1

Reconfigurable Intelligent Surfaces: Potentials, Applications,

and Challenges for 6G Wireless Networks
Sarah Basharat, Syed Ali Hassan, Haris Pervaiz, Aamir Mahmood, Zhiguo Ding, and Mikael Gidlund

Abstract—Reconfigurable intelligent surfaces (RISs), with the potential holographic MIMO surface (HMIMOS) [15], has been proposed
to realize smart radio environment, have emerged as an energy-efficient in this regard. RIS empowers the smart wireless environments by
and a cost-effective technology to support the services and demands
overcoming the stochastic nature of the propagation channel, thereby
foreseen for coming decades. By leveraging a large number of low-cost
passive reflecting elements, RISs introduce a phase-shift in the impinging improving quality-of-service (QoS) and connectivity. The wireless
signal to create a favorable propagation channel between the transmitter environment that used to be a dynamic uncontrollable factor is now
and the receiver. In this article, we provide a tutorial overview of RISs considered to be a part of the network design parameter.
for sixth-generation (6G) wireless networks. Specifically, we present a Specifically, an RIS is a software-controlled planer surface con-
arXiv:2107.05460v1 [cs.IT] 12 Jul 2021

comprehensive discussion on performance gains that can be achieved

by integrating RISs with emerging communication technologies. We sisting of a large number of low-cost passive reflecting elements.
address the practical implementation of RIS-assisted networks and expose Each element, of size smaller than the wavelength, has the capability
the crucial challenges, including the RIS reconfiguration, deployment to alter the phase of the impinging signal, creating a favorable
and size optimization, and channel estimation. Furthermore, we explore wireless environment between the transmitter and the receiver. The
the integration of RIS and non-orthogonal multiple access (NOMA)
RIS reflection adaption is programmed and controlled via a smart
under imperfect channel state information (CSI). Our numerical results
illustrate the importance of better channel estimation in RIS-assisted controller, such as a field-programmable gate array (FPGA), which
networks and indicate the various factors that impact the size of RIS. acts as a gate-way to communicate and coordinate with the BS
Finally, we present promising future research directions for realizing through a separate wireless or a wired link. In particular, RIS
RIS-assisted networks in 6G communication. receives a signal from the base station (BS), and then reflects the
incident signal by inducing phase changes, adjusted by the controller.
I. I NTRODUCTION Consequently, the reflected signal can be added coherently with the
direct signal from the BS to either boost or attenuate the overall signal

T HE global mobile traffic volume is anticipated to reach 5016

exabytes per month (Eb/mo) in 2030, which was 7.462 EB/mo
in 2010 [1]. This clearly depicts the importance of the evolution
strength at the receiver. Although RIS is theoretically passive, since it
reflects the signal without power amplification, however, in practice,
it has minimal power requirement for the operation of smart controller
and advancement of mobile communication technologies. To date, and reconfiguration of the elements for controllable reflections.
fifth-generation (5G) communication, which is expected to realize As illustrated in Fig. 1, RIS concept can be viewed to operate
the targeted 1000x increase in network capacity with new and similarly as other related wireless technologies such as, conventional
advanced services, is being rolled out in the world. However, 5G relaying, backscatter communication (BackCom), and mMIMO relay-
systems will not be able to fully support the growing demand for ing. We now present the major differences and competitive strengths
wireless communication in 2030. The core 5G technologies include of RIS that make it stand out among these technologies. First,
massive multiple-input multiple-output (mMIMO) and millimeter- compared to conventional relaying that requires additional power for
wave (mmWave) communications. The mMIMO technology exploits signal transmission, amplification, and regeneration, RIS passively
the spatial domain by deploying numerous antennas to enable parallel reflects the impinging signal by inducing intelligent phase-shifts,
transmission to multiple users using the same frequency-time resource without the need for an additional radio frequency (RF) source.
block. The mmWaves, on the other hand, offer plenteous spare Moreover, RIS operates in full-duplex (FD) mode, free from noise
spectrum in high frequency bands, thus resolving spectrum scarcity amplification and self-interference. Secondly, compared to traditional
issues at microwave frequencies. Although mMIMO and mmWaves BackCom such as radio frequency identification (RFID) tag that loads
significantly improve spectral efficiency (SE), high hardware cost its own information on the incident signal and then backscatters the
and complexity are major hurdles in their practical implementation. modulated signal to the receiver, RIS reflects the incident signal
Moreover, mmWaves are highly vulnerable to signal blockage and to assist the communication between the source and the receiver
attenuation. Therefore, reliable and efficient communication is still without sending any information of its own. BackCom also requires
not guaranteed. sophisticated signal processing for self-interference cancellation in
While 5G is yet to be realized fully, the researchers have already order to decode the tag message. Finally, unlike mMIMO relaying,
started looking for energy and spectral-efficient solution for sixth- RIS can be implemented at a much low hardware cost and power
generation (6G) systems. In addition to the energy and spectral consumption. Although signal-to-noise ratio (SNR) achieved through
efficiency, the new paradigm is smart and reconfigurable wireless RIS is less than the equal-sized mMIMO counterpart [6], however,
environments [2]. Recently, a cost-effective and energy-efficient the SNR of RIS-assisted system can be improved by increasing the
technology, reconfigurable intelligent surface (RIS) [3]–[5], also reflecting elements, since the cost per reflecting element of RIS is
called intelligent reflecting surface (IRS) [2], [6]–[14], or passive much less than the cost per antenna element in mMIMO relaying.
S. Basharat and S. A. Hassan are with the School of Electrical Engineering Architecturally, RIS is lightweight with conformal geometry and can
and Computer Science (SEECS), National University of Sciences and Tech- be easily mounted on the ceilings, walls, and building facades.
nology (NUST), Islamabad 44000, Pakistan. Inspired by the RIS potential to realize smart wireless environments
H. Pervaiz is with the School of Computing and Communications, Lancaster and its compatibility with other technologies, the main contributions
University, Lancaster LA1 4YW, U.K. of this study can be summarized as follows.
A. Mahmood and M. Gidlund are with Mid Sweden University, Sweden.
Z. Ding is with the School of Electrical and Electronic Engineering, The • We provide a comprehensive discussion on integrating RIS with
University of Manchester, Manchester M13 9PL, U.K. emerging communication technologies for realizing 6G wireless
 2

