10, article no. 1801696, 2019.
[16] L. L. Huang, X. Z. Chen, H. Mühlenbernd, et al., “Three-dimen- [36] M. Y. Shalaginov, S. S. An, Y. F. Zhang, et al., “Reconfigurable
sional optical holography using a plasmonic metasurface,” Nature all-dielectric metalens with diffraction-limited performance,” Na-
Communications, vol. 4, no. 1, article no. 2808, 2013. ture Communications, vol. 12, no. 1, article no. 1225, 2021.
[17] X. J. Ni, Z. J. Wong, M. Mrejen, et al., “An ultrathin invisibility [37] G. Y. Qu, W. H. Yang, Q. H. Song, et al., “Reprogrammable meta-
skin cloak for visible light,” Science, vol. 349, no. 6254, pp. hologram for optical encryption,” Nature Communications, vol.
1310–1314, 2015. 11, no. 1, article no. 5484, 2020.
[18] X. Wan, X. P. Shen, Y. Luo, et al., “Planar bifunctional Luneburg- [38] C. Huang, B. Sun, W. B. Pan, et al., “Dynamical beam manipula-
fisheye lens made of an anisotropic metasurface,” Laser & Pho- tion based on 2-bit digitally-controlled coding metasurface,” Sci-
tonics Reviews, vol. 8, no. 5, pp. 757–765, 2014. entific Reports, vol. 7, article no. 42302, 2017.
[19] F. Aieta, P. Genevet, M. A. Kats, et al., “Aberration-free ultrathin [39] X. Y. Luo, W. L. Guo, K. Chen, et al., “Active cylindrical meta-
flat lenses and axicons at telecom wavelengths based on plasmon- surface with spatial reconfigurability for tunable backward scatter-
ic metasurfaces,” Nano Letters, vol. 12, no. 9, pp. 4932–4936, ing reduction,” IEEE Transactions on Antennas and Propagation,
2012. vol. 69, no. 6, pp. 3332–3340, 2021.
[20] W. T. Chen, A. Y. Zhu, V. Sanjeev, et al., “A broadband achro- [40] N. Zhang, K. Chen, J. M. Zhao, et al., “A dual-polarized reconfig-
matic metalens for focusing and imaging in the visible,” Nature urable reflectarray antenna based on dual-channel programmable
Nanotechnology, vol. 13, no. 3, pp. 220–226, 2018. metasurface,” IEEE Transactions on Antennas and Propagation,
[21] M. Khorasaninejad, W. T. Chen, R. C. Devlin, et al., “Metalenses vol. 70, no. 9, pp. 7403–7412, 2022.
at visible wavelengths: Diffraction-limited focusing and subwave- [41] L. L. Li, T. Jun Cui, W. Ji, et al., “Electromagnetic repro-
ELECTROMAGNETIC SCIENCE

length resolution imaging,” Science, vol. 352, no. 6290, pp. grammable coding-metasurface holograms,” Nature Communica-
1190–1194, 2016. tions, vol. 8, no. 1, article no. 197, 2017.
[22] K. Zhang, Y. Y. Yuan, X. M. Ding, et al., “Polarization-engi- [42] M. F. Imani, T. Sleasman, and D. R. Smith, “Two-dimensional dy-
neered noninterleaved metasurface for integer and fractional or- namic metasurface apertures for computational microwave imag-
bital angular momentum multiplexing,” Laser & Photonics Re- ing,” IEEE Antennas and Wireless Propagation Letters, vol. 17,
views, vol. 15, no. 1, article no. 2000351, 2021. no. 12, pp. 2299–2303, 2018.
[23] G. W. Ding, K. Chen, X. Y. Luo, et al., “Dual-helicity decoupled [43] J. Zhang, X. Z. Wei, I. D. Rukhlenko, et al., “Electrically tunable
coding metasurface for independent spin-to-orbital angular mo- metasurface with independent frequency and amplitude modula-
mentum conversion,” Physical Review Applied, vol. 11, no. 4, arti- tions,” ACS Photonics, vol. 7, no. 1, pp. 265–271, 2020.
cle no. 044043, 2019. [44] H. Chen, Z. G. Liu, W. B. Lu, et al., “Full polarization transforma-
[24] W. L. Guo, K. Chen, G. M. Wang, et al., “Transmission–reflec- tion using graphene in microwave frequencies,” IEEE Transac-
tion-selective metasurface and its application to RCS reduction of tions on Antennas and Propagation, vol. 68, no. 5, pp. 3760–3769,
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-17

2020. nas and Propagation, vol. 69, no. 2, pp. 1188–1192, 2021.
[45] O. Buchnev, N. Podoliak, M. Kaczmarek, et al., “Electrically con- [66] L. Boccia, I. Russo, G. Amendola, et al., “Multilayer antenna-fil-
trolled nanostructured metasurface loaded with liquid crystal: To- ter antenna for beam-steering transmit-array applications,” IEEE
ward multifunctional photonic switch,” Advanced Optical Materi- Transactions on Microwave Theory and Techniques, vol. 60, no. 7,
als, vol. 3, no. 5, pp. 674–679, 2015. pp. 2287–2300, 2012.
[46] A. Komar, R. Paniagua-Domínguez, A. Miroshnichenko, et al., [67] J. Y. Lau and S. V. Hum, “A planar reconfigurable aperture with
“Dynamic beam switching by liquid crystal tunable dielectric lens and reflectarray modes of operation,” IEEE Transactions on
metasurfaces,” ACS Photonics, vol. 5, no. 5, pp. 1742–1748, 2018. Microwave Theory and Techniques, vol. 58, no. 12, pp.
