Reconfigurable

Intelligent Surfaces:
Simplified-Architecture
Transmitters—From Theory
to Implementations
This article provides an overview of the theoretical models for space–time digital
metasurfaces and information metasurfaces, the mechanisms for wavefront shaping,
and the signal modulation models in space and time for wave–matter interactions.
By Q IANG C HENG , Senior Member IEEE, L EI Z HANG , Member IEEE, J UN YAN DAI ,
WANKAI TANG , J UN C HEN K E , S HUO L IU , J ING C HENG L IANG , S HI J IN , Senior Member IEEE,
AND T IE J UN C UI , Fellow IEEE

ABSTRACT | Reconfigurable intelligent surfaces (RISs) offer manipulation is not the only skill of the RISs. With the rapid
an entirely new route to alter the propagation properties developments of space–time digital metasurface and informa-
of electromagnetic waves and thus control their reflection, tion metasurface, there has been increasing attention focused
refraction, and scattering features in arbitrary manners. Such on the information manipulation via these artificial surfaces.
physical attributes are perceived to bring about fundamen- In this article, we provide an overview of the theoretical
tal influence on the modern wireless communication system models of the space–time digital metasurface and information
due to the possibilities to establish artificial and controllable metasurface, the mechanisms of wavefront shaping, and the
propagation environments for radio signals, which no longer signal modulations in space and time domains during the
rely on the complex encoding, decoding, and other signal wave–matter interactions. We will also address some practical
processing techniques. Recent studies reveal that the wave issues during implementations of the reconfigurable intelligent
metasurfaces and the associated hardware architectures at
microwave frequencies to realize simplified radio frequency
Manuscript received 19 September 2021; revised 16 January 2022; accepted
17 April 2022. Date of publication 5 May 2022; date of current version
transmitters. Several modulation schemes and the correspond-
19 September 2022. This work was supported in part by the National Key ing demonstration systems are introduced to illustrate the
Research and Development Program of China under Grant 2017YFA0700201,
Grant 2018YFA0701904, Grant 2017YFA0700202, and Grant 2017YFA0700203;
powerful abilities of the reconfigurable intelligent metasur-
in part by the National Natural Science Foundation of China under Grant faces. Potential research directions of this technique are briefly
61631007, Grant 61731010, and Grant 11227904; in part by the China
Postdoctoral Science Foundation under Grant 2020M680062; in part by the
discussed for their potential applications in future wireless
Jiangsu Province Frontier Leading Technology Basic Research Project under Grant networks.
BK20212002; in part by the 111 Project under Grant 111-2-05; and in part by the
Southeast University-China Mobile Research Institute Joint Innovation Center. KEYWORDS | Digital-coding metasurface; information meta-
(Qiang Cheng, Lei Zhang, and Jun Yan Dai contributed equally to this work.)
(Corresponding author: Tie Jun Cui.) surface; reconfigurable intelligent surface (RIS); space cod-
Qiang Cheng, Lei Zhang, Jun Yan Dai, Jun Chen Ke, Shuo Liu, Jing Cheng ing; spectral modulation; time coding; transmitter; wavefront
Liang, and Tie Jun Cui are with the State Key Laboratory of Millimeter Waves,
Southeast University, Nanjing 210096, China (e-mail: [email protected]; shaping; wireless network.
[email protected]).
Wankai Tang and Shi Jin are with the National Mobile Communications
Research Laboratory, Southeast University, Nanjing 210096, China (e-mail: I. I N T R O D U C T I O N
[email protected]). With the rapid growth of wireless communication
Digital Object Identifier 10.1109/JPROC.2022.3170498 technology, reconfigurable intelligent surface (RIS) has

0018-9219 © 2022 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

1266 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

been recognized as an innovative paradigm for the meta-atom is quite suitable for RIS implementation [31].
electromagnetic (EM) wave manipulation, signal To avoid complex feeding network, it is also feasible
modulation, and smart radio environment reconfiguration to realize the digital-coding metasurface remotely with
[1]–[5]. RIS is a 2-D artificial metasurface composed light-addressable meta-atoms [26], [27]. By elaborately
of periodic or aperiodic subwavelength electric or arranging the digital elements on the metasurface, we can
magnetic resonators, which can exhibit dynamic spatially engineer the reflected/transmitted wavefront
reflection/transmission amplitude and phase responses freely in real time.
[6]–[9], thereby enabling software-defined phase The coding strategies of digital-coding metasurfaces can
discontinuities during wave–matter interactions, reshaping be applied in space domain, time domain, or in both
the reflected/transmitted wavefronts, and creating diverse domains simultaneously. Early studies in this area were
scattering patterns for incident waves from different focused on the arrangement of “0” and “1” elements
directions [6], [9]–[13]. Due to the advantages of at different positions on the metasurface, i.e., the cod-
ultrathin profile, low loss, and easy fabrication of the ing sequence in space domain [6]. It was later found
2-D RISs, it is promising that RISs are favorable for that such coding strategies can be further extended into
the communication applications to construct novel the time domain. Instead of constructing the spatially
wireless networks with improved spectrum and energy varying phase profile, we can also vary the phase peri-
efficiency [3], [14]–[20]. odically in time [30], [34]–[42]. Due to the periodic
A simple way to realize RISs is to employ reconfig- nature of the phase function, nonlinear harmonic com-
urable metasurfaces with varactor diodes embedded into ponents are consequently generated, implying that the
the meta-atoms. These varactor diodes are generally driven metasurface can interact with the incoming wave in a
by biasing voltages from the digital signal processors and nonlinear style. Tremendous efforts have been devoted
collectively form an inhomogeneous metasurface with spe- to independent control of the harmonic amplitudes and
cific phase profiles [21]–[25]. Although such structures phases, as well as the enhancement of the harmonic
can in principle realize continuous phase tuning in full conversion efficiency [39]–[41]. The time-varying meta-
phase range, there are significant challenges for global surface can also be constructed via temporally variant
performance optimizations when millions of elements are constitutive parameters, including the permittivity or sur-
included in the surfaces [22], [25]. In addition, since the face impedance [43]–[45]. A variety of unique functional-
desired scattering patterns of RISs may require abrupt ities have been demonstrated such as nonreciprocal and
phase changes for adjacent unit cells, it poses great chal- full-duplex systems [38], [46]. Furthermore, the com-
lenges in the design of biasing networks [23], [26], [27]. bination of coding operations in both space and time
These problems can be well solved by digital-coding meta- domains opens a new avenue for manipulating wavefronts
surfaces, which are the basis of the RIS. The concept and frequency spectra simultaneously, which is difficult to
of the digital-coding metasurface was first proposed by accomplish with naturally occurring nonlinear materials.
Cui et al. [6]. It allows wideband phase modulations A straightforward application of RISs and digital-coding
and scattering pattern syntheses in real time. Digital- metasurfaces is to build programmable wireless environ-
coding metasurface can realize complex wave manipu- ment or smart radio environment by coating them on
lations by simply changing the coding sequences in a the surfaces of large scatterers [47]–[51]. It allows us to
discretized manner, which can greatly simplify the design directly reconfigure the EM properties of incident wave by
and optimization process and can also be implemented software. Different from the uncontrollable natural prop-
with simple hardware. This new metasurface has been agation environment, the programmable wireless channel
widely used in the past few years, including RCS reduction, can be made adaptive by optimizing the scattering prop-
reflection/transmission control, beam steering, polariza- erties of built-in RISs to maximize the performance of
tion conversion, holograms, imaging, nonreciprocity, non- entire communication networks [50]. The radio frequency
linearity, information processing, wireless communication, (RF) signals from the transmitters can bypass the physical
and many other applications [28], [29]. obstacles in the propagation path, thereby helping estab-
For an n-bit digital-coding unit cell, there are 2n lish virtual links between base stations (BSs) and users,
phase states when the switching states of incorporated which is especially appealing for the coverage extension
positive-intrinsic-negative (p-i-n) diodes, field-effect tran- of wireless networks [51]. Moreover, smart radio envi-
sistors (FETs), or microelectromechanical system (MEMS) ronment facilitates multipath mitigation with RISs, which
switches are changed [30]. Although the phase quantiza- renders a novel and potential method to resist fast fading
tion errors may lead to problems, including gain reduc- caused by multipath effects and thus improve the quality
tion, broadened beamwidth, and pointing-direction error, of received signals [53]–[55]. Huang et al. [56] further
raising the number of bits n provides a good route to proposed the emerging concept of holographic multiple-
effectively address these issues and achieve a satisfac- input–multiple-output (MIMO) wireless communications
tory performance [31]–[33]. By considering the balance that utilize RISs for sensing, shaping, and optimizing
between the system cost, design complexity, and overall the wireless propagation environment. The above con-
performance, it has been found that a 2-bit digital-coding cepts and technologies still face many practical challenges

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1267

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

such as advanced hardware implementation and channel independently controlled to switch among discretized EM
estimation [57], [58]. responses that collectively shape the wavefront of illumi-
Besides wave manipulation, information modulation nated wave. The functionality of such a metasurface is
and processing are also manifested as prominent therefore determined by the digital status of each unit
capabilities of RISs and digital-coding metasurfaces cell, which can be dynamically controlled by many digital
[34], [59]–[63]. Such RISs are also known as infor- logic devices that support multiple output channels with
mation metasurfaces [64]–[66]. Under the illumination programmable functions, as shown in Fig. 1(a). The digital
of a monochromatic carrier signal, the baseband infor- characterization of metasurface also helps motivate inter-
mation will be conveyed into the reflected/transmitted disciplinary research between the electromagnetism and
signals because the amplitude and phase of reflec- the information theory.
tion/transmission will be modulated accordingly with the Fig. 1(b) shows the unit cell of the first digital-coding
baseband signal [59], [60], [67]–[69]. This configuration metasurface, which is a dielectric slab (F4B, εr = 2.65 and
greatly simplifies the system architectures of modern wire- δ = 0.001) sandwiched between two metal layers. On the
less communications. Different modulation techniques, top layer, a p-i-n diode (Skyworks, SMP-1320) is mounted
such as binary frequency shift keying (BFSK), quadrature between the two metallic structures, with both electrodes
phase shift keying (QPSK), 8PSK, 16 quadrature amplitude connected to the bottom metal layer through two metallic
modulation (QAM), and 256QAM, have been experimen- vias. The p-i-n diode exhibits different capacitance when
tally demonstrated on information metasurface platforms, it is biased in the “ ON ” and “OFF” states under the control
proving them as low-cost and low-complexity architec- of an external voltage, and therefore, by interacting with
tures for the energy-efficient transmitters [34], [59]–[62]. the structure, the incoming waves will be reflected with
Furthermore, the space–time-coding (STC) digital meta- nearly 180◦ phase difference at the working frequency
surface provides an additional degree of freedom to 8.6 GHz, as shown in Fig. 1(c). One can further develop an
design a new communication system with both space- n-bit digital-coding metasurface by carefully engineering
and frequency-division multiplexing for system capacity the reflection phase among discretized values of 360◦ /2n
improvement [63]. Due to the anisotropic feature of the using multiple p-i-n diodes or varactor diodes. Note that
digital meta-atom [76], the information metasurface can the metallic bias lines at the back side of the substrate are
respond distinctly to incident waves with different polar- spaced closely enough to act as a single metal plate, which
izations. Thus, dual-polarized RF transmitters can be con- helps achieve sufficient amplitude for refection.
structed to send the information via different polarization Designing the coding pattern is critical for the digital-
channels independently [70]. coding metasurface to generate various radiation patterns
In this article, we will elaborate on the mechanisms that fit for specific applications. Several methods have been
of RISs, including the space-domain coding, time-domain proposed for synthesizing the coding pattern to achieve the
coding, and STC digital metasurfaces, and provide a desired radiation [71]–[73]. For example, Liu et al. [71]
review of the wave manipulation methods via differ- suggested an approach to redirect the radiation pattern of
ent digital-coding strategies in Section II. On this basis, a coding metasurface to an arbitrary direction by simply
we will discuss the simplified-architecture transmitters in making modulo operation on two coding patterns, with
detail for the wireless communication systems based on one being the pattern to be manipulated and the other
the proposed metasurfaces in Sections III and IV. Finally, being a gradient pattern. Due to the Fourier transform
we outlook the future prospects of RISs and information relationship between the surface current and far field of
metasurfaces and summarize this article in Sections V and the coding metasurface, the modulo of two coding patterns
VI, respectively. in the real space is equivalent to the convolution of their
spatial spectra in reciprocal space, which consequently
II. W A V E M A N I P U L A T I O N S B A S E D O N results in a rotation of the radiation pattern in the far
T H E S P A C E–T I M E-C O D I N G D I G I T A L field. Experimental results showed excellent performance
M E TA S U R FA C E S of such an algorithm in arbitrarily manipulating the ori-
A. Digital-Coding Metasurfaces in Space and Time ented radiation patterns with the complex beam shapes
Domains using only 2-bit coding meta-atoms.
In this section, we make a brief introduction on
2) Nonlinear Harmonic Manipulations: In [34], a new
the working mechanism and general design principle of
branch of the digital-coding metasurface was proposed to
digital-coding metasurfaces in space and time domains,
perform the coding process in the time domain, i.e., time-
respectively, and introduce the interesting functionalities
domain-coding digital metasurface (TDCM). Compared
such as wavefront shaping, nonlinear harmonic manipula-
with the space-domain digital-coding metasurface, TDCM
tions, frequency translation, polarization conversion, and
is a time-varying device using temporally modulated para-
convolution operation.
meters to realize spectrum regulations of the EM waves.
1) Wavefront Shaping: The digital-coding metasur- By temporally modulating the reflection/transmission
face is composed of periodic meta-atoms, which are coefficients with periodic functions, a series of nonlinear

