Weighted Sum-Rate Maximization for Intelligent User A Intelligent Reflecting Surface User B

Reflecting Surface Enhanced Wireless Networks Tx

AS BS
Tx

AA BB

Huayan Guo† , Ying-Chang Liang† , Jie Chen† , and Erik G. Larsson∗

SA SB
† University
of Electronic Science and Technology of China (UESTC), Chengdu, P. R. China
Rx Rx

Email: guohuayan@pku.edu.cn, liangyc@ieee.org, chenjie.ay@gmail.com A to B B to A

∗ Department of Electrical Engineering (ISY), Linköping University, SE-581 83 Linköping, Sweden

Email: erik.g.larsson@liu.se MM

SINR P
Abstract—Intelligent reflecting surface (IRS) is a promising
solution to build a programmable wireless environment for 1 2 N
future communication systems, in which the reflector elements IRS
IRS
steer the incident signal in fully customizable ways by passive G
1 2 M
beamforming. This work focuses on the downlink of an IRS-
aided multiuser multiple-input single-output (MISO) system. A
practical IRS assumption is considered, in which the incident hd,k hr,k
User 1
signal can only be shifted with discrete phase levels. Then, the
weighted sum-rate of all users is maximized by joint optimizing
the active beamforming at the base-station (BS) and the passive BS User k

beamforming at the IRS. This non-convex problem is firstly

decomposed via Lagrangian dual transform, and then the active
and passive beamforming can be optimized alternatingly. In
addition, an efficient algorithm with closed-form solutions is
proposed for the passive beamforming, which is applicable to
both the discrete phase-shift IRS and the continuous phase-
shift IRS. Simulation results have verified the effectiveness of
the proposed algorithm as compared to different benchmark
schemes.
Index Terms—Intelligent reflecting surface (IRS), Reconfig-
urable Intelligent Surface (RIS), passive radio, multiple-input-
multiple-output (MIMO), discrete phase shifts, beamforming.
I. I NTRODUCTION
Passive radio technique has been considered as the key
enabler for the smart radio environment, based on which
the passive device reflects the incident radio-frequency (RF)
signal with the real-time programmable reflection coefficient
(RC), thanks to the recent advances in meta-materials and
electromagnetic scattering theory. One key application of
the passive radio is ambient backscatter communications or
symbiotic radio network [1]–[4], which are used to support
low-power communications in Internet of Things (IoT). In
particular, the data of the IoT devices are embedded into the
reflected signal from the environment rather than emitting a
new radio carrier. Recently, a novel passive-radio application
called Intelligent reflecting surface (IRS), also known as
reconfigurable intelligent surface (RIS), attracts increasing
attention from both academia and industry [5]–[7]. The IRS
is comprised of massive passive reflector elements with re-
configurable RC. Instead of data transmission, the IRS is
utilized to enhance the performance of the wireless system via
control over the behavior of wireless environment by passive
beamforming [8]–[10].
Fig. 1 illustrates a typical IRS-aided wireless communi-
cation system, in which one multi-antenna base station (BS)
User K
Fig. 1. An IRS-aided multiuser MISO communication system.
serves multiple single-antenna users. The direct links between
the BS and the mobile users may suffer from deep fading and
shadowing. The IRS is deployed on a surrounding building’s
facade to assist the BS in overcoming the unfavorable propa-
gation conditions by providing high-quality virtual links from
the BS to the users. The objective of this paper is to maximize
the weighted sum-rate (WSR) of the mobile users by jointly
optimizing the active beamforming at the BS and the passive
beamforming at the IRS.
Most existing works assume that the reflector element on
the IRS is able to induce a certain phase shift on the incident
RF signal. Thus the IRS passive beamforming is equivalent to
optimizing the phase-shift matrix. In [11], the transmit power
of the BS is minimized subject to the rate requirements of
mobile users by jointly optimizing the active beamforming
at the BS and the passive beamforming at the IRS. In [12],
the energy efficiency is maximized via joint beamforming
based on gradient descent and the majorization-minimization
approach. In [13], the minimum signal-to-interference-plus-
noise ratio (SINR) of mobile users is maximized under some
practical considerations about the channel model. To the best
of our knowledge, the joint beamforming for maximizing the
WSR of all users has not been investigated before.
In practice, the continuous phase-shift setup is difficult to
realize due to the hardware limitation, and thus the reflector
element may only shift the incident signal with discrete RC
values. In [14] and [15], quantization approach is proposed
to extract the discrete value from the optimized solution
for the continuous phase-shift setup. However, this method

978-1-7281-0962-6/19/$31.00 ©2019 IEEE

Authorized licensed use limited to: ULAKBIM UASL - KOCAELI UNIVERSITESI. Downloaded on October 28,2020 at 11:42:01 UTC from IEEE Xplore. Restrictions apply.