[47] M. Decker, C. Kremers, A. Minovich, et al., “Electro-optical 3547–3555, 2010.
switching by liquid-crystal controlled metasurfaces,” Optics Ex- [68] Q. Hu, K. Chen, N. Zhang, et al., “Arbitrary and dynamic poincaré
press, vol. 21, no. 7, pp. 8879–8885, 2013. sphere polarization converter with a time-varying metasurface,”
[48] J. B. Wu, Z. Shen, S. J. Ge, et al., “Liquid crystal programmable Advanced Optical Materials, vol. 10, no. 4, article no. 2101915,
metasurface for terahertz beam steering,” Applied Physics Letters, 2022.
vol. 116, no. 13, article no. 131104, 2020. [69] C. Huang, C. L. Zhang, J. N. Yang, et al., “Reconfigurable meta-
[49] B. W. Chen, X. R. Wang, W. L. Li, et al., “Electrically address- surface for multifunctional control of electromagnetic waves,” Ad-
able integrated intelligent terahertz metasurface,” Science Ad- vanced Optical Materials, vol. 5, no. 22, article no. 1700485,
vances, vol. 8, no. 41, article no. eadd1296, 2022. 2017.
[50] E. Takou, A. C. Tasolamprou, O. Tsilipakos, et al., “Anapole tol- [70] Z. N. Wu, Y. Ra’di, and A. Grbic, “Tunable metasurfaces: A po-
erance to dissipation losses in thermally tunable water-based meta- larization rotator design,” Physical Review X, vol. 9, no. 1, article
surfaces,” Physical Review Applied, vol. 15, no. 1, article no. no. 011036, 2019.
014043, 2021. [71] W. Li, S. Xia, B. He, et al., “A reconfigurable polarization con-
[51] A. E. Cardin, S. R. Silva, S. R. Vardeny, et al., “Surface-wave-as- verter using active metasurface and its application in horn
sisted nonreciprocity in spatio-temporally modulated antenna,” IEEE Transactions on Antennas and Propagation, vol.
metasurfaces,” Nature Communications, vol. 11, no. 1, article no. 64, no. 12, pp. 5281–5290, 2016.
1469, 2020. [72] Q. Hu, K. Chen, J. M. Zhao, et al., “On-demand dynamic polariza-
[52] W. X. Yang, K. Chen, Y. L. Zheng, et al., “Angular-adaptive re- tion meta-transformer,” Laser & Photonics Reviews, vol. 17, no. 1,
configurable spin-locked metasurface retroreflector,” Advanced article no. 2200479, 2023.
Science, vol. 8, no. 21, article no. 2100885, 2021. [73] M. Riel and J. J. Laurin, “Design of an electronically beam scan-
[53] Y. L. Zheng, K. Chen, W. X. Yang, et al., “Kirigami reconfig- ning reflectarray using aperture-coupled elements,” IEEE Transac-
urable gradient metasurface,” Advanced Functional Materials, vol. tions on Antennas and Propagation, vol. 55, no. 5, pp. 1260–1266,
32, no. 5, article no. 2107699, 2022. 2007.
[54] E. Arbabi, A. Arbabi, S. M. Kamali, et al., “MEMS-tunable dielec- [74] A. Clemente, L. Dussopt, R. Sauleau, et al., “Wideband 400-ele-
tric metasurface lens,” Nature Communications, vol. 9, no. 1, arti- ment electronically reconfigurable transmitarray in X band,” IEEE
cle no. 812, 2018. Transactions on Antennas and Propagation, vol. 61, no. 10, pp.
[55] I. V. Shadrivov, P. V. Kapitano.a, S. I. Maslovski, et al., “Metama- 5017–5027, 2013.
terials controlled with light,” Physical Review Letters, vol. 109, [75] J. Q. Han, L. Li, G. Y. Liu, et al., “A wideband 1 bit 12 × 12 re-
no. 8, article no. 083902, 2012. configurable beam-scanning reflectarray: Design, fabrication, and
[56] X. G. Zhang, W. X. Jiang, H. L. Jiang, et al., “An optically driven measurement,” IEEE Antennas and Wireless Propagation Letters,
digital metasurface for programming electromagnetic functions,” vol. 18, no. 6, pp. 1268–1272, 2019.
Nature Electronics, vol. 3, no. 3, pp. 165–171, 2020. [76] L. Z. Song, P. Y. Qin, H. Zhu, et al., “Wideband conformal trans-
[57] L. Chen, Q. F. Nie, Y. Ruan, et al., “Light-controllable metasur- mitarrays for E-band multi-beam applications,” IEEE Transac-
face for microwave wavefront manipulation,” Optics Express, vol. tions on Antennas and Propagation, vol. 70, no. 11, pp.
28, no. 13, pp. 18742–18749, 2020. 10417–10425, 2022.
[58] X. Y. Duan, S. Kamin, and N. Liu, “Dynamic plasmonic colour [77] X. Wang, P. Y. Qin, A. Tuyen Le, et al., “Beam scanning transmi-
display,” Nature Communications, vol. 8, article no. 14606, 2017. tarray employing reconfigurable dual-layer Huygens element,”
[59] L. L. Li, H. X. Ruan, C. Liu, et al., “Machine-learning repro- IEEE Transactions on Antennas and Propagation, vol. 70, no. 9,
grammable metasurface imager,” Nature Communications, vol. 10, pp. 7491–7500, 2022.
no. 1, article no. 1082, 2019. [78] L. Dussopt, A. Moknache, J. Saily, et al., “A V-band switched-
[60] H. H. Yang, X. Y. Cao, F. Yang, et al., “A programmable metasur- beam linearly polarized transmit-array antenna for wireless back-
face with dynamic polarization, scattering and focusing control,” haul applications,” IEEE Transactions on Antennas and Propaga-
Scientific Reports, vol. 6, article no. 35692, 2016. tion, vol. 65, no. 12, pp. 6788–6793, 2017.