1268 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 1. (a) Working flow of the digital-coding and programmable metasurface, where the preset code sequences are directly applied to the
metasurface through an FPGA. (b) Structure of the meta-atom. (c) Reflected phase spectra of the meta-atom in ON and OFF states.

harmonics will be generated and manipulated when the additional phase shift for the spectral component while
metasurface is illuminated by plane waves, as shown in maintaining its amplitude. As a result, the harmonic ampli-
Fig. 2(a). tudes and phases can be independently synthesized based
Here, we take the reflective phase-coding TDCM for on the combinations of phase difference and time shift of
example. The mechanism of nonlinear harmonic manipula- the TDC sequence. A practical implementation of TDCM
tion can be simply understood by a representative example was proposed in [39] for experimental validation, which
shown in Fig. 2(b), in which the temporal and spectral employs varactor diodes as the tunable components to
waveforms of the reflection coefficient of the proposed modulate the reflection phase via biasing voltages.
TDCM are shown in the left and right columns, respec- To verify the validity of the harmonic amplitude control,
tively. It is obvious that the spectrum of the reflected the reflection spectra under varied phase differences are
wave is that of the reflection coefficient shifted by the measured in Fig. 2(f). Different combinations of bias-
incident frequency, as can be regarded as a spatially mixing ing voltages correspond to different phase differences.
effect. It is worth noting that the above conclusion is only Then, several sets of phase difference and time shift
valid when the temporal modulation frequency is much are introduced in the generation of the TDC sequence
smaller than the incident frequency to fulfill the adiabatic of TDCM columns. The measured scattering patterns of
approximation [74]. In addition, to demonstrate nonlinear the +1st-order harmonic are shown in Fig. 2(g), which
harmonic manipulations of the EM waves by time-domain- clearly shows the independent controls of harmonic power
coding (TDC) sequences, Fig. 2(c) and (d) shows the phase intensity and beam shape. The good agreement between
waveforms and harmonic intensities in the case of a 1-bit theoretical calculations and experimental results validates
(“. . . 0101. . . ”) and a 2-bit (“. . . -00-01-00-01-. . . ”) TDC the efficiency and simplicity of this coding strategy in the
sequences, respectively. It is obvious that the harmonic nonlinear harmonic manipulations.
amplitudes can be regulated flexibly according to the phase
difference between two phase states of the TDC sequences. 3) Frequency Translation: It is an important application
However, the harmonic phases are also changed during the of the TDCM for new harmonic synthesis [34], [35],
amplitude regulation, which is unfavorable for nonlinear [39]. In general, frequency translation with high energy
harmonic manipulation in real applications. conversion efficiency and excellent harmonic suppression
To overcome the inherent coupling between the har- is especially favored for engineering. Actually, the reflec-
monic amplitudes and phases, a novel method was pro- tion spectra of the TDCM are highly related to the code
posed to obtain independent controls of the two properties length L, period T , and arrangement of the temporal
[39], as shown in Fig. 2(e). According to the Fourier coding sequence. For example, the 1-bit temporal coding
transform theory, the time delay in temporal signals brings sequence “. . . 0101. . . ” with L = 2 and T = 1 μs will

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1269

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 2. (a) Schematic of nonlinear harmonic manipulation based on TDCM. (b) Temporal waveforms (right) and spectra (left) of the
reflection coefficient, incident wave, and reflected wave when the TDCM is under the illumination of a plane wave. (c) and (d) Phase
waveforms and harmonic intensities in the case of 1- and 2-bit TDC sequences. (e) Conceptual illustration of independent control of harmonic
amplitudes and phases. (f) Measured harmonic distributions of the reflected wave with different phase differences. (g) Measured 1st-order
harmonic scattering patterns with different combinations of phase difference and time delay. (h) Schematic of efficient frequency
translation. (i) and (j) Phase waveforms and harmonic amplitude distributions under different bit depths. (k) Measured reflection amplitude
spectra for frequency translation.

generate a series of symmetrically distributed odd-order Fig. 2(i) and (j) shows the phase waveforms and the corre-
harmonics with the frequency interval of 1000 kHz, and sponding harmonic amplitude distributions under different
the ±1st-order harmonics are dominant. While the 2-bit bit depths. It can be seen that the energy conversion
temporal coding sequence “. . . -11-10-01-00-. . . ” with L = efficiency of the target +1st-order harmonic increases with
4 and T = 2 μs will lead to an asymmetrical harmonic the growth of bit number. In particular, if a 3-bit TDC
distribution, where the frequency interval is 500 kHz and sequence is employed, the conversion efficiency reaches
only the −1st-order harmonic is dominant. The energy above 90%. It can even approach 100% by further raising
conversion efficiency and high-order harmonic suppression the bit depth. In addition, the spurious interference at
ratio of the maximum spurious interference in both cases higher bit depths will also be effectively suppressed. The
are nearly 40%, 0 dB and 80%, 9.5 dB, respectively. measured result shows an energy conversion efficiency of
Therefore, it is reasonable to infer that better performance more than 88% and a spurious suppression ratio of over
can be obtained if higher bit-depth TDC sequences are 21 dB in Fig. 2(k).
adopted. To this end, a TDCM is proposed with contin-
uous phase tuning range and excellent phase linearity. 4) Polarization Conversion: To date, most of the
It enables the realization of a frequency synthesizer [40] reconfigurable metasurfaces polarizers are sensitive to
with ultrahigh conversion efficiency, as shown in Fig. 2(h). incident polarization [67]–[69], and their capabilities in

1270 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 3. (a) Illustration of anisotropic TDCM for linear and nonlinear polarization syntheses. (b) Polarization conversion diagram between
two LP waves of any polarization direction or between LP wave and LHCP/RHCP wave. (c) Generation of the −1st and 1st harmonic
generation with specific polarization states by altering the time delay factor t0 and initial phase factor ϕ0 of different periodic pulses.

polarization manipulations are very limited. In addition, waves can be steered to the desired directions, which
they cannot control the polarization state of nonlinear provides an accurate and efficient way for scattering
harmonics. Recently, Ke et al. [79] proposed a generalized manipulation [71]. As a counterpart, with the aid of
polarization synthesis scheme based on anisotropic time- TDCM, Zhang et al. [80] proposed a time-domain convo-
domain digital-coding metasurface, which provides the lution theorem to achieve flexible and accurate control
first realization of the programmable linear and nonlinear of the nonlinear harmonics [see Fig. 4(a)]. As mentioned
polarization syntheses. The main functions include the in [39], when a time delay t0 is introduced to the time-
real-time conversions among multiple polarization states varying reflection coefficient Γ(t), the reflection spec-
of the fundamental wave and nonlinear harmonic polar- trum will have an additional phase factor e−2πkf0 t0 at
ization syntheses [see Fig. 3(a)]. Since the anisotropic the kth-order harmonic frequency without affecting the
meta-atoms can control the time-varying reflection coef- reflection amplitude. Consequently, by dynamically updat-
ficients Γxx (t) and Γyy (t) independently along two orthog- ing the time-delay gradients among meta-atoms, the har-
onal directions, the periodic phase pulse modulation was monic phase profile of metasurface can be adjusted in
introduced along one polarization direction, whereas in real time. Note that the observed phenomenon can be
the other polarization, no modulation is applied. For the viewed as a nonlinear extension of the generalized Snell’s
fundamental harmonic, by changing the duty cycle τ of law [81]. Furthermore, if the initial harmonic coding
the pulse signal, one can realize the conversion between sequence is convolved with a preset harmonic beam deflec-
two linearly polarized (LP) waves with arbitrary angle or tion sequence, the harmonic beams can be steered along
the conversion between the LP wave and left/right-hand the desired directions with little aiming error.
circularly polarized (LHCP/RHCP) wave [see Fig. 3(b)]. Fig. 4(b)–(g) shows the scattering patterns of the
To further manipulate the high-order harmonic polariza- −1st-order harmonic based on the time-domain convolu-
tion, the independent phase pulse modulation signals with tion theorem. Four different time delays t0 = 0, T /4, T /2,
the same periods are adopted in two polarization direc- and 3T /4 are employed to construct the 2-bit harmonic
tions, which is different from the modulation scheme of phase states “00,” “01,” “10,” and “11,” as indicated by four
the polarization syntheses at the fundamental harmonic. different gray levels. When the initial coding sequence S1
With the time delay t0 and initial phase factor ϕ0 of “0 0 . . . 1 1 . . . ” is convolved with another beam deflection
different periodic pulses, the metasurface can generate sequence S2 “00 00 01 01 10 10 11 11 00 00 01 01 10 10
the desired polarization states for −1st- and +1st-order 11 11 . . . ,” a new harmonic coding sequence is produced
harmonics, respectively [see Fig. 3(c)]. In the experiment, [see Fig. 4(d)]. From Fig. 4(e) to (g), it can be observed
an anisotropic metasurface with large phase range and that the original harmonic mainlobes are steered away
low return loss was manufactured to verify the proposed from the normal axis by θ2 , which is consistent with the
general polarization synthesis scheme, and both the lin- theoretical analysis [80].
ear and nonlinear harmonic polarization syntheses were
clearly demonstrated [79].
B. Advanced Wave Manipulations Based on
5) Convolution Operation: In [71], a convolution Space–Time-Coding Digital Metasurfaces
theorem was proposed in the space domain. By performing In 2018, Zhang et al. [35] proposed the general theory
the convolution operation, the scattering patterns of EM of STC digital metasurface, which can control both spatial

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1271

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 4. (a) Illustration of the TDCM-based time-domain convolution theorem for harmonic beam manipulation [80]. (b)–(d) Calculated 3-D
scattering patterns under different time-domain convolution strategies. (e)–(g) Measured 2-D scattering patterns under corresponding
time-domain convolution strategies.

and temporal characteristics of the scattering pattern. generally comprises a 2-D array of digital meta-atoms
It has been successfully applied for complex wave manipu- loaded with tunable devices (e.g., p-i-n diodes). Each
lations in both space and frequency domains. As schemat- of them is independently controlled by a periodic time-
ically shown in Fig. 5(a), the STC digital metasurface coding sequence. By applying different control signals to

Fig. 5. Harmonic beam steering, beam shaping, and programmable nonreciprocity based on the STC digital metasurfaces. (a) Conceptual
illustration of the STC digital metasurfaces. (b) Optimized STC matrix for realizing harmonic beam steering. (c) Corresponding 2-D scattering
patterns of harmonic beam steering. (d) Optimized STC matrix for beam shaping. (e) Corresponding 2-D scattering patterns of the single
deflected beam. (f) and (g) Corresponding 2-D and 3-D scattering patterns of the vortex beam, respectively. (h) Conceptual illustration of
nonreciprocal effect based on an STC digital metasurface. (i) and (j) Numerically calculated scattering patterns for the forward scenario
(excitation from Port 1 at fc and θi  60◦ ) and the time-reversal scenario (excitation from Port 2 at fc f0 and θi  20.3◦ ), respectively.