is heuristic with unpredictable performance loss. In [16], a
sequential algorithm is proposed to directly deal with the
discrete feasible set for single-user system via optimizing the
element’s phase one by one. However, the performance of
√
Denote by Θ = ηdiag(θ1 , · · · , θn , · · · , θN ) the phase-
shift matrix of the IRS, where η < 1 indicates the reflection
efficiency, and θn is the RC of the n-th reflector element. For
ease of practical implementation, θn only takes d finite values,
the sequential algorithm is vulnerable to be trapped by the which are equally 
{ spaced in [0, 2π). Then the set of θn is giv-
}
 2π(d−1)
uninteresting solutions.
In this paper, we consider the practical discrete phase-shift
setup for the WSR maximization problem. Apart from the
phase-shift matrix optimization on the non-convex discrete
feasible set, the WSR problem is still NP-hard, which is
rather elusive compared with the min-max fairness problem
or the transmit power minimization problem [17]. We try
to design an unified algorithm to find a suboptimal solu-
tion with low computational complexity for both continuous
and discrete phase-shift setups. To be specific, the active
beamforming at BS and the passive beamforming at IRS are
decomposed based on the Lagrangian dual transform proposed
in [18]. Then, the active beamforming subproblem is solved
with closed-form solutions using the multi-ratio quadratic
transform [18]. The RC adjustment subproblem is reformu-
lated as the non-convex quadratically constrained quadratic
program (QCQP), and is further solved via the alternating
direction method of multipliers (ADMM) instead of the high-
en by FD = θn θn = ejφn , φn ∈ {0, 2π d ,··· , d } .
Note that, the algorithm developed in this paper for the
discrete phase-shift assumption is also applicable to the con-
tinuous phase-shift cases by letting d approach to infinity.
Denote the information bearing symbol of user k by sk ,
which is independent random variable with zero mean and
unit variance. We consider linear beamforming at the BS, and
the beamforming vector for user k is denoted by wk ∈ CM ×1 .
Then, the signal received at user k is expressed as
( )∑
K
yk = hH
d,k + hH
r,k ΘH
G wk sk + uk , (1)
k=1
where the superscript H denotes the conjugate transpose, and
uk ∼ CN (0, σ02 ) denotes the additive white Gaussian noise
(AWGN) at the k-th user receiver. The k-th user treats all
the signals from other users as interference. Accordingly, the
decoding SINR is
complexity semidefinite relaxation (SDR) technique suggested  2
 H 
by [16]. The main contributions of this work are summarized (hd,k + hHr,k Θ H
G)wk 
as follows: γk = ∑  2 . (2)
K  H 
i=1,i̸=k (hd,k + hr,k Θ G)wi  + σ0
H H 2
• Firstly, this paper investigates the WSR maximization
problem for the multiuser downlink MISO system aided
by IRS. An iterative algorithm with closed-form solutions III. WSR M AXIMIZATION FOR D OWNLINK
is proposed to jointly optimize the active and passive T RANSMISSION
beamforming. In this paper, we aim to maximize the WSR of all the K
• Secondly, we design a low-complexity algorithm for the users by jointly designing the beamforming vectors at the BS
passive beamforming, which is applicable to both the and the RC∑matrix at the IRS, subject to the transmit power
K
k=1 ∥wk ∥ ≤ PT of BS. Mathematically, the
discrete and the continuous phase-shift IRS. constraint 2
• Finally, simulation results have verified the effectiveness WSR maximization problem is thus formulated as
of the proposed algorithm. In particular, the 2-bit phase
shifter may work well with only a small performance ∑
K
(P1) max f1 (W, Θ) = ωk log(1 + γk )
degradation compared with the performance achieved by W,Θ
k=1
the ideal phase shifter assumption.
s.t. θn ∈ FD , ∀n = 1, · · · , N, (3a)
The rest of the paper is organized as follows. Section
II outlines the system model. The joint active and passive ∑
K
∥wk ∥2 ≤ PT , (3b)
beamforming problem is formulated and solved in Section III.
k=1
Simulation results are provided in Section IV, and Section V
concludes the paper. where W = [w1 , w2 , · · · , wK ] ∈ CM ×K , and the constant
weight ωk is used to represent the priority of user k.
II. S YSTEM M ODEL
Despite the conciseness of (P1), it is generally very hard
As shown in Fig. 1, we investigate an IRS-aided multiuser to obtain the optimal solution due to the non-convex objective
MISO communication system, in which one multi-antenna function f1 (W, Θ) and the non-convex constraint sets FD . In
BS serves K single-antenna users. The numbers of transmit this paper, we propose a low-complexity algorithm to solve
antennas at the BS and reflector elements at the IRS are (P1) sub-optimally based on the fractional programming tech-
denoted by M and N . In addition, the baseband equivalent nique [18]. Applying the Lagrangian dual transform proposed
channels from BS to user k, from BS to IRS, and from IRS in [18], (P1) can be equivalently written as
to user k are denoted by hd,k ∈ CM ×1 , G ∈ CN ×M , and
hr,k ∈ CN ×1 , respectively. We assume a quasi-static flat- (P1′ ) max f1a (W, Θ, α)
W,Θ,α
fading model, and the channel state information (CSI) of all
channels involved is perfectly known by the BS and the IRS. s.t. (3a), (3b),