[61] K. Chen, N. Zhang, G. W. Ding, et al., “Active anisotropic coding [79] M. Yi, Y. Bae, S. Yoo, et al., “Digitized reconfigurable metal Re-
metasurface with independent real-time reconfigurability for dual flectarray surfaces for millimeter-wave beam-engineering,” Ap-
polarized waves,” Advanced Materials Technologies, vol. 5, no. 2, plied Sciences, vol. 11, no. 13, article no. 5811, 2021.
article no. 1900930, 2020. [80] O. Tsilipakos, A. C. Tasolamprou, A. Pitilakis, et al., “Toward in-
[62] L. Zhang, X. Q. Chen, S. Liu, et al., “Space-time-coding digital telligent metasurfaces: The progress from globally tunable meta-
ELECTROMAGNETIC SCIENCE

metasurfaces,” Nature Communications, vol. 9, no. 1, article no. surfaces to software-defined metasurfaces with an embedded net-
4334, 2018. work of controllers,” Advanced Optical Materials, vol. 8, no. 17,
[63] N. Zhang, K. Chen, Q. Hu, et al., “Spatiotemporal metasurface to article no. 2000783, 2020.
control electromagnetic wave scattering,” Physical Review Ap- [81] H. P. Wang, Y. B. Li, H. Li, et al., “Intelligent metasurface with
plied, vol. 17, no. 5, article no. 054001, 2022. frequency recognition for adaptive manipulation of electromagnet-
[64] X. D. Bai, F. W. Kong, Y. T. Sun, et al., “High-efficiency trans- ic wave,” Nanophotonics, vol. 11, no. 7, pp. 1401–1411, 2022.
missive programmable metasurface for multimode OAM genera- [82] J. B. Gros, V. Popov, M. A. Odit, et al., “A reconfigurable intelli-
tion,” Advanced Optical Materials, vol. 8, no. 17, article no. gent surface at mmWave based on a binary phase tunable metasur-
2000570, 2020. face,” IEEE Open Journal of the Communications Society, vol. 2,
[65] Y. Wang, S. H. Xu, F. Yang, et al., “1 bit dual-linear polarized re- pp. 1055–1064, 2021.
configurable Transmitarray antenna using asymmetric dipole ele- [83] H. H. Yang, F. Yang, X. Y. Cao, et al., “A 1600-element dual-fre-
ments with parasitic bypass dipoles,” IEEE Transactions on Anten- quency electronically reconfigurable reflectarray at X/Ku-band,”
 0020111-18 Electromagnetic Science, vol.1, no.2

IEEE Transactions on Antennas and Propagation, vol. 65, no. 6, [101] L. Yang, J. X. Yang, W. W. Xie, et al., “Secrecy performance
pp. 3024–3032, 2017. analysis of RIS-aided wireless communication systems,” IEEE
[84] H. Lin, W. Yu, R. X. Tang, et al., “A dual-band reconfigurable in- Transactions on Vehicular Technology, vol. 69, no. 10, pp.
telligent metasurface with beam steering,” Journal of Physics D: 12296–12300, 2020.
Applied Physics, vol. 55, no. 24, article no. 245002, 2022. [102] H. L. Zhang, B. Y. Di, L. Y. Song, et al., “Reconfigurable intelli-
[85] Q. Ma, G. D. Bai, H. B. Jing, et al., “Smart metasurface with self- gent surfaces assisted communications with limited phase shifts:
adaptively reprogrammable functions,” Light: Science & Applica- How many phase shifts are enough?,” IEEE Transactions on Ve-
tions, vol. 8, article no. 98, 2019. hicular Technology, vol. 69, no. 4, pp. 4498–4502, 2020.
[86] J. Y. Liu, Y. P. Duan, T. Zhang, et al., “Dual-polarized and real- [103] C. W. Huang, A. Zappone, G. C. Alexandropoulos, et al., “Recon-
time reconfigurable metasurface absorber with infrared-coded re- figurable intelligent surfaces for energy efficiency in wireless
mote-control system,” Nano Research, vol. 15, no. 8, pp. communication,” IEEE Transactions on Wireless Communica-
7498–7505, 2022. tions, vol. 18, no. 8, pp. 4157–4170, 2019.
[87] E. Basar, M. D. Renzo, J. De Rosny, et al., “Wireless communica- [104] Q. Q. Wu and R. Zhang, “Intelligent reflecting surface enhanced
tions through reconfigurable intelligent surfaces,” IEEE Access, wireless network via joint active and passive beamforming,” IEEE
vol. 7, pp. 116753–116773, 2019. Transactions on Wireless Communications, vol. 18, no. 11, pp.
[88] Q. Q. Wu and R. Zhang, “Towards smart and reconfigurable envi- 5394–5409, 2019.
ronment: Intelligent reflecting surface aided wireless network,” [105] S. H. Zhang, H. L. Zhang, B. Y. Di, et al., “Intelligent Omni-sur-
IEEE Communications Magazine, vol. 58, no. 1, pp. 106–112, faces: Ubiquitous wireless transmission by reflective-refractive
2020. metasurfaces,” IEEE Transactions on Wireless Communications,
[89] J. Z. Hu, H. L. Zhang, B. Y. Di, et al., “Reconfigurable intelligent vol. 21, no. 1, pp. 219–233, 2022.
surface based RF sensing: Design, optimization, and implementa- [106] J. P. Qiao and M. S. Alouini, “Secure transmission for intelligent
tion,” IEEE Journal on Selected Areas in Communications, vol. reflecting surface-assisted mmWave and terahertz systems,” IEEE
38, no. 11, pp. 2700–2716, 2020. Wireless Communications Letters, vol. 9, no. 10, pp. 1743–1747,
[90] M. Di Renzo, A. Zappone, M. Debbah, et al., “Smart radio envi- 2020.
ronments empowered by reconfigurable intelligent surfaces: How [107] E. Basar, “Reconfigurable intelligent surfaces for Doppler effect
it works, state of research, and the road ahead,” IEEE Journal on and multipath fading mitigation,” Frontiers in Communications
Selected Areas in Communications, vol. 38, no. 11, pp. and Networks, vol. 2, article no. 672857, 2021.