1272 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

the programmable meta-atoms, the digital states of the The STC digital metasurface can generate Doppler-like
STC digital metasurface are jointly encoded in the space frequency shift so that the reflection carrier frequency has
and time domains, resulting in simultaneous manipulation a slight offset compared to the incident signal, resulting in
of the reflected wavefront and spectrum. a change in the normal momentum component. Therefore,
In this section, we will focus on several representa- the STC digital metasurface can achieve the simultaneous
tive applications based on the STC digital metasurfaces control of the tangential and normal momentum compo-
for advanced wave manipulations [82], [83], including nents, which breaks the time-reversal reciprocity. In [38],
the harmonic beam shaping/steering, programmable non- a 2-bit STC digital metasurface was designed to realize
reciprocal effect, arbitrary multibit phase generations, the nonreciprocal effect in a programmable way. Fig. 5(h)
independent control of dual harmonics, and joint mul- conceptually shows the process of the nonreciprocal reflec-
tifrequency syntheses, which have promising application tions in a two-port metasurface system. By controlling the
potentials in wireless communications. coding element’s phase responses and the corresponding
space–time gradient coding sequence, the incident plane
1) Harmonic Beam Steering and Beam Shaping With the
wave with the incidence angle θ1 and frequency f1 will be
Optimization Method: Here, we introduce the optimization
reflected to a main beam with the reflection angle θ2 and
method for harmonic beam steering [Fig. 5(a)]. We con-
frequency f2 . For the time-reversal case, an incident plane
sider a reflection-type STC digital metasurface that has
wave with the angle θ2 and frequency f2 will be reflected
eight columns of programmable digital meta-atoms, and
to the main beam with angle θ3 and frequency f3 . Due to
1-bit temporal coding sequences are used in the design.
the space–time modulation introduced in the STC digital
The STC matrix obtained by the binary particle swarm opti-
metasurface, the frequency f3 and angle θ3 are not equal
mization (BPSO) algorithm is shown in Fig. 5(b), in which
to the original frequency f1 and angle θ1 in this two-port
the green and red squares represent “0” and “1” phase
reflection system, that is to say, the reflected wave in the
states, respectively. When the STC digital metasurface is
time-reversal case does not propagate along the direction
modulated by the STC matrix, its far-field scattering pat-
of the original obliquely incident wave in the forward
terns make great difference at various harmonic frequen-
case and is also accompanied by high-efficiency frequency
cies. Fig. 5(c) shows the numerically calculated harmonic
conversion. To illustrate this effect, Fig. 5(i) and (j) shows
steering scattering patterns, from which one can see that
the numerically computed scattering patterns of the non-
the main beams at different harmonic frequencies point to
reciprocal reflection. In the forward scenario, the original
different spatial directions with uniform power levels.
incident angle from port 1 is θ1 = 60◦ , and the reflec-
Due to the time modulation, the STC metasurfaces
tion angle of the main beam is θ2 = 20.3◦ , as shown
can realize beam shaping at both fundamental and har-
in Fig. 5(i). However, for the time-reversal scenario, the
monic frequencies with high accuracy [35]. For exam-
reflection angle of the main beam is θ3 = 51.93◦ . It can be
ple, Fig. 5(d) shows an STC matrix that is composed of
observed from Fig. 5(j) that the angle difference between
eight sets of 2-bit temporal coding sequences, which are
θ3 and θ1 reaches a fairly large value, thereby breaking
elaborately designed to realize the equivalent 3-bit phase
the Lorentz reciprocity. Moreover, the STC digital meta-
quantization at the fundamental frequency. When the STC
surface can easily realize the nonreciprocal effect and the
metasurface is divided into eight columns with the gradi-
frequency conversion in a programmable fashion through
ent phase distribution, the phase discontinuity will lead to
the field-programmable gate array (FPGA), which could
a single beam with a deflection angle of 14.5◦ according
isolate the signal transmissions in both space and fre-
to the generalized Snell’s law [81]. The corresponding 2-D
quency domains and will find promising applications in
scattering pattern at the fundamental frequency is shown
RIS-assisted wireless communications for separating the
in Fig. 5(e), in which the angle of main beam agrees well
transmitting and receiving channels.
with theoretical prediction. Besides, when the STC meta-
surface is divided into eight sectors with a rotated phase
3) Arbitrary Multibit Phase Generations: The time
distribution, the corresponding spiral-like phase profile can
modulation will also help to raise the equivalent phase
generate vortex beam carrying orbital angular momentum
quantization bits with 1- or 2-bit metasurface only, for the
(OAM). Fig. 5(f) and (g) shows the corresponding 2-D and
purpose of precise EM control. In [30], a 2-bit STC digital
3-D scattering patterns at the fundamental frequency, from
metasurface was proposed to realize arbitrary multibit
which a hollow-center profile can be clearly observed in
quantization in the phase range of 360◦ at both funda-
the patterns.
mental and harmonic frequencies, as shown in Fig. 6(a).
2) Programmable Nonreciprocal Effect: Breaking reci- Fig. 6(b) shows the phasor diagram to illustrate the
procity has always been a crucial issue in the EM theory mechanism, in which the physical 2-bit reflection coeffi-
and engineering. How to realize the nonreciprocal effect in cients ejϕ00 , ejϕ01 , ejϕ10 , and ejϕ11 are represented by
some specific applications has also attracted great research four red-arrow vectors. By carefully designing the 2-bit
interests. Instead of using conventional magnetic or non- temporal coding sequences, the new green-arrow vectors
linear materials, nonreciprocal effect can also be realized can be successfully synthesized by superimposing the four
by introducing spatiotemporal modulations [45], [84]. red-arrow vectors. Fig. 6(c) shows the equivalent phase

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1273

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 6. Arbitrary multibit phase state generation, dual-harmonic independent control, and joint multifrequency syntheses based on the
STC digital metasurfaces. (a) Schematic of arbitrary multibit phase generation based on a 2-bit programmable metasurface. (b) Complex
plane representation based on the vector analytical method. (c) Equivalent phases and amplitudes at the fundamental frequency.
(d) Schematic of independent control of dual-harmonic based on STC digital metasurface. (e) and (f) Calculated combinations of initial phase
and time delay, and corresponding 2-D far-field patterns in independent control of 1st- and 2nd-order harmonics. (g) Conceptual
illustration of joint multifrequency syntheses. (h) Schematic of the intertwined time-coding subsequences, in which each subsequence only
affects one specific harmonic frequency. (i) Example of multibeam generations at the four frequencies.

and amplitude at the fundamental frequency under the different influences on the harmonic phases: the former
modulation of different time-coding sequences. It can be introduced the same additional phase to all harmonics,
found that the phase coverage can reach almost 360◦ while the latter brought different phase shifts according
as expected. Therefore, this temporal coding strategy to the harmonic orders. By solving binary linear equations,
provides a new way for the design of arbitrary multi- arbitrary phase responses of the dual harmonics could be
bit coding programmable metasurfaces, which does not synthesized, resulting in the independent controls of dual
require complicated structural designs and control circuit harmonics. Fig. 6(d) shows the schematic of this work,
systems. which realized flexible manipulations of dual-harmonic
beams, including the directions, intensities, and shapes.
4) Independent Control of Dual Harmonics: In [39], extra For visual demonstration, a 3-bit STC metasurface com-
time delay was used to obtain independent control of posed of 8 × 8 elements was taken as an example to
harmonic amplitudes and phases. However, this method synthesize a vortex beam and a deflected beam at the
is only valid for single harmonic manipulation. To increase +1st- and +2nd-order harmonics, respectively. This con-
the harmonic number for independent control, a novel STC figuration required completely different phase profiles for
strategy was proposed in [41]. The initial phase was intro- the two harmonics, indicating that different combinations
duced into the coding sequence to regulate the harmonic of initial phase and time delay need to be introduced into
phases while maintaining the amplitudes, which was very the sequence of each element, as shown in Fig. 6(e). The
similar to the role of time delay. However, they showed corresponding far-field patterns of the two harmonics are

1274 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

shown in Fig. 6(f). Good agreements can be found between transmitter and system prototypes to demonstrate the
the simulations and measurements [41]. potentials of the STC metasurfaces in future wireless
communication technologies, including the BFSK
5) Joint Multifrequency Syntheses: As stated before, transmitter, PSK transmitters, and QAM transmitters.
numerical optimization algorithm can be used to realize
harmonic beam steering at multiple frequencies. However,
such an approach is greatly limited on its general applica-
A. BFSK Transmitter
bility due to the high computational complexity in the The first metasurface-based simplified-architecture
optimization process. The scattering properties of the STC BFSK transmitter was proposed in [34]. According to
digital metasurface were inherently entangled at multi- the feature of BFSK modulation, we need two harmonic
ple frequencies [85], and hence, the joint multifrequency frequencies generated by the STC metasurface to represent
syntheses are hindered. To relax this restriction, a new the baseband digits “0” and “1.” Zhao et al. [34] showed
coding strategy was proposed to realize independent and two types of 2-bit temporal coding sequences, i.e.,
simultaneous syntheses of scattering patterns at multiple “00-01-10-11. . . ” and “11-10-01-00. . . ” for dual harmonic
frequencies. Fig. 6(g) shows the conceptual illustration of generation, which can be mapped to the binary digits
this approach, in which the beam numbers, directions, and “0” and “1” as expected. Specifically speaking, since the
shapes can be simultaneously engineered at the desired two sequences will lead to the −1st - and +1st-order
frequencies. The new coding strategy relies on a more harmonics, it can be categorized as the BFSK modulation
sophisticated STC matrix, which is made up of inter- scheme. As a demonstration, a proof-of-concept prototype
twined time-coding subsequences, as schematically shown was constructed to reach real-time data transmission in
in Fig. 6(h). Suppose that the joint multifrequency synthe- a microwave anechoic chamber, as shown in Fig. 7(c).
ses are considered at Q harmonic frequencies, the temporal During the experiment, a horn antenna was placed
coding sequence is decomposed into Q sets of intertwined in front of the metasurface to feed the carrier wave.
subsequences (represented by the same color square) Meanwhile, the information to be transmitted was
with length P . By carefully designing the individual compiled into a binary bit stream and further mapped
subsequence, one can adjust any specific harmonic [85], into the control signals of the metasurface. Fig. 7(d)
thereby enabling the independent controls of the scat- shows the received spectrum when the system was
tering patterns at multiple frequencies. To demonstrate transmitting a binary digit “0.” Since the power of
the potential of the proposed coding strategy, Fig. 6(i) −1st-order harmonic (−312.5 kHz) was higher than that
shows an illustrative example of multibeam generations at of +1st-order harmonic (+312.5 kHz) by nearly 20 dB,
four frequencies, in which different scattering patterns are the receiver could easily judge the transmitted digit to be
independently synthesized. It is worth noting that the inde- “0.” The system successfully demonstrated real-time image
pendently controlled harmonic number is limited by the transmission with a data rate of 78.125 kb/s. Fig. 7(e)
number of discrete phase states and the switching speed of shows the restored images from two different receiving
the programmable meta-atom. If more harmonics need to angles. It can be seen that the original image can be
be manipulated independently, we need to further increase recovered successfully at the receiver end. To evaluate
the phase bit width and the switching speed of diodes. the performance of the wireless communication system
More details of this approach can be found in [85]. Overall, more quantitatively, the bit error rate (BER) was measured
the approach of joint multifrequency syntheses signifi- under the condition of different transmitting powers,
cantly enhances the wave-manipulation capabilities of STC distances, and receiving angles, as plotted in Fig. 7(f).
digital metasurfaces, which is promising for frequency- The measured waterfall curves demonstrated comparable
division multiplexing in wireless communications. performance to conventional transmitters. The prototype
verified the feasibility to build the STC metasurface-based
architecture transmitter for the first time, providing
III. S I M P L I F I E D-A R C H I T E C T U R E
significant guidance for the subsequent investigations.
TRANSMITTERS AND SYSTEMS
In Section II, we demonstrate the outstanding capability
of the metasurface for wave manipulation in various B. PSK Transmitters
domains. It provides new possibilities to construct Different from the BFSK modulation, the PSK
simplified-architecture transmitters and systems for modulation scheme realizes information transfer via
wireless communications, which can accomplish the the carrier phases. Interestingly, the corresponding
information modulations during the wave–matter mapping relationship is quite similar to the phase
interactions, as shown in Fig. 7(a) [59]. The system coding strategy of the digital-coding metasurface, that
diagram is shown in Fig. 7(b) [86], where the traditional is to say, the binary digits “0” and “1” stand for the
RF components are replaced by the STC metasurface, phase states 0 and π , respectively. Therefore, the STC
with the advantages of high efficiency, low cost, easy digital metasurface is extremely suitable for spatial PSK
assembling, and reduced power consumption. Here, modulation. In addition, such modulation only requires
we introduce some representative simplified-architecture one phase state instead of a certain length of TDC

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1275

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 7. (a) Schematic of simplified-architecture transmitter based on STCM (RIS). (b) Block diagram of this kind of system. (c) Experimental
environments of the proof-of-concept prototype and details of the BFSK transmitter. (d) Received spectrum under transmission of binary
digits “0.” (e) Results of real-time image transmission at the receiving angles of 0◦ (left) and 30◦ (right). (f) BER performance of the system
with different transmitting powers, distances, and receiving angles.

sequence to represent each message symbol, which owns “10,” and “11.” It can be found that both constellation
higher modulation efficiency and transmission rate than diagrams are not very standard according to the require-
the previous BFSK modulation. In [59], a transmitter ment of QPSK modulation scheme, which is caused by the
was proposed to implement the QPSK modulation nonuniform amplitude response of the four phase states
at 4 GHz. Due to the introduction of some traditional for the STC metasurface. In addition, the scatter clusters at
communication technologies, the reliability and BER low data rates are more concentrated than those at higher
of the transmitter have been greatly improved, making data rates, leading to better BER performance and higher
it possible to carry out real-time high-speed wireless reliability. The deterioration at high data rates is caused
communications in the indoor scenario, as shown in by the rapidly distorted control signal within one symbol
Fig. 8(a). duration. Fig. 8(c) shows the measured waveforms of
The distance between the transmitter and the receiver control signal corresponding to a digital symbol at two data
during the demonstration was 2.5 m. The proposed rates, which are actually biasing voltage pulses. The insets
transmitter finally obtained a higher transmission rate illustrate the rise time of the realistic pulses. Obviously,
of 1.6 Mb/s in the verification experiment [59]. As a the signal distortion is worse at higher data rates. Conse-
modulation scheme that requires coherent demodulation quently, it will introduce more noise into the modulated
to restore the digital messages, the constellation diagram signal, thereby reducing the performance of data trans-
is also needed for performance evaluation. In general, the mission. Such issue can be alleviated by applying more
constellation diagram for ideal QPSK modulation contains advanced hardware, so as to reach better performance with
four scatters that have the same amplitude but different faster transmission speed. Following this idea, a similar
phases (45◦ , 135◦ , 215◦ , and 315◦ ). As shown in Fig. 8(b), transmitter of the same architecture was further realized
the received constellation diagrams of the prototype at using an upgraded control platform [60]. It implemented
different data rates are plotted. Obviously, there are four an 8PSK modulation scheme with an improved data rate
scatter clusters in the four quadrants. Each of them stands of 6 Mb/s [60]. In terms of theoretical modeling and
for one digital symbol of QPSK modulation, i.e., “00,” “01,” analysis, Basar [61] developed a mathematical framework