RIS self-interference
(FD mode)
Relay
Passive Active
Energy Efficient Low energy
Cost Effective consumption
Free from self- Low hardware cost Source%3Cm
Reflects signal interference Suffers from noise
` amplification &
Receives/transmits signal
self-interference (FD
mode) Receiver

RIS reflected link

Direct link Source-Relay-Receiver link
RIS control link Direct link
RIS

RIS aided Communication Conventional Relaying

RIS controller

self-interference
MIMO Relay BackCom Relay
(FD mode) Active Passive
High energy Energy efficient
consumption Cost effective
High hardware cost Short operating
Suffers from noise range MIMO relay
Reflects signal
amplification & Suffers from self-
Receives/transmits signal self-interference (FD interference
mode)
Backscatter tag

Source-MIMO Relay-Receiver link Backscatter reflected link

Direct link Direct link

Massive MIMO Relaying Backscatter Communication

Fig. 1: RIS comparison with existing related technologies.

networks, namely, non-orthogonal multiple access (NOMA), (e.g., time, frequency, code) block, hence improves both spectral
simultaneous wireless information and power transfer (SWIPT), and energy efficiency. In a downlink NOMA system, superposition
unmanned aerial vehicles (UAVs), BackCom, mmWaves, and coding (SC) is used at the BS to multiplex the data of the users
multi-antenna systems. with different channel gains, and successive interference cancellation
• For practical implementation of RIS-assisted networks, we iden- (SIC) is employed at the receiver to decode the message signals. Al-
tify three crucial challenges, including RIS reconfiguration for though NOMA provides sufficient performance gains over OMA, the
controllable reflections, deployment and size optimization, and stringent demands on data rate and connectivity for B5G/6G systems
channel estimation. compel to shift to smart and reconfigurable wireless networks; hence,
• We present a novel case study for RIS-assisted NOMA network the RIS-assisted NOMA system has been proposed by the research
with imperfect channel state information (CSI) to highlight the community.
impact of channel estimation errors on the performance of RIS- To fully exploit the gains of the RIS-assisted NOMA system,
assisted NOMA networks. We further determine the various appropriate phase-shift matrix and beamforming vectors need to
factors that affect the size of RIS, i.e., the number of RIS be designed. The two types of phase-shifting designs, coherent
elements. phase-shifting and random phase-shifting achieve different trade-offs
• To provide effective guidance for future research, we introduce between performance and complexity for an RIS-assisted NOMA
five promising research directions for realizing RIS-assisted system [7]. For coherent phase-shifting, each RIS element introduces
networks. a phase that matches the phase of fading channels from RIS-to-
BS and users. Despite the superior performance, it might not be
II. I NTEGRATING RIS WITH EMERGING COMMUNICATION feasible to implement coherent shifting design because of hardware
TECHNOLOGIES limitations of phase shifters and excess system overhead of acquiring
CSI. Although the random phase-shifting design reduces the system
Current research contributions have revealed RIS to be a cutting-
complexity and overhead of acquiring CSI, it degrades the system per-
edge technology, opening new promising research opportunities on
formance. Furthermore, the RIS phase-shift matrix and beamforming
the road towards 6G. In this section, we elaborate on the perfor-
vectors can be jointly optimized to minimize the transmit power of
mance gains that can be achieved by integrating RIS with emerging
the BS using semi-definite relaxation (SDR). However, for large-scale
communication technologies.
RIS-assisted networks, the computational complexity of the SDR
technique is extremely high and the probability of returning rank-
A. RIS and NOMA one optimum solution is extremely small. Therefore, a difference-of-
NOMA has emerged as a promising technology for future genera- convex (DC) method has been proposed that overcomes the limitation
tion networks to support massive connectivity. Power domain NOMA of the SDR method and outperforms in terms of minimizing the BS
(PD-NOMA) enables multiple users to share the same resource transmit power [8].
 3