2450–2525, 2020. [108] F. Liu, O. Tsilipakos, A. Pitilakis, et al., “Intelligent metasurfaces
[91] M. A. ElMossallamy, H. L. Zhang, L. Y. Song, et al., “Reconfig- with continuously tunable local surface impedance for multiple re-
urable intelligent surfaces for wireless communications: Principles, configurable functions,” Physical Review Applied, vol. 11, no. 4,
challenges, and opportunities,” IEEE Transactions on Cognitive article no. 044024, 2019.
Communications and Networking, vol. 6, no. 3, pp. 990–1002, [109] P. del Hougne, M. Fink, and G. Lerosey, “Optimally diverse com-
2020. munication channels in disordered environments with tuned ran-
[92] J. Y. Dai, W. K. Tang, M. Z. Chen, et al., “Wireless communica- domness,” Nature Electronics, vol. 2, no. 1, pp. 36–41, 2019.
tion based on information metasurfaces,” IEEE Transactions on [110] G. C. Trichopoulos, P. Theofanopoulos, B. Kashyap, et al., “De-
Microwave Theory and Techniques, vol. 69, no. 3, pp. 1493–1510, sign and evaluation of reconfigurable intelligent surfaces in real-
2021. world environment,” IEEE Open Journal of the Communications
[93] C. Liaskos, S. Nie, A. Tsioliaridou, et al., “A new wireless com- Society, vol. 3, pp. 462–474, 2022.
munication paradigm through software-controlled metasurfaces,” [111] T. Sleasman, M. F. Imani, M. Boyarsky, et al., “Computational
IEEE Communications Magazine, vol. 56, no. 9, pp. 162–169, through-wall imaging using a dynamic metasurface antenna,” OSA
2018. Continuum, vol. 2, no. 12, pp. 3499–3513, 2019.
[94] S. W. Zhang and R. Zhang, “Capacity characterization for intelli- [112] Y. L. Zheng, K. Chen, Z. Y. Xu, et al., “Metasurface-assisted
gent reflecting surface aided MIMO communication,” IEEE Jour- wireless communication with physical level information encryp-
nal on Selected Areas in Communications, vol. 38, no. 8, pp. tion,” Advanced Science, vol. 9, no. 34, article no. 2204558, 2022.
1823–1838, 2020. [113] W. K. Tang, J. Y. Dai, M. Z. Chen, et al., “MIMO transmission
[95] X. D. Mu, Y. W. Liu, L. Guo, et al., “Capacity and optimal re- through reconfigurable intelligent surface: System design, analy-
source allocation for IRS-assisted multi-user communication sys- sis, and implementation,” IEEE Journal on Selected Areas in Com-
tems,” IEEE Transactions on Communications, vol. 69, no. 6, pp. munications, vol. 38, no. 11, pp. 2683–2699, 2020.
3771–3786, 2021. [114] J. Y. Dai, W. K. Tang, L. X. Yang, et al., “Realization of multi-
[96] G. P. Wang, F. F. Gao, R. F. Fan, et al., “Ambient backscatter modulation schemes for wireless communication by time-domain
communication systems: Detection and performance analysis,” digital coding metasurface,” IEEE Transactions on Antennas and
IEEE Transactions on Communications, vol. 64, no. 11, pp. Propagation, vol. 68, no. 3, pp. 1618–1627, 2020.
4836–4846, 2016. [115] J. Zhao, X. Yang, J. Y. Dai, et al., “Programmable time-domain
[97] L. Wei, C. W. Huang, G. C. Alexandropoulos, et al., “Channel es- digital-coding metasurface for non-linear harmonic manipulation
timation for RIS-empowered multi-user MISO wireless communi- and new wireless communication systems,” National Science Re-
cations,” IEEE Transactions on Communications, vol. 69, no. 6, view, vol. 6, no. 2, pp. 231–238, 2019.
ELECTROMAGNETIC SCIENCE

pp. 4144–4157, 2021. [116] M. Z. Chen, W. K. Tang, J. Y. Dai, et al., “Accurate and broad-
[98] S. Hu, F. Rusek, and O. Edfors, “Beyond massive MIMO: The po- band manipulations of harmonic amplitudes and phases to reach
tential of positioning with large intelligent surfaces,” IEEE Trans- 256 QAM millimeter-wave wireless communications by time-do-
actions on Signal Processing, vol. 66, no. 7, pp. 1761–1774, 2018. main digital coding metasurface,” National Science Review, vol. 9,
[99] S. Hu, F. Rusek, and O. Edfors, “Beyond massive MIMO: The po- no. 1, article no. nwab134, 2022.
tential of data transmission with large intelligent surfaces,” IEEE [117] E. Björnson, Ö. Özdogan, and E. G. Larsson, “Reconfigurable in-
Transactions on Signal Processing, vol. 66, no. 10, pp. telligent surfaces: Three myths and two critical questions,” IEEE
2746–2758, 2018. Communications Magazine, vol. 58, no. 12, pp. 90–96, 2020.
[100] Y. Han, W. K. Tang, S. Jin, et al., “Large intelligent surface-assist- [118] Y. W. Liu, X. Liu, X. D. Mu, et al., “Reconfigurable intelligent
ed wireless communication exploiting statistical CSI,” IEEE surfaces: Principles and opportunities,” IEEE Communications
Transactions on Vehicular Technology, vol. 68, no. 8, pp. Surveys & Tutorials, vol. 23, no. 3, pp. 1546–1577, 2021.
8238–8242, 2019. [119] S. Basharat, M. Khan, M. Iqbal, et al., “Exploring reconfigurable
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-19

intelligent surfaces for 6G: State-of-the-art and the road ahead,” perimental demonstration of ultrafast THz modulation in a
IET Communications, vol. 16, no. 13, pp. 1458–1474, 2022. graphene-based thin film absorber through negative photoinduced
[120] Q. Cheng, L. Zhang, J. Y. Dai, et al., “Reconfigurable intelligent conductivity,” ACS Photonics, vol. 6, no. 3, pp. 720–727, 2019.
surfaces: Simplified-architecture transmitters—From theory to im- [140] P. J. Guo, R. D. Schaller, J. B. Ketterson, et al., “Ultrafast switch-
plementations,” Proceedings of the IEEE, vol. 110, no. 9, pp. ing of tunable infrared plasmons in indium tin oxide nanorod ar-
1266–1289, 2022. rays with large absolute amplitude,” Nature Photonics, vol. 10, no.