1276 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 8. (a) Photograph of the QPSK transmitter during real-time video transmission in the in-door scenario. (b) and (c) Received
constellation diagrams at different data rates and corresponding waveforms of control signals, respectively. (d) Measured constellation
diagrams of QAM transmitter with different modulation schemes using the harmonic modulation method. (f) Illustration of real-time indoor
image transmission using millimeter-wave 256QAM transmitter. (g) Received constellation diagrams of millimeter-wave 256QAM transmitter
under different carrier frequencies.

to evaluate the theoretical symbol error probability (SEP) this method was designed at 4.25 GHz and experimentally
of an RIS-based M-PSK transmitter by using probabilistic verified in the indoor environment. The final realistic
tools. transmission rate was improved up to 10 Mb/s with
16QAM modulation scheme. During the experiment,
a linearly varying phase waveform was used as the
C. QAM Transmitters basic function for amplitude modulation due to its high
As a high-order modulation scheme, the QAM efficiency in the energy conversion from the carrier wave
modulation requires two independent amplitude to the +1st-order harmonic [40]. In addition, the time
modulation signals with the phase difference of π /2. delays were introduced into the basic function for phase
Based on previous studies, an STC digital metasurface modulation. Finally, different modulation schemes were
capable of joint modulations of carrier amplitude successfully implemented for demonstration, including
and phase is required for this purpose. However, the QPSK, 8PSK, and 16QAM. The measured results of
previous prototypes still require the additional correction constellation diagrams are shown in Fig. 8(d), which
strategies during the demodulation process due to the shows standard distributions of the constellation points as
nonstandard constellation diagrams [Fig. 8(d)], which expected. Fig. 8(e) further shows the BER performance of
further reduced the demodulation efficiency, especially the prototype under different transmitting powers, symbol
for the high-order modulation schemes. To overcome rates, and modulation schemes. Clearly, the low-order
this difficulty, an improved modulation method named modulation scheme has a better performance due to the
harmonic modulation was proposed to modulate the simpler constellation. In addition, the symbol rates will
information at a specific harmonic frequency to implement also reduce the BER of the system due to the inevitable
a 16QAM modulation scheme [62]. From [39], the performance degradation of the control circuit at higher
harmonic amplitudes and phases can be precisely modulation frequencies. Moreover, such amplitude and
controlled by tuning phase differences and time delays, phase synthesis cannot be obtained in wideband due to
which provided an efficient way to synthesize arbitrary the element dispersion. Recently, a novel approach is
constellation diagrams. The proof-of-concept prototype of proposed for accurate and broadband manipulations of

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1277

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 9. (a) Illustrations of conventional modulation techniques ASK, FSK, PSK, and the radiation pattern-based modulation technique (last
row). (b) Photograph of the fabricated programmable metamaterial for the new wireless communication system. (c) Schematic illustration of
the application scenario for satellite-to-Earth security communication. (d) Original image to be sent for testing the performance of the
prototype system. Received images at the receiver for three test scenarios: (e) without obstacle, (f) with obstacle, and (g) with obstacle and
channel adaptive algorithm.

harmonic amplitudes and phases [64]. A periodic square modulation, without exploiting the unique capability of
wavefrom was applied as the basic function phase, which wavefront manipulation. Actually, information modulation
used the duty circle and time delay for amplitude and and wavefront manipulation can be realized simultane-
phase modulations, respectively. The main contribution of ously by the STC metasurfaces, leading to new-architecture
this method lies in the independent manipulation of the transmitters and systems. Here, we further present some
harmonic amplitude and phase in broadband, which can prototypes, such as the MIMO system, pattern modula-
reach up to 256QAM modulation scheme at the harmonic tion system, multichannel direct transmission system, and
frequency. Finally, a new STC digital metasurface in the space-/frequency-division multiplexing system. In addi-
Ka-band is fabricated to prototype a millimeter-wave tion, we also demonstrate the path loss models of the RIS
256QAM transmitter [64]. Fig. 8(f) shows the experiment systems as a complement to better demonstrate the func-
configuration of the system for real-time indoor image tionality of RIS in the wireless communication systems.
transmission at 27 GHz. Fig. 8(g) shows the measured
constellation diagrams at the carrier frequencies of 27 and A. Pattern Modulation System
31.15 GHz. Both of them show the standard 256QAM
Fig. 9(a) summarizes some popular modulation tech-
constellation point distributions. It can be concluded
niques, including the amplitude-shift keying (ASK), FSK,
that the harmonic modulation method can not only
PSK, and QAM, which make use of time-varying ampli-
implement the high-order modulation schemes but also
tude, frequency, and phase of a signal, or the combina-
can synthesize the standard constellation diagrams, which
tion of them, of the carrier signal. In this configuration,
is promising to relax the restrictions of the STC digital
the information is modulated only in the time domain,
metasurface in joint modulations of the amplitude and
leading to identical information in all directions where the
phase to some extent.
antenna radiates. The three dimensions of space, which
potentially convey much higher capacity of information
IV. N E W-A R C H I T E C T U R E than the time dimension, are not exploited for signal
TRANSMITTERS AND SYSTEMS modulation until the emergence of MIMO that utilizes spa-
Among those prototypes demonstrated in Section III, the tially separated transmitters and receivers to increase the
STC metasurfaces are only responsible for the information channel capacity [89]. However, MIMO does not exploit

1278 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

the far-field of antennas, which can be considered as an the mapping between the radiation pattern and the code
infinite-dimensional Hilbert space. This indicates that we [Fig. 9(c)]. This makes it a promising secrete communica-
can transmit, in theory, an infinite amount of information tion method at the physical layer and may find applications
through the spatial variation of the far field of an antenna in satellite-to-Earth communications and military wireless
at a single time frame. This idea is schematically shown in communications.
the last row Fig. 9(a), where two distinct radiation patterns Note that a conceptually similar approach, called the
are used for modulating the digital signals “0” and “1.” media-based modulation [91], [92], was proposed prior
The digital-coding metasurface featured by dynamic to the pattern modulation system. It exploits different real-
control of the radiation pattern emerges as an ideal device izations of the wireless channels to create its modulation
for realizing the new communication system with the alphabet by selectively controlling the connection status
radiation pattern modulation technique. Cui et al. [90] of an array of RF mirrors, which are implemented by a
reported a digital wireless communication system based pin-diode-loaded patch antenna.
on the radiation pattern modulation technique using the
digital-coding programmable metasurface. The signal is
directly modulated through the ever-changing radiation B. Multichannel Direct Transmission System
pattern and is radiated to free space under the illumination As a potential solution to extend communication capac-
of an external feeding antenna. Many of the conventional ity, the multichannel direct transmission focuses on synthe-
modules in modern communication systems can thus be sizing different near-/far-field patterns to directly transmit
avoided, such as the mixer and D/A convertors, which the information of different channels to corresponding
leads to a greatly simplified system architecture for digital users located in different directions/positions [93], [94].
communication systems. Compared with the traditional array-based approaches,
A metasurface consisting of 35 × 35 digital units was the digital-coding metasurface is more favorable for the
used to realize the transmitter [Fig. 9(b)], which were multichannel direct transmissions from the perspective of
divided into seven columns, with the digital status of each cost, weight, complexity, and power consumption.
column being independently controlled by FPGA. This con- As shown in Fig. 10(a), a multichannel direct
figuration generated a total number of 26 different radia- transmission system was proposed based on the digital-
tion patterns that support a maximum modulation depth coding metasurfsce, which is able to synthesize different
of 6 bit/symbol. In the real environment, the maximum near-field patterns to control the intensity of the corre-
number of radiation patterns that can be ambiguously sponding pixels in the focus plane. In this way, each user
distinguished is limited by the noise level and the receiver in these positions can have an independent channel to
sensitivity. Note that the modulation depth can be further receive different information through the ASK modula-
enhanced by increasing the number of independent control tion. A proof-of-concept prototype is realized with a 20 ×
blocks of the transmitter and the number of receivers, 20 1-bit reflective programmable metasurface to demon-
as long as the receiving side can distinguish all different strate experiments of three-channel direct transmissions
radiation patterns at a given noise level. Fig. 9(d)–(f) at 10 GHz. The information of the three channels is mod-
shows the system performance of the prototype, in which ulated by a binary ASK (BASK) modulation scheme, which
a binary image [Fig. 9(d)] was chosen as the original indicates independent binary amplitude control at three
data and was sent to the receivers using 2-bit mode trans- specific locations. Therefore, it requires eight coding pat-
mission. The system performance was evaluated by three terns to correspond to every possible case of digital infor-
different test scenarios: without obstacle, with obstacle, mation for the three channels, i.e., “000,” “001,” “010,”
and with obstacle and channel adaptive algorithm. The “011,” “100,” “101,” “110,” and “111.” In order to better
transmission process proceeded correctly in the line-of- fulfill the requirement, a modified Gerchberg–Saxton (GS)
sight propagation [Fig. 9(e)], but was interrupted once a algorithm is applied to optimize the coding patterns. The
metallic obstacle was placed between the transmitter and optimized results and the corresponding near-filed ampli-
the receiver [Fig. 9(f)]. To allow the system correctly work tude distributions are shown in Fig. 10(b), which visually
under the real environment where the wireless channel exhibits the independent controls of intensities at three
was partially blocked, the authors developed a channel desired locations. During experiments, three receivers are
optimization algorithm, which can dynamically find all dis- deployed at these locations to receive the RF signals and
tinguishable radiation patterns for an ambiguous transmis- demodulate the information in their own wireless chan-
sion, and thereby improved the system robustness against nel. As shown in Fig. 10(c), the received waveforms of
channel interference. Fig. 9(g) shows that the image is the three channels after demodulation and digitalization
correctly received with an employment of the developed have shown three-channel direct transmissions straightfor-
channel optimization algorithm. wardly. In addition, it is also worth noting that this method
Most importantly, the new modulation technique pro- is not only promising in capacity expansion but also in
tects the information from being intercepted from either secure communication because only the user located at
a single or multiple position, as the eavesdropper would the right position can recover the correct message, while
not be able to crack the information without knowing the eavesdroppers at other locations will just get invalid

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1279

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 10. (a) Schematic of multichannel direct transmission system. (b) Eight coding patterns for three-channel direct transmissions and
the corresponding simulated near-field amplitude distributions. (c) Received waveforms of the three channels after demodulation and
digitalization. (d) Conceptual illustration of multichannel backscatter wireless communications. (e) Eight coding patterns for three-channel
backscatter wireless communication. (f) Simulated far-field amplitude and phase patterns corresponding to different coding patterns.
(g) Constellation diagrams of the system implementing BPSK modulation scheme in the case of single channel, double channels, and three
channels. (h) Received individual images from three users and the corresponding synthetized results.

information [93]. In particular, similar idea was adopted it can further implement the PSK modulation scheme in
in [94] to further realize multichannel backscatter wireless these wireless channels, such as BPSK and QPSK [94].
communications with enhanced privacy. The transmitter For example, if the system adopts the BPSK modulation
does not need a specific carrier source as before. It is able scheme, it will require independent binary phase control
to modulate the propagation environment of ambient stray at the locations of the three users. As shown in Fig. 10(e),
waves to realize information transmission. eight coding patterns are optimized accordingly, each of
A prototype was constructed to exhibit three-channel which can transmit specific digital information to the three
backscatter wireless communication in the environment channels. The corresponding simulated far-field amplitude
of commodity wireless fidelity (Wi-Fi) signals at 2.4 GHz. and phase patterns are shown in Fig. 10(f). It is obvious
As shown in Fig. 10(d), a wireless router emits stan- to find that the amplitudes of the three users are kept
dard Wi-Fi signals to the environment, which include a quite high in all cases, while their phases are controlled
metasurface-based transmitter (“Alice”), a Wi-Fi user (“lis- independently with the coding patterns. Likewise, a proof-
tener”), and three registered users (“Bob-R,” “Bob-G,” and of-concept system is constructed to show the backscatter
“Bob-B”). Alice can transmit three information streams to wireless communication. Fig. 10(e) shows the measured
these users by modulating the stray Wi-Fi signal reflected constellation of the system for BPSK modulation in single
from the metasurface. The listener is able to perform channel, double channels, and three channels, in which
signal cancellation to mitigate the influence of the envi- the iconic constellation distributions are quite remarkable.
ronment, making it possible to demodulate the infor- Furthermore, a full-color image is decomposed into mono-
mation in the Wi-Fi signal at three receiving terminals. chrome ones and transmitted through the three channels.
Like the three-channel wireless communication system The received individual images from the three users and
discussed above, this one is also able to realize informa- the final synthesized results are shown in Fig. 10(h). This
tion transmission using the ASK modulation. Moreover, system has the potential to increase the communication