B. RIS and SWIPT

SWIPT is an effective solution to power massive devices in
RIS
a wireless-powered Internet-of-things (IoT) network. In practice,
the significant power loss over long distances reduces the energy
harvested at the energy receiver, which degrades the performance
of SWIPT systems. However, the limitations of practical SWIPT
systems can be compensated via RIS, as illustrated in Fig. 2. Through Reflected IF
intelligent signal reflections, RIS boosts the signal strength both at Direct IF
the information receiver (IR) and the energy receiver (ER), thereby
improving the energy efficiency of SWIPT system. Reflected EF
Recently, [9] unleashes the benefits of RIS-assisted SWIPT net- BS IRs
work, where a multi-antenna BS serves several multi-antenna in- ERs
formation users, while satisfying the energy requirements of the EF (Energy flow)
IF (Information flow Direct EF
energy users. The authors proposed a weighted sum-rate maximiza-
tion problem for the joint optimization of BS’s transmit precoding
matrices and RIS phase-shifts. The proposed problem is challenging RIS-assisted SWIPT system
to solve for optimal solution, owing to highly coupled optimization
parameters. Therefore, the authors developed an iterative solution
using a block coordinate descent (BCD) algorithm that converges Controllable UAV-RIS link RIS
rapidly and outperforms the benchmark approaches, i.e., fixed RIS
phase-shifts and conventional SWIPT networks without RIS. Such a UAV
performance is quite appealing for practical applications. Blocked Direct link

C. RIS and UAVs

` Controllable
RIS can even be applied to UAV-enabled communication systems RIS-UE link
to improve propagation environment and enhance communication
quality, as shown in Fig. 2. In a dense urban environment, the line-
of-sight (LoS) links between the UAV and the ground users may be
blocked, which deteriorates the channel gains. However, the RIS-
assisted UAV system can enable virtual LoS paths by reflecting
the signal received from UAV to the ground users. The received Ground user
signal power at the ground user can be significantly enhanced by RIS-assisted UAV communication
the joint optimization of RIS beamforming and UAV trajectory. In
this regard, Li et al. [3] proposed a iterative algorithm for achievable Fig. 2: RIS-assisted SWIPT and UAV communication.
rate maximization under UAV mobility and the RIS’s phase-shift
constraints, with the results confirming significant improvement in
achievable rate with the aid of RIS. to optimize RIS phase-shifts and the source transmit beamforming.
RIS can also enhance the cellular communication of UAVs [10], The application of RIS to BackCom can significantly reduce the
which suffers from down-tilted BS antennas, i.e., the main lobes transmit power, which can be mapped to improve the operational
of antennas are optimized to serve the ground users, while UAVs range.
communication is supported by side lobes only. Through intelligent Furthermore, the ability of RIS to steer the signals in different
and optimized signal reflections, controlled via cellular BS, RIS directions to reduce the inter-user interference can be utilized to
can direct the impinging BS signal towards a specific UAV. The improve the detection performance of the ambient BackCom systems.
RIS reflected signal combines coherently with the direct BS-UAV In this regard, Jia et al. [12] proposed a deep reinforcement learning
signal, thus improves the received signal strength at the UAV. Even (DRL) based approach, namely, the deep deterministic policy gradient
a small-sized RIS, deployed on a building facade, can improve the (DDPG) algorithm, to jointly optimize the RIS and reader beamform-
cellular communication of the UAV flying substantially above the ing for RIS-assisted ambient BackCom system with no knowledge
BS. Moreover, RIS location, i.e., RIS distance from BS distance of channels and ambient signals. The results in [12] demonstrate
and RIS deployment height, is a critical factor in such applications the significant improvement in detection performance of ambient
as performance gain achieved through RIS is maximized when RIS BackCom with the aid of RIS.
location is selected optimally.
E. RIS and mmWaves
D. RIS and BackCom The mmWave communication, with the capability to support multi-
BackCom is a promising solution towards an energy-efficient gigabits of data rate, is perceived as a potential solution for the loom-
and sustainable IoT network. Despite the extensive research on the ing capacity crunch. However, high directivity of mmWaves makes
improvement of reliability and throughput of a BackCom system, it vulnerable to blockage instants, especially in indoor and dense
its short operation range remains a key barrier towards the large- urban environments. As RIS has the capability to introduce effective
scale deployment that needs to be addressed. Recently, in [11], the additional paths, an RIS-enhanced mmWave system can overcome the
authors elaborated on the potentials of RIS-assisted monostatic and limitations of a conventional mmWave system. When the direct links
bistatic BackCom systems, where the RIS is employed to assist the from the BS to users are severely blocked, optimizing the system
communication between the tag and the reader. The authors proposed parameters can provide satisfactory performance gains. Recently, in
a joint optimization framework, under transmit power minimization, [13], the authors utilized the alternative optimization and successive
 4