[121] L. L. Li, H. T. Zhao, C. Liu, et al., “Intelligent metasurfaces: Con- 4, pp. 267–273, 2016.
trol, communication and computing,” eLight, vol. 2, no. 1, article [141] M. Taghinejad, H. Taghinejad, Z. H. Xu, et al., “Ultrafast control
no. 7, 2022. of phase and polarization of light expedited by hot-electron trans-
[122] Q. He, S. L. Sun, and L. Zhou, “Tunable/reconfigurable metasur- fer,” Nano Letters, vol. 18, no. 9, pp. 5544–5551, 2018.
faces: Physics and applications,” Research, vol. 2019, article no. [142] M. K. Liu, H. Y. Hwang, H. Tao, et al., “Terahertz-field-induced
1849272, 2019. insulator-to-metal transition in vanadium dioxide metamaterial,”
[123] Y. Saifullah, Y. J. He, A. Boag, et al., “Recent progress in recon- Nature, vol. 487, no. 7407, pp. 345–348, 2012.
figurable and intelligent metasurfaces: A comprehensive review of [143] A. Leitis, A. Heßler, S. Wahl, et al., “All-dielectric programmable
tuning mechanisms, hardware designs, and applications,” Ad- Huygens’ metasurfaces,” Advanced Functional Materials, vol. 30,
vanced Science, vol. 9, no. 33, article no. 2203747, 2022. no. 19, article no. 1910259, 2020.
[124] H. T. Chen, W. J. Padilla, M. J. Cich, et al., “A metamaterial solid- [144] P. Pitchapp., A. Kumar, S. Prakash, et al., “Chalcogenide phase
state terahertz phase modulator,” Nature Photonics, vol. 3, no. 3, change material for active terahertz photonics,” Advanced Materi-
pp. 148–151, 2009. als, vol. 31, no. 12, article no. 1808157, 2019.
[125] J. Zhang, H. Zhang, W. X. Yang, et al., “Dynamic scattering steer- [145] X. G. Zhang, Y. L. Sun, B. C. Zhu, et al., “A metasurface-based
ing with grapheme-based coding metamirror,” Advanced Optical light-to-microwave transmitter for hybrid wireless communica-
Materials, vol. 8, no. 19, article no. 2000683, 2020. tions,” Light:Science & Applications, vol. 11, no. 1, article no.
[126] G. K. Shirmanesh, R. Sokhoyan, P. C. Wu, et al., “Electro-optical- 126, 2022.
ly tunable multifunctional metasurfaces,” ACS Nano, vol. 14, no. [146] X. H. Yin, T. Steinle, L. L. Huang, et al., “Beam switching and bi-
6, pp. 6912–6920, 2020. focal zoom lensing using active plasmonic metasurfaces,” Science
[127] Q. Wang, X. G. Zhang, H. W. Tian, et al., “Millimeter-wave digi- & Applications, vol. 6, no. 7, article no. e17016, 2017.
tal coding metasurfaces based on nematic liquid crystals,” Ad- [147] K. Sun, C. A. Riedel, A. Urbani, et al., “VO2 thermochromic meta-
vanced Theory and Simulations, vol. 2, no. 12, article no. 1900141, material-based smart optical solar reflector ,” ACS Photonics, vol.
2019. 5, no. 6, pp. 2280–2286, 2018.
[128] N. Zhang, K. Chen, Y. L. Zheng, et al., “Programmable coding [148] T. Driscoll, S. Palit, M. M. Qazilbash, et al., “Dynamic tuning of
metasurface for dual-band independent real-time beam control,” an infrared hybrid-metamaterial resonance using vanadium diox-
IEEE Journal on Emerging and Selected Topics in Circuits and ide,” Applied Physics Letters, vol. 93, no. 2, article no. 024101,
Systems, vol. 10, no. 1, pp. 20–28, 2020. 2008.
[129] H. L. Wang, H. F. Ma, M. Chen, et al., “A reconfigurable multi- [149] H. T. Chen, H. Yang, R. Singh, et al., “Tuning the resonance in
functional metasurface for full-space control of electromagnetic high-temperature superconducting terahertz metamaterials,” Physi-
waves,” Advanced Functional Materials, vol. 31, no. 25, article cal Review Letters, vol. 105, no. 24, article no. 247402, 2010.
no. 2100275, 2021. [150] M. Rahmani, L. Xu, A. E. Miroshnichenko, et al., “Reversible
[130] K. Chen, Y. J. Feng, F. Monticone, et al., “A reconfigurable ac- thermal tuning of all-dielectric metasurfaces,” Advanced Function-
tive Huygens’ metalens,” Advanced Materials, vol. 29, no. 17, arti- al Materials, vol. 27, no. 31, article no. 1700580, 2017.
cle no. 1606422, 2017. [151] H. B. Wu, Q. L. Luo, H. J. Chen, et al., “Magnetically control-
[131] J. C. Liang, Q. Cheng, Y. Gao, et al., “An angle-insensitive 3-bit lable nonreciprocal Goos-Hänchen shiftsupported by a magnetic
reconfigurable intelligent surface,” IEEE Transactions on Anten- plasmonic gradient metasurface,” Physical Review A, vol. 99, no.
nas and Propagation, vol. 70, no. 10, pp. 8798–8808, 2022. 3, article no. 033820, 2019.
[132] H. K. Kim, D. Lee, and S. Lim, “Frequency-tunable metamaterial [152] M. Lei, Y. N. Feng, Q. M. Wang, et al., “Magnetically tunable
absorber using a varactor-loaded fishnet-like resonator,” Applied metamaterial perfect absorber,” Journal of Applied Physics, vol.