1280 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 11. (a) Conceptual illustration of the space- and frequency-division multiplexing for the multichannel wireless communication.
(b) Photograph of dual-channel wireless communication system based on a 2-bit STC digital metasurface. (c) and (d) Block diagrams of the
communication system and the transmitting/receiving processes, respectively. Experiments of dual-channel wireless communication system
for validating the reprogrammble features: (e) two users in nonsymmetric directions and (f) two users in the same direction.

security for its “invisibility” to the existed wireless envi- communication system. Fig. 11(a) shows the conceptual
ronment, which offers a new role of the digital-coding illustration of the multichannel system with the feature of
metasurface in wireless communications. direct data transmission. With the aid of the STC digital
metasurface, different digital messages are directly routed
to multiple users located at different locations indepen-
C. Space-/Frequency-Division Multiplexing dently and simultaneously.
System Without the loss of generality, a dual-channel
To further improve the capacity of wireless communi- information-encoding scheme was first adapted for
cation, multiplexing technique has been widely studied. illustrating the direct transmission of digital messages.
The STC digital metasurface can control the spatial and By optimizing the 2-bit STC matrices of the digital
spectral features of EM waves in space and frequency metasurfaces, the main beams at designated frequencies
domains, which is suitable for implementing the space- can be engineered to point to target users. Therefore,
and frequency-division multiplexing with the advantages the power intensities transmitted to two users can be
of low cost and simple fabrications. In [63], a new independently controlled by altering the STC matrices.
information-encoding scheme based on the STC digital For experimental demonstration, a 2-bit STC metasurface
metasurface was put forward to realize the multiplexing was used to build a realistic dual-channel wireless
techniques in the space and frequency domains, which communication system, in which two different color
can be further used to construct a multichannel wireless pictures (the logo of the State Key Laboratory of Millimeter

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1281

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Waves and Southeast University) were successfully proved the feasibility of using an RIS-based transmitter to
transmitted to two users at different locations, as shown in achieve the high-order modulation and MIMO transmis-
Fig. 11(b). The corresponding transmitting and receiving sion simultaneously. By considering the physics and EM
processes of the dual-channel wireless communication are nature of RISs, the system model of the MIMO wireless
shown in Fig. 11(c) and (d), respectively. communication system empowered by an RIS-based trans-
In this multiplexed communication system, each desired mitter was proposed. According to the proposed analytical
user has a specific frequency to construct its own depen- model, an RIS-based MIMO transmitter architecture was
dent channel and the interference between each other is designed and analyzed, in which each unit cell of RIS can
very low. Besides, this information-encoding scheme has act as an independent transmitter. In other words, each
the characteristic of directional modulation, which can be unit cell can directly modulate the baseband information
further explored for secure communications. onto the reflected wireless signal by adjusting its phase or
To show the reprogrammable features of the proposed amplitude EM responses.
wireless communication scheme, more experimental As shown in Fig. 12(a), by combining the conventional
results with two users located at different directions were MIMO transmission scheme with the RIS-based transmit-
also conducted. Fig. 11(e) and (f) shows the cases of ter, the hardware architecture of an RIS-based MIMO
two users located in nonsymmetric directions and the transmission prototype system was proposed. In order to
same direction, respectively. In both cases, two different achieve the purpose of rapid experimental verification, the
pictures can be successfully transmitted to two users in prototype system realizes basic 2 × 2 MIMO wireless com-
the experiments. Moreover, this space- and frequency- munication. Therefore, the RIS is divided into two parts,
division multiplexing strategy can be extended to multi- and their reflection coefficients are controlled by two inde-
channel wireless communication by elaborately selected pendent and dedicated control signals from the control
the STC matrices. In general, the STC digital meta- circuit board. The central controller together with several
surface provides a low-cost solution for realizing the off-the-shelf modules generates two source bit streams,
space- and frequency-division multiplexing techniques, map digital baseband signals to control signals for the RIS,
without the need for antenna arrays and the expensive RF and deliver them to the control circuit board. Then, the RIS
modules [95]. performs 2 × 2 MIMO transmission through the reflected
waves once it is illuminated by carrier signal at 4.25 GHz
provided by the RF signal generator. As a matter of fact,
D. MIMO Systems the proposed architecture can be generalized for any size
The MIMO theory and technology have gone through of MIMO.
several transformations to upgrade the recent generations Fig. 12(b) shows a picture of the prototype system. The
of wireless communication networks. By exploiting the RIS-based MIMO transmitter is located on the left and
spatial dimension of wireless channels, the MIMO tech- the conventional receiver is on the right. The left and
nology uses multichannel transmitters and receivers to right half parts of the RIS are controlled by an external
reach higher data rate, better link performance, and wider control signal and thus modulate two different bit streams
communication range. In the fifth-generation (5G) com- onto the reflected wireless signals at 20-Mb/s data rate
mercial networks, MIMO has been upgraded to massive when ignoring the overhead of synchronization and pilot
MIMO. By deploying up to 64 antennas and transceiver symbols. After transmission over air, the two bit streams
RF chains at the BS, 5G offers the performance leaps and can be successfully recovered by the receiver and their
currently delivers the enhanced mobile broadband (eMBB) constellations are clear. As shown in Fig. 12(c), the BER
communication service to customers. However, the mas- performances of the two transmitted bit streams were
sive MIMO is facing the issues of high hardware complexity measured at the receiver and then compared to theo-
and high power consumption. Therefore, it is very inter- retical ones. It can be observed that the measured BER
esting and necessary to explore the combination of MIMO curves agree well with the theoretical curves, which indi-
technology and metasurface-based simplified-architecture cates that the RIS-based MIMO transmission scheme can
transmitters. achieve the same BER performance with the conventional
The aforementioned works on the digital metasurface- one under the same signal-to-noise ratio condition, but
based transmitters in Section III belong to the basic with less hardware complexity. The experimental results
single-input–single-output (SISO) wireless communica- given in Fig. 12(d) validate that the proposed transmis-
tions. Fortunately, the 2-D surface/array structures of STC sion scheme is robust with the discrete phase shifts of
digital metasurfaces or RISs naturally bring about new the RIS.
degrees of freedom in the spatial dimension, which facili- The RIS-based MIMO transmitter can support vari-
tates the combination between the MIMO signal processing ous space–time coding schemes. Taking the prototype
algorithms and metasurface hardware resources. presented in [88] as an example, the classic Alamouti
Some pioneering works have carried out preliminary space–time coding transmission scheme was achieved by
studies on the design, analysis, and implementation of RIS- using the RIS-based transmitter, which validates con-
based MIMO transmission systems. Tang et al. [5], [87] vincingly that the extensively studied space–time coding

1282 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 12. MIMO transmission through RIS-based transmitter. (a) Hardware architecture of the prototype. (b) Photograph of the prototype.
(c) Comparison of the measured and theoretical BER performances. (d) BER under different discrete phase shift steps with 2.5-Mbd symbol
rate [5].

t
schemes can be applied in the RIS-based transmitter The larger the angle of incidence θn,m and the angle of
r
architecture. reception θn,m , the larger the path loss. When considering
the entire RIS, the overall path loss is related to the reflec-
tion coefficients of all unit cells, which are determined
E. Path Loss Models of the RIS Systems by the specific application scenario of the RIS. When an
The key feature of RIS systems is the construction of RIS is employed in far-field beamforming scenario, the
the digital space based on the physics of RISs, which overall path loss of the RIS is proportional to (d1 d2 )2 (the
enables various applications in the field of wireless com- square of the product of the two distances) and inversely
munications, such as simplified-architecture transmitters proportional to (M N dx dy )2 , i.e., the square of the total
discussed so far. In addition, by deploying them in the wire- area of the RIS. In contrast, when an RIS is employed in
less propagation paths, RISs are expected to proactively the near-field broadcasting scenario, the overall path loss
customize and enhance the radio environment, thus lead- is approximately proportional to (d1 + d2 )2 (the square of
ing to the emerging concept of “smart radio environments” the sum of the two distances) and is not related to the total
[1], [2], [96]. For example, RISs can perform spatial beam- area of the RIS.
forming to steer wireless signals toward users, thereby To validate the proposed theoretical path loss models
improving the user experience [3], [17]. For RIS-based for the RIS, Tang et al. [4], [97] built a free-space path loss
transmitters and smart radio environment, the path loss measurement system for measuring the amount of signal
characteristics of RISs determine the performance limits of power reflected from the RIS and received by the receiver.
these RIS systems. As shown in Fig. 13(b), the measurement system is com-
Tang et al. [4], [97] proposed free-space path loss posed of an RF signal generator, a Tx horn antenna, an RIS,
models for the RIS systems and reported the measurement an Rx horn antenna, an RF signal analyzer, and some RF
campaigns that validate the proposed models. As shown cables. The two antennas can be moved flexibly to carry
in Fig. 13(a), there are many key parameters that need out path loss measurement for different configurations of
to be considered, including various distances and angles. distances d1 and d2 . Fig. 13(c) shows one fabricated RIS
In particular, the path loss of the subchannel provided by that is employed in the path loss measurement campaigns.
a single unit cell Un,m has been proven to be proportional It consists of 1600 unit cells with 1-bit reflection phase
t r
to (rn,m rn,m )2 (the square of the product of the distances control capability and its operating frequency is 33 GHz
from the transmitter/receiver to this unit cell), and to be in the millimeter waveband.
inversely proportional to (dx dy )2 (the square of the area By configuring the abovementioned RIS into a far-field
of the unit cell) and to an angle-dependent loss factor. beamforming function and setting d1 = 5 m, Fig. 13(d)

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1283

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Fig. 13. Path loss modeling for RIS systems. (a) RIS-assisted wireless communication system. (b) Free-space path loss measurement
system for measuring the amount of power reflected from the RIS to the receiver. (c) Photograph of the utilized RIS. (d) and
(e) Measurements versus RIS path loss modeling.

shows the curve of the measured received signal power as and blockage effect from the feeding/receiving antennas
a function of d2 . It can be observed that the measurement are two main disadvantages. Thus, transmissive [98],
results agree well with the proposed theoretical path loss radiative [99], and waveguide-fed [100] designs should
models for the RIS. By configuring the RIS into a near- be considered as the candidates for the profile reduc-
field broadcasting function and setting d1 = 0.25 m, the tion and blockage elimination. In addition, the tunable
measurement results shown in Fig. 13(e) also validate the component integrated in the meta-atom is another piv-
proposed path loss models for the RIS in the general case otal point, which determines the modulation speed and
and a special case. In general, the measurement campaigns the working frequency of digital-coding metasurfaces. The
reported in [4] and [97] prove that the proposed path loss commonly used tunable devices are p-i-n diodes and var-
models yield the good estimation of the received signal actor diodes, which are usually designed in the microwave
power in RIS-assisted link, which may help researchers to and millimeter-wave bands with the modulation speed less
conduct the link budget calculation and performance limit than 100 MHz. To further improve the performance, new
analysis of the RIS systems. materials or devices will be employed in the meta-atom
design, such as transistor [101], photoconductors [27],
V. F U T U R E R E S E A R C H D I R E C T I O N S graphene [102], and vanadium dioxide [103].
Recent highlighted applications of RISs and digital-coding
metasurfaces have demonstrated their immense potentials
from wave manipulations to wireless communications. B. Advanced STC Strategies
Despite the promising prospects, this emerging con- If the digital meta-atom lays the physical foundation
cept still has several challenges worth investigating. of the STC digital metasurface, the coding strategy
In this section, we will discuss some future research is undoubtedly the algorithmic core to realize various
directions that may be of great importance for pro- functionalities. In previous abundant attempts in spa-
moting the development of RISs and digital-coding tial and spectral EM manipulations, only the phase
metasurfaces in wireless communications in the next coding strategy has been widely applied at fundamen-
stage. tal and harmonic frequencies in the current STC dig-
ital metasurface-based prototypes for implementations.
A. Improved Designs of Digital Meta-Atoms However, the joint amplitude–phase coding method is
Since the majority of the metasurface prototypes are more important to realize arbitrary temporal waveforms
implemented with reflective architectures, the high profile and spatial profiles, which is also very promising in