Power
RIS Controller RIS
cluster 1

cluster M

Frequency
e
m
Ti

Power U1 UK

UK Cluster M
BS U1 UK

U1 U1 UK
Cluster 1
e Frequency
m
Ti

RIS control link BS-UE link BS-RIS link RIS-UE link Unblocked Cluster Blocked Cluster

Fig. 3: An illustration of RIS-assisted downlink NOMA network.

convex approximation (SCA) to jointly optimize the beamforming RIS. The observations indicated in [5] provide useful insights into
vectors and power allocation for the RIS-assisted mmWave-NOMA the design of RIS-based systems. Firstly, the system performance is
system. The results confirm the RIS’s ability to enhance the coverage influenced by the size of RIS and the number of quantization bits for
range of the mmWave-NOMA system, especially when the direct BS discrete phase-shifts. Secondly, the RIS-based hybrid beamforming
to users’ links are blocked. design can greatly reduce the requirement of dedicated hardware
Furthermore, the hybrid precoding design for multi-user RIS- while providing the satisfactory sum-rate.
assisted mmWave communication system is presented in [4], where
the direct links from the BS and the users are assumed to be blocked. III. P RACTICAL I MPLEMENTATION OF RIS- ASSISTED N ETWORKS
The authors jointly optimized the hybrid precoding at the BS and
In this section, we identify and discuss the crucial challenges for
phase-shifts at the RIS to minimize the mean square error (MSE)
the practical implementation of RIS-assisted networks.
between the transmitted and the received symbols. The gradient-
projection (GP) method, based on alternating optimization (AO), is A. RIS Reconfiguration for Controllable Reflections
adopted to address the non-convex constraint for the analog precoding
The RIS phase-shift per element can be tuned for controllable
and the phase-shifts. The results illustrate the significant performance
reflections through three main approaches, namely, mechanical ac-
gains of the proposed design. However, the efficacy of the proposed
tuation, functional materials, and electronic devices. Besides the
design under imperfect CSI needs further investigation. Moreover,
phase control, the reflection amplitude can be adjusted by varying
the proposed design for the hybrid precoding and phase-shifts can be
the load impedance in each element. Thus, the reflection amplitude
extended to a system with the direct links between the BS and the
and phase-shift can be realized in the range of [0, 1] and [0, 2π),
users.
respectively. The continuous variation of reflection coefficients is
usually beneficial from the communication performance perspective.
F. RIS and Multi-antenna Systems
However, for practical RISs with a massive number of reflecting
The multi-antenna systems aim at actively improving the signal elements, it is desirable to implement only a finite number of dis-
quality by employing a large number of antennas and exploiting the crete phase-shift and amplitude levels, since high-resolution elements
spatial domain for transmit beamforming. However, the conventional increase the hardware cost, design complexity, and control overhead.
multiple-input single-output (MISO) systems suffer from wireless In addition, the optimization of reflection amplitudes and phase-
channel randomness, limiting their performance. Therefore, for an shifts becomes more challenging with discrete variables. Furthermore,
energy-efficient solution, RIS can be applied to MISO systems to most of the works on RIS assume ideal phase-shift model with
improve the network performance at significantly low hardware cost full signal reflection at each element regardless of the phase-shift
and energy consumption. Different from the conventional systems, at each element, which is difficult to realize in practice because
an RIS-aided MISO system can guarantee the users’ QoS, with less of the strong coupling amongst the reflection amplitude and phase-
number of BS antennas, by utilizing the smart passive reflections of shift. Therefore, realistic phase-shift models, with phase-dependent
RIS. amplitude response, have to be conceived for accurate performance
Recently, [5] presented a hybrid beamforming design for a multi- analysis.
user RIS-assisted MISO system, where the communication is es-
tablished via RIS only due to the presence of obstacles between B. RIS Deployment and Size Optimization
the BS and the various users. The authors proposed a two-step The RIS deployment in a hybrid network, with both active BSs
sum-rate maximization algorithm to design the continuous digital and passive RISs, is another crucial problem. The deployment strat-
beamforming at the BS and discrete analog beamforming at the egy significantly influences the distribution of the RIS associated
 5