Optics, vol. 55, no. 15, pp. 4113–4118, 2016. 119, no. 24, article no. 244504, 2016.
[133] Y. Kim, P. C. Wu, R. Sokhoyan, et al., “Phase modulation with [153] A. E. Serebryannikov, A. Lakhtakia, and E. Ozbay, “Characteris-
electrically tunable vanadium dioxide phase-change tic attributes of multiple cascaded terahertz metasurfaces with
metasurfaces,” Nano Letters, vol. 19, no. 6, pp. 3961–3968, 2019. magnetically tunable subwavelength resonators,” Annalen der
[134] J. Shabanpour, “Programmable anisotropic digital metasurface for Physik, vol. 530, no. 3, article no. 1700252, 2018.
independent manipulation of dual-polarized THz waves based on a [154] S. S. Liu, Y. Long, C. Y. Liu, et al., “Bioinspired adaptive mi-
voltage-controlled phase transition of VO2 microwires ,” Journal croplate arrays for magnetically tuned optics,” Advanced Optical
of Materials Chemistry C, vol. 8, no. 21, pp. 7189–7199, 2020. Materials, vol. 5, no. 11, article no. 1601043, 2017.
[135] Y. F. Wang, P. Landreman, D. Schoen, et al., “Electrical tuning of [155] J. Y. Ou, E. Plum, J. F. Zhang, et al., “An electromechanically re-
phase-change antennas and metasurfaces,” Nature Nanotechnolo- configurable plasmonic metamaterial operating in the near-in-
gy, vol. 16, no. 6, pp. 667–672, 2021. frared,” Nature Nanotechnology, vol. 8, no. 4, pp. 252–255, 2013.
ELECTROMAGNETIC SCIENCE

[136] S. Abdollahramezani, O. Hemmatyar, M. Taghinejad, et al., “Elec- [156] H. S. Ee and R. Agarwal, “Tunable metasurface and flat optical
trically driven reprogrammable phase-change metasurface reach- zoom lens on a stretchable substrate,” Nano Letters, vol. 16, no. 4,
ing 80% efficiency,” Nature Communications, vol. 13, no. 1, arti- pp. 2818–2823, 2016.
cle no. 1696, 2022. [157] L. B. Yan, W. M. Zhu, M. F. Karim, et al., “0.2 λ0 thick adaptive
[137] Y. Z. Hu, T. Jiang, J. H. Zhou, et al., “Ultrafast terahertz frequen- retroreflector made of spin-locked metasurface,” Advanced Mate-
cy and phase tuning by all-optical molecularization of metasur- rials, vol. 30, no. 39, article no. 1802721, 2018.
faces,” Advanced Optical Materials, vol. 7, no. 22, article no. [158] S. Sun, W. H. Yang, C. Zhang, et al., “Real-time tunable colors
1901050, 2019. from microfluidic reconfigurable all-dielectric metasurfaces,” ACS
[138] P. P. Vabishchevich, A. Vaskin, N. Karl, et al., “Ultrafast all-opti- Nano, vol. 12, no. 3, pp. 2151–2159, 2018.
cal diffraction switching using semiconductor metasurfaces,” Ap- [159] C. Li, J. B. Wu, S. L. Jiang, et al., “Electrical dynamic modulation
plied Physics Letters, vol. 118, no. 21, article no. 211105, 2021. of THz radiation based on superconducting metamaterials,” Ap-
[139] A. C. Tasolamprou, A. D. Koulouklidis, C. Daskalaki, et al., “Ex- plied Physics Letters, vol. 111, no. 9, article no. 092601, 2017.
 0020111-20 Electromagnetic Science, vol.1, no.2

[160] K. Shih, P. Pitchapp., M. Manjapp., et al., “Active MEMS meta- [179] V. Arun and H. Balakrishnan, “RFocus: Beamforming using thou-
materials for THz bandwidth control,” Applied Physics Letters, sands of passive antennas,” in Proceedings of the 17th USENIX
vol. 110, no. 16, article no. 161108, 2017. Symposium on Networked Systems Design and Implementation,
[161] M. Li, L. Shen, L. Q. Jing, et al., “Origami metawall: Mechanical- Santa Clara, CA, USA, pp. 1047–1062, 2020.
ly controlled absorption and deflection of light,” Advanced Sci- [180] K. Tang, N. Zhang, K. Chen, et al., “Reconfigurable intelligent
ence, vol. 6, no. 23, article no. 1901434, 2019. surface enhancing in-door wireless communication,” in Proceed-
[162] S. Bildik, S. Dieter, C. Fritzsch, et al., “Reconfigurable folded re- ings of the 2021 IEEE International Workshop on Electromagnet-
flectarray antenna based upon liquid crystal technology,” IEEE ics: Applications and Student Innovation Competition, Guangzhou,
Transactions on Antennas and Propagation, vol. 63, no. 1, pp. China, pp. 1–3, 2021.
122–132, 2015. [181] L. Zhang, M. Z. Chen, W. K. Tang, et al., “A wireless communi-
[163] M. H. Dahri, M. H. Jamaluddin, M. I. Abbasi, et al., “A review of cation scheme based on space- and frequency-division multiplex-
wideband reflectarray antennas for 5G communication systems,” ing using digital metasurfaces,” Nature Electronics, vol. 4, no. 3,
IEEE Access, vol. 5, pp. 17803–17815, 2017. pp. 218–227, 2021.
[164] J. Ning, K. Chen, W. B. Zhao, et al., “An ultrathin tunable meta- [182] D. Kitayama, Y. Hama, K. Goto, et al., “Transparent dynamic
material absorber for lower microwave band based on magnetic metasurface for a visually unaffected reconfigurable intelligent
nanomaterial,” Nanomaterials, vol. 12, no. 13, article no. 2135, surface: Controlling transmission/reflection and making a window
2022. into an RF lens,” Optics Express, vol. 29, no. 18, pp.