1284 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

band-limited spectrum synthesis and sidelobe suppres- networks are expected to simultaneously support com-
sion [104]. The frequency coding approach is also munication, sensing, computing, and many other func-
highly desired to control the digital state of element at tions. RISs may play an important role in this technol-
different frequencies independently, leading to the advan- ogy evolution and deserve more interdisciplinary research
tages in multiband applications [105]. In addition, it is efforts.
also valuable to involve intelligent methods in the design
of coding sequences, such as deep learning or machine
learning [107].
E. Electromagnetic Information Theory
C. Practical and Tractable Signal Models As discussed in Sections II, III, and IV, the information
RISs and information metasurfaces are usually com- metasurface and RIS have an important feature to bridge
posed of numerous EM resonant structures. The distance the EM physics world and digital information world, which
between adjacent unit cells is usually kept at the subwave- helps to establish the EM information theory [118], [119].
length scale, which leads to strong mutual coupling among The current Shannon information theory is the foundation
the unit cells. Some researchers have proposed preliminary for modern wireless communications, but it may reach
EM models of mutual coupling [109], [110] and experi- bottleneck in the 5G and 6G wireless technologies because
mentally observed this phenomenon [111]. In addition, a the EM knowledge becomes more important. The informa-
practical phase shift model of the unit cell that captures the tion metasurface and RIS provide a solution for solving
phase-dependent amplitude variation has been developed this problem since both digital-stream information and EM
in [112]. Meanwhile, the anisotropic scattering pattern of information can be defined on a platform of information
the unit cell inherently makes its reflection/transmission metasurface or RIS [118], [119]. The information capac-
coefficient related to the angles of incidence and recep- ity (upper and lower bounds) and orthogonal radiation
tion, even when the control signal applied to this unit patterns can be predicted by the proposed theory [118],
cell remains constant. This angle-dependent feature has [119]. In the future, a general EM information theory
been only initially discussed in some studies [4], [113], should be realized by considering the statistical model,
[114]. Hence, it is necessary to develop practical and determined EM model, signals and EM waves, and com-
tractable signal models that comprehensively consider the plicated environment, which requires the hard work from
abovementioned hardware constraints of RISs, which may the communication and EM communities.
pave the way for understanding the performance limits
and optimizing the transmission schemes of RIS-based VI. C O N C L U S I O N
transmitters systems. In this article, we provide an overview of the simplified-
architecture transmitters based on RISs or the digital-
D. Multifunctional Integration for Future coding metasurfaces. The hardware implementations of
Networks RISs and the mechanisms of wave–matter interactions
The current research efforts on RIS-empowered in STC metasurfaces are discussed in detail. We demon-
simplified-architecture transmitters are mainly focused on strate that by dynamically engineering the amplitude and
the design of direct space–time modulation and transmis- phase responses of the digital meta-atoms, it is possi-
sion schemes. It is worth further exploring their appli- ble to establish a mapping relationship between the EM
cation scenarios in the future wireless networks, such properties and the baseband digital information and thus
as intelligent Internet of Things (IoT), where the RIS- realizing the direct information modulations in the wave
based transmitters have potentials to make use of the reflection or transmission process without using mixer and
wireless signals that already exist in the radio environ- D/A converters required in the traditional superheterodyne
ments for green communications and networking [115]. communication systems. Different mainstream modulation
On the other hand, RISs offer a novel EM platform for schemes have been implemented on the digital-coding
supporting various functions beyond the communications. metasurface platform, and the corresponding RF trans-
For instance, the capability of large-scale RISs to generate mitters have been introduced in detail. In the end, the
and steer narrow beams offers the opportunity for wireless future trend of the RIS-based information systems is dis-
power transfer [116]. In addition, the spatial dimension cussed to further enhance the channel capacity as well
of RISs offers the opportunity to realize indoor position- as the overall performance of wireless communication
ing with improved performance [117]. Future wireless systems.

REFERENCES
[1] E. Basar, M. Di Renzo, J. De Rosny, M. Debbah, empowered by reconfigurable intelligent surfaces: intelligent surfaces for energy efficiency in
M. Alouini, and R. Zhang, “Wireless How it works, state of research, and the road wireless communication,” IEEE Trans. Wireless
communications through reconfigurable ahead,” IEEE J. Sel. Areas Commun., vol. 38, Commun., vol. 18, no. 8, pp. 4157–4170,
intelligent surfaces,” IEEE Access, vol. 7, no. 11, pp. 2450–2525, Nov. 2020. Aug. 2019.
pp. 116753–116773, 2019. [3] C. Huang, A. Zappone, G. C. Alexandropoulos, [4] W. Tang et al., “Wireless communications with
[2] M. D. Renzo et al., “Smart radio environments M. Debbah, and C. Yuen, “Reconfigurable reconfigurable intelligent surface: Path loss

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1285

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

modeling and experimental measurement,” IEEE [24] X. Wu et al., “Active microwave absorber with the Dec. 2014.
Trans. Wireless Commun., vol. 20, no. 1, dual-ability of dividable modulation in absorbing [45] Y. Hadad, D. L. Sounas, and A. Alu, “Space-time
pp. 421–439, Jan. 2021. intensity and frequency,” AIP Adv., vol. 3, no. 2, gradient metasurfaces,” Phys. Rev. B, Condens.
[5] W. Tang et al., “MIMO transmission through Feb. 2013, Art. no. 022114. Matter, vol. 92, no. 10, Sep. 2015,
reconfigurable intelligent surface: System design, [25] K. Chen et al., “A reconfigurable active Huygens’ Art. no. 100304.
analysis, and implementation,” IEEE J. Sel. Areas metalens,” Adv. Mater., vol. 29, no. 17, May 2017, [46] S. Taravati and G. V. Eleftheriades, “Full-duplex
Commun., vol. 38, no. 11, pp. 2683–2699, Art. no. 1606422. nonreciprocal beam steering by time-modulated
Nov. 2020. [26] X. G. Zhang et al., “Light-controllable digital phase-gradient metasurfaces,” Phys. Rev. A, Gen.
[6] T. J. Cui, M. Q. Qi, X. Wan, J. Zhao, and Q. Cheng, coding metasurfaces,” Adv. Sci., vol. 5, no. 11, Phys., vol. 14, no. 1, Jul. 2020, Art. no. 014027.
“Coding metamaterials, digital metamaterials and Nov. 2018, Art. no. 1801028. [47] M. D. Renzo et al., “Smart radio environments
programmable metamaterials,” Light, Sci. Appl., [27] X. G. Zhang et al., “An optically driven digital empowered by reconfigurable AI meta-surfaces:
vol. 3, no. 10, p. e218, Oct. 2014. metasurface for programming electromagnetic An idea whose time has come,” EURASIP J.
[7] C. Huang, B. Sun, W. Pan, J. Cui, X. Wu, and functions,” Nature Electron., vol. 3, no. 3, Wireless Commun. Netw., vol. 2019, no. 1, p. 129,
X. Luo, “Dynamical beam manipulation based on pp. 165–171, Mar. 2020. May 2019.
2-bit digitally-controlled coding metasurface,” Sci. [28] T. J. Cui, S. Liu, and L. Zhang, “Information [48] L. Subrt and P. Pechac, “Intelligent walls as
Rep., vol. 7, no. 1, pp. 1–8, Feb. 2017. metamaterials and metasurfaces,” J. Mater. Chem. autonomous parts of smart indoor environments,”
[8] T.-T. Kim et al., “Amplitude modulation of C, vol. 5, no. 15, pp. 3644–3668, 2017. IET Commun., vol. 6, no. 8, pp. 1004–1010,
anomalously refracted terahertz waves with [29] T. J. Cui et al., “Information metamaterial May 2012.
gated-graphene metasurfaces,” Adv. Opt. Mater., systems,” iScience, vol. 23, no. 8, Aug. 2020, [49] P. del Hougne, M. Fink, and G. Lerosey, “Optimally
vol. 6, no. 1, Jan. 2018, Art. no. 1700507. Art. no. 101403. diverse communication channels in disordered
[9] Y. B. Li et al., “Transmission-type 2bit [30] L. Zhang et al., “Dynamically realizing arbitrary environments with tuned randomness,” Nature
programmable metasurface for single-sensor and multi-bit programmable phases using a 2-bit Electron., vol. 2, no. 1, pp. 36–41, 2019.
single-frequency microwave imaging,” Sci. Rep., time-domain coding metasurface,” IEEE Trans. [50] C. Liaskos et al., “A novel communication
vol. 6, no. 1, Mar. 2016, Art. no. 23731. Antennas Propag., vol. 68, no. 4, pp. 2984–2992, paradigm for high capacity and security via
[10] L. Li et al., “Electromagnetic reprogrammable Apr. 2020. programmable indoor wireless environments in
coding-metasurface holograms,” Nature Commun., [31] B. Wu, A. Sutinjo, M. E. Potter, and next generation wireless systems,” Ad Hoc Netw.,
vol. 8, no. 1, p. 197, Aug. 2017. M. Okoniewski, “On the selection of the number vol. 87, pp. 1–16, May 2019.
[11] X. Cai, Z. Zhou, and T. H. Tao, “Programmable of bits to control a dynamic digital MEMS [51] M. Di Renzo and J. Song, “Reflection probability
vanishing multifunctional optics,” Adv. Sci., vol. 6, reflectarray,” IEEE Antennas Wireless Propag. Lett., in wireless networks with metasurface-coated
no. 4, Feb. 2019, Art. no. 1801746. vol. 7, pp. 183–186, 2008. environmental objects: An approach based on
[12] S. Liu and T. J. Cui, “Concepts, working principles, [32] C. R. Dietlein, A. S. Hedden, and D. A. Wikner, random spatial processes,” EURASIP J. Wireless
and applications of coding and programmable “Digital reflectarray considerations for terrestrial Commun. Netw., vol. 2019, no. 1, p. 99, Apr. 2019.
metamaterials,” Adv. Opt. Mater., vol. 5, no. 22, millimeter-wave imaging,” IEEE Antennas Wireless [52] W. Khawaja, O. Ozdemir, Y. Yapici, F. Erden, and
Nov. 2017, Art. no. 1700624. Propag. Lett., vol. 11, pp. 272–275, 2012. I. Guvenc, “Coverage enhancement for NLOS
[13] D. F. Sievenpiper, J. H. Schaffner, H. J. Song, [33] Q. Wu and R. Zhang, “Beamforming optimization mmWave links using passive reflectors,” IEEE
R. Y. Loo, and G. Tangonan, “Two-dimensional for intelligent reflecting surface with discrete Open J. Commun. Soc., vol. 1, pp. 263–281, 2020.
beam steering using an electrically tunable phase shifts,” in Proc. IEEE Int. Conf. Acoust., [53] E. Basar, “Reconfigurable intelligent surfaces for
impedance surface,” IEEE Trans. Antennas Propag., Speech Signal Process. (ICASSP), May 2019, Doppler effect and multipath fading mitigation,”
vol. 51, no. 10, pp. 2713–2722, Oct. 2003. pp. 7830–7833. Frontiers Commun. Netw., vol. 2, May 2021,
[14] L. Li and T. J. Cui, “Information [34] J. Zhao et al., “Programmable time-domain Art. no. 672857.
metamaterials—From effective media to real-time digital-coding metasurface for non-linear [54] H. Zhang, L. Song, Z. Han, and H. V. Poor, “Spatial
information processing systems,” Nanophotonics, harmonic manipulation and new wireless equalization before reception: Reconfigurable
vol. 8, no. 5, pp. 703–724, Mar. 2019. communication systems,” Nat. Sci. Rev., vol. 6, intelligent surfaces for multi-path mitigation,” in
[15] S. Hu, F. Rusek, and O. Edfors, “Beyond massive no. 2, pp. 231–238, Mar. 2019. Proc. IEEE Int. Conf. Acoust., Speech Signal Process.
MIMO: The potential of data transmission with [35] L. Zhang et al., “Space-time-coding digital (ICASSP), Jun. 2021, pp. 8062–8066.
large intelligent surfaces,” IEEE Trans. Signal metasurfaces,” Nature Commun., vol. 9, no. 1, [55] R. Zhou et al., “Modeling and measurements for
Process., vol. 66, no. 10, pp. 2746–2758, p. 4334, 2018. multi-path mitigation with reconfigurable
May 2018. [36] A. Tennant and B. Chambers, “Time-switched intelligent surfaces,” in Proc. 16th Eur. Conf.
[16] Q. Wu and R. Zhang, “Intelligent reflecting array analysis of phase-switched screens,” IEEE Antennas Propag. (EuCAP), 2022, pp. 1–5.
surface enhanced wireless network via joint active Trans. Antennas Propag., vol. 57, no. 3, [56] C. Huang et al., “Holographic MIMO surfaces for
and passive beamforming,” IEEE Trans. Wireless pp. 808–812, Mar. 2009. 6G wireless networks: Opportunities, challenges,
Commun., vol. 18, no. 11, pp. 5394–5409, [37] W. H. Kummer, A. T. Villeneuve, T. Fong, and and trends,” IEEE Wireless Commun., vol. 27,
Nov. 2019. F. Terrio, “Ultra-low sidelobes from no. 5, pp. 118–125, Oct. 2020.
[17] Y. Han, W. Tang, S. Jin, C. Wen, and X. Ma, “Large time-modulated arrays,” IEEE Trans. Antennas [57] L. Wei, C. Huang, G. C. Alexandropoulos, C. Yuen,
intelligent surface-assisted wireless Propag., vol. AP-11, no. 6, pp. 633–639, Z. Zhang, and M. Debbah, “Channel estimation for
communication exploiting statistical CSI,” IEEE Nov. 1963. RIS-empowered multi-user MISO wireless
Trans. Veh. Technol., vol. 68, no. 8, [38] L. Zhang et al., “Breaking reciprocity with communications,” IEEE Trans. Commun., vol. 69,
pp. 8238–8242, Jun. 2019. space-time-coding digital metasurfaces,” Adv. no. 6, pp. 4144–4157, Mar. 2021.
[18] C. Liaskos, S. Nie, A. Tsioliaridou, A. Pitsillides, Mater., vol. 31, Oct. 2019, Art. no. 1904069. [58] Z. Wang, L. Liu, and S. Cui, “Channel estimation
S. Ioannidis, and I. Akyildiz, “A new wireless [39] J. Y. Dai, J. Zhao, Q. Cheng, and T. J. Cui, for intelligent reflecting surface assisted multiuser
communication paradigm through “Independent control of harmonic amplitudes and communications: Framework, algorithms, and
software-controlled metasurfaces,” IEEE Commun. phases via a time-domain digital coding analysis,” IEEE Trans. Wireless Commun., vol. 19,
Mag., vol. 56, no. 9, pp. 162–169, Sep. 2018. metasurface,” Light, Sci. Appl., vol. 7, no. 1, p. 90, no. 10, pp. 6607–6620, Oct. 2020.
[19] R. Heath, S. Peters, Y. Wang, and J. Zhang, Nov. 2018. [59] J. Y. Dai et al., “Wireless communications through
“A current perspective on distributed antenna [40] J. Y. Dai et al., “High-efficiency synthesizer for a simplified architecture based on time-domain
systems for the downlink of cellular systems,” spatial waves based on space-time-coding digital digital coding metasurface,” Adv. Mater. Technol.,
IEEE Commun. Mag., vol. 51, no. 4, pp. 161–167, metasurface,” Laser Photon. Rev., vol. 14, no. 6, vol. 4, no. 7, Jul. 2019, Art. no. 1900044.
Apr. 2013. Jun. 2020, Art. no. 1900133. [60] W. K. Tang et al., “Programmable
[20] S. Hu, F. Rusek, and O. Edfors, “The potential of [41] J. Y. Dai et al., “Arbitrary manipulations of dual metasurface-based RF chain-free 8PSK wireless
positioning with large intelligent surfaces,” IEEE harmonics and their wave behaviors based on transmitter,” Electron. Lett., vol. 55, no. 7,
Trans. Signal Process., vol. 66, no. 7, space-time-coding digital metasurface,” Appl. pp. 417–419, Apr. 2019.
pp. 1761–1774, Apr. 2018. Phys. Rev., vol. 7, no. 4, Dec. 2020, [61] E. Basar, “Transmission through large intelligent
[21] J. Y. Lau and S. V. Hum, “A wideband Art. no. 041408. surfaces: A new frontier in wireless
reconfigurable transmitarray element,” IEEE [42] H. Rajabalipanah, A. Abdolali, S. Iqbal, L. Zhang, communications,” in Proc. Eur. Conf. Netw.
Trans. Antennas Propag., vol. 60, no. 3, and T. J. Cui, “Analog signal processing through Commun. (EuCNC), Jun. 2019, pp. 112–117.
pp. 1303–1311, Mar. 2012. space-time digital metasurfaces,” Nanophotonics, [62] J. Y. Dai et al., “Realization of multi-modulation
[22] L. Boccia, I. Russo, G. Amendola, and G. D. Massa, vol. 10, no. 6, pp. 1753–1764, Apr. 2021. schemes for wireless communication by
“Multilayer antenna-filter antenna for [43] D. L. Sounas, C. Caloz, and A. Alù, “Giant time-domain digital coding metasurface,” IEEE
beam-steering transmit-array applications,” IEEE non-reciprocity at the subwavelength scale using Trans. Antennas Propag., vol. 68, no. 3,
Trans. Microw. Theory Techn., vol. 60, no. 7, angular momentum-biased metamaterials,” pp. 1618–1627, May 2020.
pp. 2287–2300, Jul. 2012. Nature Commun., vol. 4, p. 2407, Sep. 2013. [63] L. Zhang et al., “A wireless communication scheme
[23] I. V. Shadrivov, P. V. Kapitanova, S. I. Maslovski, [44] N. A. Estep, D. L. Sounas, J. Soric, and A. Alù, based on space- and frequency-division
and Y. S. Kivshar, “Metamaterials controlled with “Magnetic-free non-reciprocity and isolation based multiplexing using digital metasurfaces,” Nature
light,” Phys. Rev. Lett., vol. 109, no. 8, Aug. 2012, on parametrically modulated coupled-resonator Electron., vol. 4, no. 3, pp. 218–227, Mar. 2021.
Art. no. 083902. loops,” Nature Phys., vol. 10, no. 12, pp. 923–927, [64] M. Z. Chen et al., “Accurate and broadband