9 1.2
Transmit Power,P
29 (dBm)

37

32

31
28

30
34
N=100

36
s

33
35
N=150

(bps/Hz)
8
N=200
1.1
N=250
27
Total spectral efficiency (bps/Hz)

7 N=300
N=350

Per user target spectral efficiency,

1
6
26

28
29
30
31
32
34

33
5 0.9

27
25
4
0.8
3

26
24

28
29
0.7

30
2

31
32
23

25
27
1 0.6
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 100 150 200 250 300
2 -3
Channel estimation error variance, 10 Number of RIS elements,N
e

Fig. 4: The total SE versus channel estimation error variance for Fig. 5: Required transmit power for varying per user target SE and
different number of RIS elements (N ) with transmit power (Ps ) = number of RIS elements with channel estimation error variance (σe2 )
30 dBm, and users (K) = 3. = 0.0005, and users (K) = 3.

propagation channels, i.e., BS-RIS and RIS-users’ channels. There- of low-overhead channel estimation protocols for future networks
fore, an effective RIS deployment strategy needs to be adopted with a large number of users remains an open problem. Moreover,
to guarantee the performance enhancements promised by the RIS most of the existing works on RIS assume the availability of perfect
technology. Moreover, the RIS deployment must also consider the CSI at the BS, which is practically not feasible. Therefore, the
practical factors, namely, deployment cost, user distribution, and channel estimation errors should be taken into account for accurate
available space, along with propagation conditions. For the single- performance analysis.
user design, the given number of RIS elements can be either grouped
as a single RIS or partitioned into multiple cooperative RISs for IV. RIS- ASSISTED NOMA N ETWORK UNDER IMPERFECT CSI: A
reaping the cooperative passive beamforming gain. Furthermore, for CASE STUDY
the multi-user design, the given number of RIS elements can be
either grouped as a single RIS placed in the vicinity of the BS The integration of RIS with NOMA can provide a potential
(centralized deployment) or partitioned into multiple RISs placed multiple access solution for future networks. RIS-assisted NOMA
closed to users’ hot spots (distributed deployment). In general, the networks can play a significant role in improving network coverage
distributed RISs have a greater probability of establishing LoS links and capacity in urban areas where high-rise buildings and structures
with the BS and the users than the centralized RIS. However, in disrupt wireless services. As illustrated in Fig. 3, we consider an
distributed deployment, the communication between the RISs and the RIS-assisted downlink PD-NOMA network, where a macro BS is
BS, and the coordination among the multiple RISs greatly increase equipped with a single transmit antenna and an RIS equipped with N
the signalling overhead. Moreover, how to optimally select the active reconfigurable passive reflecting elements serve uniformly distributed
number of RIS elements in both centralized and distributed placement single-antenna users, grouped into clusters. The RIS is connected to
is another design challenge. a smart controller, which communicates with the BS and changes
the phase of the incident signal based on coherent phase-shifting
design [7]. The total system bandwidth is equally divided into
C. Channel Estimation in RIS-assisted Networks orthogonal frequency resource blocks, such that each NOMA cluster
The channel estimation for RIS-assisted networks is performed is allocated a single frequency resource block. The available BS
at the BS station, and the acquired CSI is communicated to the power is equally divided among the resource blocks, such that the
RIS controller via control link, which adjusts the phase-shifts. One power allocated to a single resource block is Ps . In the considered
practical approach for RIS’s channel estimation is by employing an network, the direct links from the BS to users are blocked for some
element-by-element ON/OFF-based channel estimation scheme, i.e., clusters, due to the blocking objects, hence such users exploit RIS
only one RIS element is turned on each time while all other elements to establish communication. On the other hand, some clusters have
are set OFF, consequently, the direct channels from the BS to users strong and direct BS-to-user links such that the signal received via
and the RIS reflected channels are estimated separately. However, RIS is negligible because of the increased path loss.
the ON/OFF-based scheme is not cost-effective for large-scale RISs To better explore the role of RIS, we focus on a single blocked
because of training overhead. Furthermore, the RIS reflected signal cluster with K users, where we assumed the user 1 and user K to
suffers from substantial power loss as each time only one element is be the strongest and the weakest users, respectively, based on their
turned on. This weakens the received signal strength, which degrades overall channel gains. Following the fixed power allocation approach
the channel estimation accuracy. To reduce the training overhead for NOMA users, the highest channel gain user, i.e., the strongest
for practical scale RISs, the RIS elements are grouped into sub- user, gets the minimum share of transmit power while the lowest
surfaces [14]; as a result, only the RIS reflected channel associated channel gain user, i.e., the weakest user, gets the maximum share
with the sub-surface needs to be estimated. However, the design of the transmit power. All wireless links, i.e., BS to RIS link and
 6