[165] X. Yang, S. H. Xu, F. Yang, et al., “A broadband high-efficiency 29292–29307, 2021.
reconfigurable reflectarray antenna using mechanically rotational [183] J. W. Tang, M. Y. Cui, S. H. Xu, et al., “Transmissive RIS for 6G
elements,” IEEE Transactions on Antennas and Propagation, vol. communications: Design, prototyping, and experimental demon-
65, no. 8, pp. 3959–3966, 2017. strations,” arXiv preprint, arXiv: 2206.15133, 2022.
[166] E. Vassos, J. Churm, and A. Feresidis, “Ultra-low-loss tunable [184] A. Araghi, M. Khalily, M. Safaei, et al., “Reconfigurable intelli-
piezoelectric-actuated metasurfaces achieving 360° or 180° dy- gent surface (RIS) in the Sub-6 GHz band: Design, implementa-
namic phase shift at millimeter-waves,” Scientific Reports, vol. 10, tion, and real-world demonstration,” IEEE Access, vol. 10, pp.
no. 1, article no. 15679, 2020. 2646–2655, 2022.
[167] X. Yang, S. H. Xu, F. Yang, et al., “A mechanically reconfig- [185] H. Jeong, E. Park, R. Phon, et al., “Mechatronic reconfigurable in-
urable reflectarray with slotted patches of tunable height,” IEEE telligent-surface-driven indoor fifth-generation wireless communi-
Antennas and Wireless Propagation Letters, vol. 17, no. 4, pp. cation,” Advanced Intelligent Systems, vol. 4, no. 12, article no.
555–558, 2018. 2200185, 2022.
[168] W. Lee and Y. J. Yoon, “A broadband dual-metallic-reflectarray [186] F. Liu, A. Pitilakis, M. S. Mirmoosa, et al. , “Programmable meta-
antenna for millimeter-wave applications,” IEEE Antennas and surfaces: State of the art and prospects,” in Proceedings of the
Wireless Propagation Letters, vol. 16, pp. 856–859, 2017. 2018 IEEE International Symposium on Circuits and Systems, Flo-
[169] V. F. Fusco, “Mechanical beam scanning reflectarray,” IEEE rence, Italy, pp. 1–5, 2018.
Transactions on Antennas and Propagation, vol. 53, no. 11, pp. [187] M. D. Renzo, M. Debbah, D. T. Phan-Huy, et al., “Smart radio en-
3842–3844, 2005. vironments empowered by reconfigurable AI meta-surfaces: An
[170] R. Y. Deng, F. Yang, S. H. Xu, et al., “A 100-GHz metal-only re- idea whose time has come,” EURASIP Journal on Wireless Com-
flectarray for high-gain antenna applications,” IEEE Antennas and munications and Networking, vol. 2019, no. 1, article no. 129,
Wireless Propagation Letters, vol. 15, pp. 178–181, 2016. 2019.
[171] Q. Hu, J. M. Zhao, K. Chen, et al., “An intelligent programmable [188] C. W. Huang, S. Hu, G. C. Alexandropoulos, et al., “Holographic
Omni-metasurface,” Laser & Photonics Reviews, vol. 16, no. 6, ar- MIMO surfaces for 6G wireless networks: Opportunities, chal-
ticle no. 2100718, 2022. lenges, and trends,” IEEE Wireless Communications, vol. 27, no.
[172] W. K. Tang, M. Z. Chen, J. Y. Dai, et al., “Wireless communica- 5, pp. 118–125, 2020.
tions with programmable metasurface: New paradigms, opportuni- [189] Y. C. Liang, R. Z. Long, Q. Q. Zhang, et al., “Large intelligent
ties, and challenges on transceiver design,” IEEE Wireless Com- surface/antennas (LISA): Making reflective radios smart,” Jour-
munications, vol. 27, no. 2, pp. 180–187, 2020. nal of Communications and Information Networks, vol. 4, no. 2,
[173] S. Foo, “Liquid-crystal reconfigurable metasurface reflectors,” in pp. 40–50, 2019.
Proceedings of the 2017 IEEE International Symposium on Anten- [190] D. Bhandari, C. A. Murthy, and S. K. Pal, “Genetic algorithm with
nas and Propagation & USNC/URSI National Radio Science elitist model and its convergence,” International Journal of Pat-
Meeting, San Diego, CA, USA, pp. 2069–2070, 2017. tern Recognition and Artificial Intelligence, vol. 10, no. 6, pp.
[174] C. Liaskos, A. Tsioliaridou, and S. Ioannidis, “Towards a circular 731–747, 1996.
economy via intelligent metamaterials,” in Proceedings of the [191] X. D. Meng, R. X. Liu, H. C. Chu, et al., “Through-wall wireless
2018 IEEE 23rd International Workshop on Computer Aided Mod- communication enabled by a metalens,” Physical Review Applied,
eling and Design of Communication Links and Networks, vol. 17, no. 6, article no. 064027, 2022.
Barcelona, Spain, pp. 1–6, 2018. [192] B. Sanz-Izquierdo and E. A. Parker, “Dual polarized reconfig-
[175] L. L. Dai, B. C. Wang, M. Wang, et al., “Reconfigurable intelli- urable frequency selective surfaces,” IEEE Transactions on Anten-
gent surface-based wireless communications: Antenna design, pro- nas and Propagation, vol. 62, no. 2, pp. 764–771, 2014.
totyping, and experimental results,” IEEE Access, vol. 8, pp. [193] M. Bouslama, M. Traii, T. A. Denidni, et al., “Beam-switching an-
ELECTROMAGNETIC SCIENCE

45913–45923, 2020. tenna with a new reconfigurable frequency selective surface,”

[176] J. Y. Lau and S. V. Hum, “A wideband reconfigurable transmitar- IEEE Antennas and Wireless Propagation Letters, vol. 15, pp.
ray element,” IEEE Transactions on Antennas and Propagation, 1159–1162, 2016.