1286 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

manipulations of harmonic amplitudes and phases metasurfaces: Principles and applications,” [102] M. M. Salary, S. Jafar-Zanjani, and H. Mosallaei,
to reach 256 QAM millimeter-wave wireless Research, vol. 2021, pp. 1–25, May 2021. “Time-varying metamaterials based on
communications by time-domain digital coding [83] L. Zhang, J. Y. Dai, M. Moccia, G. Castaldi, graphene-wrapped microwires: Modeling and
metasurface,” Nat. Sci. Rev., vol. 9, no. 1, T. J. Cui, and V. Galdi, “Recent advances and potential applications,” Phys. Rev. B, vol. 97,
Jan. 2022, Art. no. nwab134. perspectives on space-time coding digital no. 11, Mar. 2018, Art. no. 115421.
[65] Q. Ma and T. J. Cui, “Information metamaterials: metasurfaces,” EPJ Appl. Metamater., vol. 7, p. 7, [103] Y. Zhao et al., “Dynamic photoinduced controlling
Bridging the physical world and digital world,” Dec. 2020. of the large phase shift of terahertz waves via
PhotoniX, vol. 1, no. 1, pp. 1–32, Dec. 2020. [84] A. Shaltout, A. Kildishev, and V. Shalaev, vanadium dioxide coupling nanostructures,” ACS
[66] J. Y. Dai et al., “Wireless communication based on “Time-varying metasurfaces and Lorentz Photon., vol. 5, no. 8, pp. 3040–3050, Aug. 2018.
information metasurfaces,” IEEE Trans. Microw. non-reciprocity,” Opt. Mater. Exp., vol. 5, no. 11, [104] L. Bao et al., “Design of digital coding
Theory Techn., vol. 69, no. 3, pp. 1493–1510, p. 2459, 2015. metasurfaces with independent controls of phase
Mar. 2021. [85] G. Castaldi et al., “Joint multi-frequency beam and amplitude responses,” Appl. Phys. Lett.,
[67] S. Henthorn, K. L. Ford, and T. O’Farrell, shaping and steering via space–time-coding digital vol. 113, no. 6, Aug. 2018, Art. no. 063502.
“Bit-error-rate performance of quadrature metasurfaces,” Adv. Funct. Mater., vol. 31, no. 6, [105] H. Wu et al., “Controlling energy radiations of
modulation transmission using reconfigurable Feb. 2021, Art. no. 2007620. electromagnetic waves via frequency coding
frequency selective surfaces,” IEEE Antennas [86] W. Tang et al., “Wireless communications with metamaterials,” Adv. Sci., vol. 4, no. 9, Sep. 2017,
Wireless Propag. Lett., vol. 16, pp. 2038–2041, programmable metasurface: Transceiver design Art. no. 1700098.
2017. and experimental results,” China Commun., [106] T. Qiu et al., “Deep learning: A rapid and efficient
[68] S. Henthorn, K. L. Ford, and T. O’Farrell, “Direct vol. 16, no. 5, pp. 46–61, May 2019. route to automatic metasurface design,” Adv. Sci.,
antenna modulation for high-order phase shift [87] W. Tang et al., “Wireless communications with vol. 6, no. 12, Jun. 2019, Art. no. 1900128.
keying,” IEEE Trans. Antennas Propag., vol. 68, programmable metasurface: New paradigms, [107] Q. Zhang et al., “Machine-learning designs of
no. 1, pp. 111–120, Jan. 2020. opportunities, and challenges on transceiver anisotropic digital coding metasurfaces,” Adv.
[69] G. I. Kiani and M. A. Hussaini, “Waveguide ASK design,” IEEE Wireless Commun., vol. 27, no. 2, Theory Simul., vol. 2, no. 2, Feb. 2019,
modulator using switchable FSS,” J. Infr., Millim., pp. 180–187, Apr. 2020. Art. no. 1800132.
THz Waves, vol. 41, no. 12, pp. 1478–1487, [88] W. Tang et al., “Realization of reconfigurable [108] J. Wang et al., “Interplay between RIS and AI in
Dec. 2020. intelligent surface-based Alamouti space-time wireless communications: Fundamentals,
[70] C. X. Huang, J. Zhang, Q. Cheng, and T. J. Cui, transmission,” in Proc. Int. Conf. Wireless Commun. architectures, applications, and open research
“Polarization modulation for wireless Signal Process. (WCSP), Oct. 2020, pp. 904–909. problems,” IEEE J. Sel. Areas Commun., vol. 39,
communications based on metasurfaces,” Adv. [89] L. Lu, G. Y. Li, A. L. Swindlehurst, A. Ashikhmin, no. 8, pp. 2271–2288, Aug. 2021.
Funct. Mater., vol. 31, no. 36, Sep. 2021, and R. Zhang, “An overview of massive MIMO: [109] G. Gradoni and M. Di Renzo, “End-to-end mutual
Art. no. 2103379. Benefits and challenges,” IEEE J. Sel. Topics Signal coupling aware communication model for
[71] S. Liu et al., “Convolution operations on coding Process., vol. 8, no. 5, pp. 742–758, Oct. 2014. reconfigurable intelligent surfaces: An
metasurface to reach flexible and continuous [90] T. J. Cui, S. Liu, G. D. Bai, and Q. Ma, “Direct electromagnetic-compliant approach based on
controls of terahertz beams,” Adv. Sci., vol. 3, transmission of digital message via programmable mutual impedances,” IEEE Wireless Commun. Lett.,
no. 10, Jul. 2016, Art. no. 1600156. coding metasurface,” Research, vol. 2019, vol. 10, no. 5, pp. 938–942, May 2021.
[72] R. Y. Wu, C. B. Shi, S. Liu, W. Wu, and T. J. Cui, Jan. 2019, Art. no. 2584509. [110] X. Qian and M. D. Renzo, “Mutual coupling and
“Addition theorem for digital coding [91] A. K. Khandani et al., “Media-based modulation: A unit cell aware optimization for reconfigurable
metamaterials,” Adv. Opt. Mater., vol. 6, no. 5, new approach to wireless transmission,” in Proc. intelligent surfaces,” IEEE Wireless Commun. Lett.,
Jan. 2018, Art. no. 1701236. IEEE Int. Symp. Inf. Theory, Jul. 2013, pp. vol. 10, no. 6, pp. 1183–1187, Jun. 2021.
[73] M. Moccia et al., “Coding metasurfaces for diffuse 3050–3054. [111] X. Chen et al., “Design and implementation of
scattering: Scaling laws, bounds, and suboptimal [92] E. Basar, “Media-based modulation for future MIMO transmission based on dual-polarized
design,” Adv. Opt. Mater., vol. 5, no. 19, Jul. 2017, wireless systems: A tutorial,” IEEE Wireless reconfigurable intelligent surface,” IEEE Wireless
Art. no. 1700455. Commun., vol. 26, no. 5, pp. 160–166, Jul. 2019. Commun. Lett., vol. 10, no. 10, pp. 2155–2159,
[74] M. Liu, D. A. Powell, Y. Zarate, and I. V. Shadrivov, [93] X. Wan et al., “Multichannel direct transmissions Oct. 2021.
“Huygens’ metadevices for parametric waves,” of near-field information,” Light, Sci. Appl., vol. 8, [112] S. Abeywickrama, R. Zhang, Q. Wu, and C. Yuen,
Phys. Rev. X, vol. 8, no. 3, Sep. 2018, no. 1, pp. 1–8, Jul. 2019. “Intelligent reflecting surface: Practical phase shift
Art. no. 031077. [94] H. Zhao, Y. Shuang, M. Wei, T. J. Cui, model and beamforming optimization,” IEEE
[75] X. Gao, W. L. Yang, H. F. Ma, Q. Cheng, X. H. Yu, P. D. Hougne, and L. Li, “Metasurface-assisted Trans. Commun., vol. 68, no. 9, pp. 5849–5863,
and T. J. Cui, “A reconfigurable broadband massive backscatter wireless communication with Sep. 2020.
polarization converter based on an active commodity Wi-Fi signals,” Nature Commun., [113] W. Chen, L. Bai, W. Tang, S. Jin, W. X. Jiang, and
metasurface,” IEEE Trans. Antennas Propag., vol. 11, no. 1, pp. 1–10, Dec. 2020. T. J. Cui, “Angle-dependent phase shifter model
vol. 66, no. 11, pp. 6086–6095, Nov. 2018. [95] S. Nie and I. F. Akyildiz, “Metasurfaces for for reconfigurable intelligent surfaces: Does the
[76] S. Liu et al., “Anisotropic coding metamaterials multiplexed communication,” Nature Electron., angle-reciprocity hold?” IEEE Commun. Lett.,
and their powerful manipulation of differently vol. 4, no. 3, pp. 177–178, Mar. 2021. vol. 24, no. 9, pp. 2060–2064, Sep. 2020.
polarized terahertz waves,” Light, Sci. Appl., [96] Q. Wu and R. Zhang, “Towards smart and [114] Z. Wang, L. Tan, H. Yin, K. Wang, X. Pei, and
vol. 5, no. 5, May 2016, Art. no. e16076. reconfigurable environment: Intelligent reflecting D. Gesbert, “A received power model for
[77] Z. L. Mei, X. M. Ma, C. Lu, and Y. D. Zhao, surface aided wireless network,” IEEE Commun. reconfigurable intelligent surface and
“High-efficiency and wide-bandwidth linear Mag., vol. 58, no. 1, pp. 106–112, Jan. 2020. measurement-based validations,” 2021,
polarization converter based on double U-shaped [97] W. Tang et al., “Path loss modeling and arXiv:2105.06082.
metasurface,” AIP Adv., vol. 7, no. 12, Dec. 2017, measurements for reconfigurable intelligent [115] H. Zhao, Y. Shuang, M. Wei, T. J. Cui,
Art. no. 125323. surfaces in the millimeter-wave frequency band,” P. D. Hougne, and L. Li, “Metasurface-assisted
[78] F. Li et al., “Dual-band reflective polarization 2021, arXiv:2101.08607. massive backscatter wireless communication with
converter based on slotted wire resonators,” Appl. [98] W. H. Gao et al., “1-bit reconfigurable commodity Wi-Fi signals,” Nature Commun.,
Phys. B, Lasers Opt., vol. 124, no. 2, p. 28, transmitarray with low loss and wide bandwidth,” vol. 11, no. 1, pp. 1–10, Aug. 2020.
Jan. 2018. New J. Phys., vol. 23, no. 6, Jun. 2021, [116] S. Yu, H. Liu, and L. Li, “Design of near-field
[79] J. C. Ke et al., “Linear and nonlinear polarization Art. no. 065006. focused metasurface for high-efficient wireless
syntheses and their programmable controls based [99] S. J. Li et al., “Programmable controls to scattering power transfer with multifocus characteristics,”
on anisotropic time-domain digital coding properties of a radiation array,” Laser Photon. Rev., IEEE Trans. Ind. Electron., vol. 66, no. 5,
metasurface,” Small Struct., vol. 2, no. 1, vol. 15, no. 2, Feb. 2021, Art. no. 2000449. pp. 3993–4002, May 2019.
Jan. 2021, Art. no. 2170003. [100] D. R. Smith, O. Yurduseven, L. P. Mancera, [117] H. Zhang et al., “MetaRadar: Indoor localization
[80] C. Zhang et al., “Convolution operations on P. Bowen, and N. B. Kundtz, “Analysis of a by reconfigurable metamaterials,” IEEE Trans.
time-domain digital coding metasurface for beam waveguide-fed metasurface antenna,” Phys. Rev. Mobile Comput., early access, Dec. 14, Dec. 2020,
manipulations of harmonics,” Nanophotonics, A, Gen. Phys., vol. 8, no. 5, Nov. 2017, doi: 10.1109/TMC.2020.3044603.
vol. 9, no. 9, pp. 2771–2781, Jul. 2020. Art. no. 054048. [118] H. T. Wu et al., “Information theory of
[81] N. Yu et al., “Light propagation with phase [101] T. Qiu, Y. Jia, J. Wang, Q. Cheng, and S. Qu, metasurfaces,” Nat. Sci. Rev., vol. 7, no. 3,
discontinuities: Generalized laws of reflection and “Controllable reflection-enhancement pp. 561–571, Mar. 2020.
refraction,” Science, vol. 334, no. 6054, metasurfaces via amplification excitation of [119] H. Wu et al., “Harmonic information transitions of
pp. 333–337, Oct. 2011. transistor circuit,” IEEE Trans. Antennas Propag., spatiotemporal metasurfaces,” Light, Sci. Appl.,
[82] L. Zhang and T. J. Cui, “Space-time-coding digital vol. 69, no. 3, pp. 1477–1482, Mar. 2021. vol. 9, no. 1, pp. 1–13, Dec. 2020.