RIS to users’ links, are modeled as Rayleigh fading channels with

path loss and perturbed by additive white Gaussian noise (AWGN). 350
RIS-NOMA, Ps = 40dBm
We modeled the channel estimates using the minimum mean square
RIS-NOMA, Ps = 30dBm
error (MMSE) channel estimation error model, whose quality of 300
RIS-OMA, P s = 40dBm
estimation is indicated by the variance of the channel estimation
RIS-OMA, P s = 30dBm
error, σe2 , smaller the error variance better the estimation. The channel

Number of RIS elements,N

250
estimation error is regarded as interference in the system, which
adversely affects the system performance.
200

A. Impact of Imperfect CSI and RIS Elements 150

The SE, i.e., the ratio of achievable rate and bandwidth, in the RIS-
assisted NOMA network highly depends on the quality of channel 100
estimation and the number of RIS elements. In Fig. 4 we demonstrate
the impact of varying channel estimation error variance, and the
50
number of RIS elements on total SE, i.e., the SE achieved by K
users. For any value of RIS elements, as error variance increases, the
0
total SE decreases, since the channel estimation error acts as a source 1 2 3 4
of interference. However, even with imperfect CSI, high SE can be Number of users in cluster,K
achieved by deploying a higher number of RIS elements.
The transmit power consumption for RIS-assisted NOMA network Fig. 6: Number of RIS elements versus the number of users for RIS-
depends upon the number of RIS elements, and the target spectral NOMA and RIS-OMA with per user target SE (τ ) = 1.2 bps/Hz, and
efficiency. In this regard, Fig. 5 shows the impact of varying per channel estimation error variance (σe2 ) = 0.0001.
user target spectral efficiency and the number of RIS elements on
BS transmit power. First, it is observed that for a given number of
RIS elements, the required transmit power increases with the increase be applied to THz communication for better coverage performance.
in target spectral efficiency. Second, the transmit power scales down However, the RIS-assisted THz communication poses a major chal-
with the increase in the number of RIS elements. For instance, for lenge of exploiting the unique propagation properties of THz in the
the same per-user target spectral efficiency of 1 bps/Hz, a transmit RIS-assisted network, which needs to be addressed.
power of 32 dBm is required for 150 RIS elements, while this value
reduces to about 26 dBm for 300 elements, which indicates 6 dB
gain by doubling the RIS elements. From this, we can conclude that B. Aerial RIS Empowered Communication
RIS passive reflection adds power gain, which can be either utilized The aerial RIS, carried by UAV or balloon, can realize full-space
to improve SE or reduce total power consumption. reflections to serve a relatively larger number of users than the
ground RIS, fixed at a location. The aerial RIS is more likely to
B. Factors Affecting the Size of RIS enjoy the LoS channel conditions, thus mitigating the blockages.
The minimum number of RIS elements required to guarantee the The high mobility of UAVs can be exploited to further expand the
per user target SE varies with many factors. In Fig. 6, we compare the coverage of aerial RIS. However, in practice, the aerial RIS brings
number of RIS elements required for RIS-NOMA and RIS-OMA with new challenges, including the three-dimensional (3D) placement and
varying number of users and BS transmit power. Here, for RIS-OMA, the channel estimation, that are thus worthy of investigation.
we consider frequency division multiple access (FDMA), where
multiple users transmit data simultaneously at different frequency
slots. It can be observed that for the same number of users and C. RIS-assisted Physical Layer Security
transmit power, the number of RIS elements required to achieve The RIS’s signal manipulation capability can even enhance the
the target SE are greater for RIS-OMA than for RIS-NOMA. This physical layer security (PLS) of the communication links, by simulta-
clearly depicts that the superiority of NOMA compared to OMA neously boosting the signal beam at the intended user and suppressing
still remains after introducing the RIS. The number of RIS elements the beam at the unintended user. The RIS-assisted PLS requires the
also increases with the increase in number of users in the cluster, information of the channels from the eavesdropper to the RIS and
owing to the increase in required orthogonal sub-bands. Moreover, the BS, which is difficult to obtain in practice. This, therefore, calls
the required number of RIS elements can be reduced by increasing for a sophisticated channel estimation and RIS passive beamforming
the BS transmit power. design for RIS-assisted PLS under imperfect CSI.