vol. 60, no. 3, pp. 1303–1311, 2012. [194] Z. J. Zhang, L. L. Dai, X. B. Chen, et al., “Active RIS vs. passive
[177] X. Tan, Z. Sun, D. Koutsonikolas, et al. , “Enabling indoor mobile RIS: Which will prevail in 6G?,” IEEE Transactions on Communi-
millimeter-wave networks based on smart reflect-arrays,” in Pro- cations, vol. 71, no. 3, pp. 1707–1725, 2023.
ceedings of the 2018 IEEE Conference on Computer Communica- [195] Q. Ma, L. Chen, H. B. Jing, et al., “Controllable and pro-
tions, Honolulu, HI, USA, pp. 270–278, 2018. grammable nonreciprocity based on detachable digital coding
[178] H. T. Zhao, Y. Shuang, M. L. Wei, et al., “Metasurface-assisted metasurface,” Advanced Optical Materials, vol. 7, no. 24, article
massive backscatter wireless communication with commodity Wi- no. 1901285, 2019.
Fi signals,” Nature Communications, vol. 11, no. 1, article no. [196] S. Taravati and G. V. Eleftheriades, “Full-duplex reflective beam-
3926, 2020. steering metasurface featuring magnetless nonreciprocal amplifica-
Reconfigurable Intelligent Surfaces: Design, Implementation, and Practical Demonstration 0020111-21

tion,” Nature Communications, vol. 12, no. 1, article no. 4414, Kai Qu received the B.S. degree in optical engin-
2021. eering from the Harbin Institute of Technology,
[197] S. Taravati and G. V. Eleftheriades, “Programmable nonreciprocal Harbin, China, in 2019. He is currently pursuing
meta-prism,” Scientific Reports, vol. 11, no. 1, article no. 7377, the Ph.D. degree in electronic science and techno-
2021. logy with Nanjing University, Nanjing, China. His
[198] R. Q. Liu, Q. Q. Wu, M. D. Renzo, et al., “A path to smart radio current research focuses on the multifunctional
environments: An industrial viewpoint on reconfigurable intelli- metasurfaces. (Email: kqu@[Link])
gent surfaces,” IEEE Wireless Communications, vol. 29, no. 1, pp.
202–208, 2022.
[199] S. M. Gong, X. Lu, D. T. Hoang, et al., “Toward smart wireless
communications via intelligent reflecting surfaces: A contempo- Weixu Yang received the B.S. degree in commu-
rary survey,” IEEE Communications Surveys & Tutorials, vol. 22,
nication engineering from Nanjing University of
no. 4, pp. 2283–2314, 2020.
Science and Technology, Nanjing, China, in 2019.
[200] C. Liaskos, L. Mamatas, A. Pourdamghani, et al., “Software-de-
He is currently pursuing the Ph.D. degree in elec-
fined reconfigurable intelligent surfaces: From theory to end-to-
tronic science and technology with Nanjing Uni-
end implementation,” Proceedings of the IEEE, vol. 110, no. 9, pp.
1466–1493, 2022. versity. His current research interests include elec-
tromagnetic metasurfaces, reflectarray/transmit-
Yijun Feng received the Ph.D. degree from the array antennas, and millimeter-wave antennas.
Department of Electronic Science and Engineer- (Email: wxyang@[Link])
ing, Nanjing University, China, in 1992. Since
then, he has been a faculty member and is a Full Yilin Zheng received the B.S. degree in telecom-
Professor and Deputy Dean of the School of Elec- munications engineering from Nanjing University,
tronic Science and Engineering, Nanjing Univer- Nanjing, China, in 2018, where she is currently
sity. From September 1995 to July 1996, he was a pursuing the Ph.D. degree in electronic science
Visiting Scientist at the Physics Department of and technology under the supervision of Prof.
Technical University of Denmark, Denmark. From Yijun Feng. She has published 4 peer-reviewed
August 2001 to August 2002, he was a Visiting Researcher at the Uni- first-author journal articles. Her current research
versity of California, Berkeley, USA. His research interests include interests include origami-inspired metamaterials
EM metamaterials and their application to microwave and photonic for beam manipulation and wireless communica-
devices, EM wave theory, and novel microwave functional materials. tion based on metamaterials. (Email: DG20230062@[Link])
Dr. Feng has served as the General Co-Chair of 2018 IEEE Interna-
tional Workshop on Antenna Technology, and Technical Program Co- Ke Chen received the B.S. and Ph.D. degrees in
Chair of 2013 International Symposium on Antennas and Propagation. electronic science and engineering from Nanjing
He is the recipient of the 2010 Science and Technology Award (first University, Nanjing, China, in 2012 and 2017, re-
grade) of Jiangsu Province, and the 2021 Science and Technology spectively. He is currently an Associate Professor
Award (first grade) of Shanxi Province, China. He has authored or co- at the Department of Electronic Engineering,
authored over 200 peer-reviewed journal papers and over 160 refer- School of Electronic Science and Engineering,
eed international conference papers. (Email: yjfeng@[Link]) Nanjing University. His research interests include
electromagnetic metamaterials and metasurfaces
Qi Hu received the B.S. degree in communica- and their applications to wireless communication
tion engineering from the Communication Uni- and photonic devices. Dr. Chen is the recipient of the Best Excellent
versity of China, Beijing, China, in 2019. She is Doctoral Dissertation Award of China Education Society of Electron-
currently pursuing the Ph.D. degree in electronic ics in 2018, the Young Scientist Award of URSI General Assembly
science and technology with Nanjing University, and Scientific Symposium (GASS) in 2021, and the Young Scientist
Nanjing, China. Her current research focuses on Award of International Applied Computational Electromagnetics So-
metasurfaces. (Email: hq1111@[Link]) ciety (ACES) in 2021. He has authored or co-authored over 70 peer-
reviewed journal articles and over 60 refereed international confer-
ence papers. (Email: [Link]@[Link])
ELECTROMAGNETIC SCIENCE