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1287

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

ABOUT THE AUTHORS

Qiang Cheng (Senior Member, IEEE) Wankai Tang received the B.S., M.S., and
received the B.S. and M.S. degrees from the Ph.D. degrees from Southeast University,
Nanjing University of Aeronautics and Astro- Nanjing, China, in 2011, 2014, and 2021,
nautics, Nanjing, Jiangsu, China, in 2001 and respectively.
2004, respectively, and the Ph.D. degree He worked at National Instruments,
from Southeast University, Nanjing, in 2008. Shanghai, China, from 2014 to 2016. He is
In 2008, he joined the State Key currently with the National Mobile Commu-
Laboratory of Millimeter Waves, Southeast nications Research Laboratory, Southeast
University, where he was involved in the University. His research interests include
development of metamaterials and metadevices. He is currently a modeling and prototyping of reconfigurable intelligent surface
Full Professor with the Radio Department, Southeast University. He (RIS)-empowered wireless communication systems.
leads a group of Ph.D. students and master students in the areas Dr. Tang received the Electronics Letters Best Paper Award in
of metamaterials, tunable microwave circuits, microwave imaging, 2020 and the China Communications Best Paper Award in 2021.
and terahertz systems. He has authored or coauthored more than
100 publications, with citation over 2000 times.
Dr. Cheng was a recipient of the 2010 Best Paper Award from
the New Journal of Physics, the China’s Top Ten Scientific Advances
of 2010, and the Second Class National Natural Science Award Jun Chen Ke received the B.S. degree
in 2014. He served as the Vice Chair for the 2008 and 2010 from the School of Instrument Science and
International Workshop on Metamaterials, Nanjing. Engineering, Southeast University, Nanjing,
China, in 2016. He is currently working
toward the Ph.D. degree at the State Key
Laboratory of Millimeter Waves, Department
of Radio Engineering, Southeast University.
His research interests include metamate-
rials, metasurfaces, and array antennas.

Lei Zhang (Member, IEEE) received the

B.E. degree in electromagnetic wave prop-
agation and antenna from Xidian University,
Shuo Liu received the B.E. and
Xi’an, China, in 2015, and the Ph.D. degree
Ph.D. degrees in electromagnetism and
in electromagnetic field and microwave
microwave engineering from Southeast
technology from the State Key Laboratory
University, Nanjing, China, in 2010 and
of Millimeter Waves, Southeast University
2017, respectively.
(SEU), Nanjing, China, in 2020.
He is currently working at Southeast
He is currently a Postdoctoral Fellow with
University on metamaterials and topological
the State Key Laboratory of Millimeter Waves, SEU. He has authored
insulators. He is the author of two books
or coauthored over 40 peer-reviewed journal articles with more
on information metamaterials published
than 3000 citations. His current research interests include meta-
at Cambridge University Press and has published over 60 peer-
materials, antenna design, and space-time metasurfaces.
reviewed journal articles with more than 4300 citations, among
Dr. Zhang was a recipient of the Best Student Paper Award at the
which eight articles enter “top 1% ESI highly cited papers.”
6th Asia-Pacific Conference on Antennas and Propagation (APCAP)
Dr. Liu received the Marie-Curie Fellow, the Grand Prize of
in 2017.
Baosteel Student Scholarship, “25 Excellent Doctoral Degree
Dissertation” by the Chinese Institute of Electronics (CIE), and
Important Achievements of Optics in China in 2016. His works on
information metamaterials were nominated by the Optical Society
of America (OSA) as the top 30 best optical achievements in 2016
(optics in 2016).

Jun Yan Dai received the B.S. degree in

information engineering from Southeast Uni-
versity, Nanjing, China, in 2013, and the Jing Cheng Liang received the B.E. degree
Ph.D. degree in electromagnetic field and from the School of Electronic Science and
microwave technology from the State Key Engineering, Nanjing University of Posts
Laboratory of Millimeter Waves, Southeast and Telecommunications, Nanjing, China, in
University, in 2019. 2016, and the M.E. degree from the School
From February 2020 to December 2021, of Electronic Science and Engineering, Uni-
he was a Postdoctoral Fellow with the State versity of Electronic Science and Technology
Key Laboratory of Terahertz and Millimeter Waves, City University of of China, Chengdu, China, in 2019. He is
Hong Kong, Hong Kong. He is currently an Associate Professor with currently working toward the Ph.D. degree at
the State Key Laboratory of Millimeter Waves, Southeast University. the State Key Laboratory of Millimeter Waves, Southeast University,
His research interests include metasurfaces, reconfigurable intel- Nanjing.
ligent surfaces, space-time modulation technology, and wireless His current research interests include metasurfaces, reconfig-
communication systems. urable intelligent surface (RIS), and time-reversal method.

1288 P ROCEEDINGS OF THE IEEE | Vol. 110, No. 9, September 2022

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.
 Cheng et al.: Reconfigurable Intelligent Surfaces: Simplified-Architecture Transmitters

Shi Jin (Senior Member, IEEE) received the of Millimeter Waves, and the Founding Director of the Institute
B.S. degree in communications engineer- of Electromagnetic Space, Southeast University. He proposed the
ing from the Guilin University of Electronic concepts of digital coding and programmable metamaterials, and
Technology, Guilin, China, in 1996, the M.S. realized their first prototypes, based on which he founded the new
degree from the Nanjing University of Posts direction of information metamaterials, bridging the physical world
and Telecommunications, Nanjing, China, in and digital world. He is the first author of books: Metamaterials:
2003, and the Ph.D. degree in informa- Theory, Design, and Applications (Springer, November 2009); Meta-
tion and communications engineering from materials: Beyond Crystals, Noncrystals, and Quasicrystals (CRC
Southeast University, Nanjing, in 2007. Press, March 2016); and Information Metamaterials (Cambridge
From June 2007 to October 2009, he was a Research Fellow with University Press, 2021). He has published over 500 peer-reviewed
the Adastral Park Research Campus, University College London, journal articles, which have been cited by more than 44 600 times
London, U.K. He is currently with the Faculty of the National Mobile (H-factor 107; Google Scholar), and licensed over 100 patents.
Communications Research Laboratory, Southeast University. His His research interests include metamaterials and computational
research interests include space-time wireless communications, electromagnetics.
random matrix theory, information theory, intelligent communica- Dr. Cui was awarded a Research Fellowship from the Alexander
tions, and intelligent surfaces. von Humboldt Foundation, Bonn, Germany, in 1995. He received
Dr. Jin and his coauthors have been awarded the 2011 IEEE the Young Scientist Award from the International Union of Radio
Communications Society Stephen O. Rice Prize Paper Award in the Science in 1999; the Cheung Kong Professor by the Ministry of
field of communication theory and the 2010 Young Author Best Education, China, in 2001; the National Science Foundation of
Paper Award by the IEEE Signal Processing Society. He serves China for Distinguished Young Scholars in 2002; the Natural Science
as an Associate Editor for the IEEE TRANSACTIONS ON WIRELESS Award (First Class) from the Ministry of Education, China, in 2011;
COMMUNICATIONS, the IEEE COMMUNICATIONS LETTERS, and IET and the National Natural Science Awards of China (Second Class,
Communications. twice) in 2014 and 2018. His research has been selected as one of
the most exciting peer-reviewed optics research “Optics in 2016”
Tie Jun Cui (Fellow, IEEE) received the by Optics and Photonics News Magazine, 10 Breakthroughs of
B.Sc., M.Sc., and Ph.D. degrees in elec- China Science in 2010, and many research highlights in a series
trical engineering from Xidian University, of journals. His work has been widely reported by Nature News, MIT
Xi’an, China, in 1987, 1990, and 1993, Technology Review, Scientific American, New Scientists, Discover,
respectively. and so on. He is the Academician of the Chinese Academy of Sci-
In March 1993, he joined the Department ence. He served as an Associate Editor for IEEE TRANSACTIONS ON
of Electromagnetic Engineering, Xidian Uni- GEOSCIENCE AND REMOTE SENSING and a Guest Editor of Science
versity, where he was promoted to an Asso- China Information Sciences, Science Bulletin, IEEE JOURNAL ON
ciate Professor in November 1993. From EMERGING AND SELECTED TOPICS IN CIRCUITS AND SYSTEMS, Engi-
1995 to 1997, he was a Research Fellow with the Institut fur neering, and Research. He is also the Chief Editor of Metamaterial
Hochstfrequenztechnik und Elektronik (IHE), University of Karl- Short Book Series in Cambridge University Press, an Editor of
sruhe, Karlsruhe, Germany. In July 1997, he joined the Center Materials Today Electronics, an Associate Editor of Research, and
for Computational Electromagnetics, Department of Electrical and an Editorial Board Member of National Science Review, eLight, Pho-
Computer Engineering, University of Illinois at Urbana–Champaign, toniX, Advanced Optical Materials, Small Structures, and Advanced
Champaign, IL, USA, first as a Postdoctoral Research Associate and Photonics Research. He presented more than 100 keynote and ple-
then as a Research Scientist. In September 2001, he was a Cheung- nary talks in academic conferences, symposiums, and workshops.
Kong Professor with the Department of Radio Engineering, South- From 2019 to 2021, he was ranked in the top 1% for the highly cited
east University, Nanjing, China. He is currently a Chief Professor papers in the field of physics by Clarivate Web of Science (Highly
with Southeast University, the Director of the State Key Laboratory Cited Researcher).

Vol. 110, No. 9, September 2022 | P ROCEEDINGS OF THE IEEE 1289

Authorized licensed use limited to: Gyeongsang National Univ. Downloaded on November 22,2023 at 06:47:41 UTC from IEEE Xplore. Restrictions apply.