V. R ESEARCH D IRECTIONS
In this section, we present promising future research directions that D. RIS-assisted Optical Wireless Communication
we consider to be of great importance to unlock the full potential of
Optical wireless communication (OWC) is a promising solution
RISs for 6G networks.
for next-generation high data rate applications at relatively low
hardware cost and complexity than the RF counterpart. Nonetheless,
A. RIS-assisted Terahertz (THz) Communication the performance of OWC is subject to the existence of LoS between
The THz communication, with ultra-wide bandwidth, is considered the transceivers. To relax this constraint, RIS can be applied to OWC
to be a promising candidate for 6G communication. Because of to mitigate the LoS blockages by directing the optical beam in a
the ultra-high frequency, the THz signal may undergo severe signal desired direction. Thus, the integration of RIS and OWC can enable
attenuation and communication interruptions. To this end, RIS can a plethora of applications for both indoor and outdoor scenarios.
 7

E. RIS-assisted mMIMO Network [15] C. Huang et al., “Holographic MIMO Surfaces for 6G Wireless Net-
works: Opportunities, Challenges, and Trends,” IEEE Wireless Commu-
mMIMO, which is the extension of MIMO technology, is one of nications, vol. 27, no. 5, pp. 118–125, 2020.
the key enablers for dramatically improving the transmission gain
and spectral efficiency. However, high hardware cost and power
consumption are the fundamental limitations towards the practical
implementation of mMIMO systems. Nevertheless, RIS can be in- Sarah Basharat (sbasharat.msee19seecs@seecs.edu.pk) received her B.E.
tegrated with mMIMO to provide required performance gains in and M.S. degrees in electrical engineering from National University of
an energy-efficient and cost-effective fashion. The low complexity Sciences and Technology (NUST), Pakistan, in 2019 and 2021, respectively.
algorithms for beamforming designs and resource allocation for RIS- Her research interests include B5G and 6G communications, non-orthogonal
multiple access (NOMA), and reconfigurable intelligent surfaces (RISs).
assisted mMIMO systems need to be studied to achieve maximum
performance.

VI. C ONCLUSION Syed Ali Hassan [S’08, M’11, SM’17] (ali.hassan@seecs.edu.pk) received
his Ph.D. in electrical engineering from Georgia Tech, Atlanta, in 2011, his
In this article, we overviewed the potentials of RIS technology M.S. in mathematics from Georgia Tech in 2011, and his M.S. in electrical
for 6G wireless networks. We first discussed the performance gains engineering from the University of Stuttgart, Germany, in 2007, and a B.E.
that can be achieved by integrating RIS with emerging commu- in electrical engineering (highest honors) from the National University of
nication technologies, such as NOMA, SWIPT, UAVs, BackCom, Sciences and Technology (NUST), Pakistan, in 2004. Currently, he is working
as an associate professor at NUST, where he is heading the IPT research group,
mmWaves, and multi-antenna systems. Despite the great potentials, which focuses on various aspects of theoretical communications.
RIS encounters new challenges to be efficiently integrated into the
wireless network. In this regard, we exposed the crucial challenges
for the practical implementation of RIS-assisted networks. A case
study for RIS-assisted NOMA network under imperfect CSI has Haris Pervaiz [S’09, M’09] (h.b.pervaiz@lancaster.ac.uk) is currently an
also been presented to demonstrate the importance of better channel assistant professor with the School of Computing and Communications (SCC),
estimation for RIS-assisted networks and to indicate the various Lancaster University, U.K. From April 2017 to October 2018, he was a
research fellow with the 5G Innovation Centre, University of Surrey, U.K.
factors affecting the size of RIS. Finally, to provide effective guidance From 2016 to 2017, he was an EPSRC Doctoral Prize Fellow with the SCC,
for future research, we highlighted promising research directions for Lancaster University. He received his Ph.D. degree from Lancaster University,
RIS-assisted networks. U.K., in 2016. His current research interests include green heterogeneous
wireless communications and networking, 5G and beyond, millimeter wave
communication, and energy and spectral efficiency.
R EFERENCES
[1] M. W. Akhtar et al., “The Shift to 6G Communications: Vision and
Requirements,” Human-centric Computing and Information Sciences,
vol. 10, 2020. Aamir Mahmood [M’18, SM’19] (aamir.Mahmood@miun.se) is an assistant
[2] Q. Wu et al., “Towards Smart and Reconfigurable Environment: Intelli- professor of communication engineering at Mid Sweden University, Sweden.
gent Reflecting Surface Aided Wireless Network,” IEEE Commun. Mag., He received the M.Sc. and D.Sc. degrees in communications engineering
vol. 58, no. 1, pp. 106–112, 2020. from Aalto University School of Electrical Engineering, Finland, in 2008
[3] S. Li et al., “Reconfigurable Intelligent Surface Assisted UAV Com- and 2014, respectively. He was a research intern at Nokia Researcher Center,
munication: Joint Trajectory Design and Passive Beamforming,” IEEE Finland and a visiting researcher at Aalto University during 2014-2016. His
Wireless Commun. Lett., vol. 9, no. 5, pp. 716–720, 2020. research interests include network time synchronization, resource allocation
[4] C. Pradhan et al., “Hybrid Precoding Design for Reconfigurable Intelli- for URLLC, and RF interference/coexistence management.
gent Surface Aided mmWave Communication Systems,” IEEE Wireless
Communications Letters, vol. 9, no. 7, pp. 1041–1045, 2020.
[5] B. Di et al., “Hybrid Beamforming for Reconfigurable Intelligent Surface
based Multi-User Communications: Achievable Rates With Limited Dis-
Zhiguo Ding [S’03, M’05, SM’17, F’20] (zhiguo.ding@manchester.ac.uk)
crete Phase Shifts,” IEEE Journal on Selected Areas in Communications,
is currently a Professor at the University of Manchester. Dr Ding research
vol. 38, no. 8, pp. 1809–1822, 2020.
interests are 5G networks, signal processing and statistical signal processing.
[6] E. Björnson et al., “Power Scaling Laws and Near-Field Behaviors of
He has been serving as an Editor for IEEE TCOM, IEEE TVT, and served
Massive MIMO and Intelligent Reflecting Surfaces,” IEEE Open Journal
as an editor for IEEE WCL and IEEE CL. He received the EU Marie Curie
of the Communications Society, vol. 1, pp. 1306–1324, 2020.
Fellowship 2012-2014, IEEE TVT Top Editor 2017, 2018 IEEE COMSOC
[7] Z. Ding et al., “On the Impact of Phase Shifting Designs on IRS-
Heinrich Hertz Award, 2018 IEEE VTS Jack Neubauer Memorial Award, and
NOMA,” IEEE Wireless Commun. Lett., vol. 9, no. 10, pp. 1596–1600,
2018 IEEE SPS Best Signal Processing Letter Award.
2020.
[8] M. Fu et al., “Intelligent Reflecting Surface for Downlink Non-
Orthogonal Multiple Access Networks,” in IEEE GC Wkshps, 2019, pp.
1–6.
[9] C. Pan et al., “Intelligent Reflecting Surface Aided MIMO Broadcasting Mikael Gidlund [M’98, SM’16] (mikael.gidlund@miun.se) is a professor of
for Simultaneous Wireless Information and Power Transfer,” IEEE J. Sel. computer engineering at Mid Sweden University, Sweden. He has worked as
Areas Commun., 2020. Senior Principal Scientist and Global Research Area Coordinator of Wireless
[10] D. Ma et al., “Enhancing Cellular Communications for UAVs via Technologies, ABB Corporate Research, Sweden, Project Manager and Senior
Intelligent Reflective Surface,” in IEEE WCNC, 2020, pp. 1–6. Specialist with Nera Networks AS, Norway, and Research Engineer and
[11] X. Jia et al., “Intelligent Reflecting Surface-Aided Backscatter Commu- Project Manager with Acreo AB, Sweden. His current research interests
nications,” arXiv preprint arXiv:2004.09059, 2020. include wireless communication and networks, wireless sensor networks,
[12] X.Jia et al., “IRS-Assisted Ambient Backscatter Communications Uti- access protocols, and security.
lizing Deep Reinforcement Learning,” arXiv preprint arXiv:2103.07083,
2021.
[13] J. Zuo et al., “Intelligent Reflecting Surface Enhanced Millimeter-Wave
NOMA Systems,” IEEE Commun. Lett., pp. 1–1, 2020.
[14] Y. Yang et al., “Intelligent Reflecting Surface Meets OFDM: Protocol
Design and Rate Maximization,” IEEE Trans Commun., 2020.