Authorized licensed use limited to: ULAKBIM UASL - KOCAELI UNIVERSITESI. Downloaded on October 28,2020 at 11:42:01 UTC from IEEE Xplore. Restrictions apply.
 User

Power allocation
User
SINR
where α refers to [α1 , · · · , αk , · · · , αK ]T , and αk is an where β is the vector of the auxiliary variables [β1 , · · · , βK ]T ,
auxiliary variable
User for the decoding SINR γk ; and the new and the superscript ∗ denotes the conjugate of scalar. Then,
IRS phase adjustment
objective function is defined by β and W can be updated with closed-form solutions.
∑ K ∑ K Lemma 1: The beamforming vector wk is updated by
f1a (W, Θ, α) = ωk log(1 + αk ) − ωk αk ( )−1
k=1 k=1 √ ∑K
◦
wk = α̃k βk λ0 IM + |βi | hi hi
2 H
hk , (8)
∑K
ωk (1 + αk )γk i=1
+ . (4)
1 + γk
k=1 where λ0 is the dual variable introduced for the transmit power
′ Update SINR ◦
alize 0 0In (P1 ), when W and Θ hold fixed, the optimal αk is αk = constraint, which is determined by
γk . Then, for aafixed
k g kα, optimizing W and Θ is reduced to { }
∑K
∑K
α̃ γ
◦
λ0 = min λ0 ≥ 0 : ∥wk ∥ ≤ PT ,
2
(9)
(P1′′ ) max , ( i +k1) , k ( i +1)
(i )
, (i ) , (i ) (i )

W,Θ 1 + γk k=1
k=1
s.t. (3a), (3b), and βk is the fractional-programming auxiliary variable,
which is given by
BSwhere α̃k = ω( ik) ,(1 (+
Transmit i 1)
α, k ().
i)
IRS RC √
To sum up, the original problem (P1) is firstly transformed α̃k hH
k wk
Beamforming Optimization
to (P1′ ), in which the active beamforming at BS and the βk◦ = ∑K   . (10)
hH wi 2 + σ 2
i=1 k 0
passive beamforming at the IRS can be decomposed. Then,
we can solve (P1′ ) in an iterative manner by alternatively Proof: βk◦ in (10) and wk◦ in (8) can be obtained by
optimizing α, W and Θ as illustrated in Fig. 2. setting ∂f2a /∂βk and ∂f2a /∂wk to zero, respectively. The
detailed derivations can be found in [18, Section IV].

Init Θ, W Update α Update W Update Θ B. Optimizing Reflection Response Matrix Θ

For a fixed W, the objective function of (P1′′ ) can be
Fig. 2. Alternating optimization for (P1′ ). expressed as a function of Θ:

∑
K
α̃k |(hH
d,k + hr,k Θ G)wk |
H H 2
A. Transmit Beamforming f3u (Θ) = ∑K . (11)
For a given Θ, the composite channel for user k is hk = k=1 i=1 |(hH
d,k + hr,k Θ G)wi | + σ0
H H 2 2

hd,k + GH Θhr,k . Then, the SINR γk in (2) becomes √

 H 2 By defining ai,k = ηdiag(hH H
r,k )Gwi , bi,k = hd,k wi , and
h wk  θ = [θ1 , · · · , θN ] , |(hd,k + hr,k Θ G)wi | in (11) becomes
T H H H
k
γk = ∑K   , (5)
 H 2 + σ 2 √ H
i=1,i̸=k hk wi 0
|(hH
d,k + hr,k Θ G)wi | = |bi,k +
H H
r,k )Gwi |
ηθ diag(hH
′′
and the objective function of (P1 ) is written as a function
= |bi,k + θ H ai,k |.
of W:  2
∑ α̃k hH  (12)
K
k wk
f2 (W) = ∑K  H 2 . (6) Using (12), f3u (Θ) in (11) is equivalently transformed to a
k=1 i=1 hk wi + σ02 new function of θ:
Thus, given α and Θ, optimizing W becomes
∑
K
α̃k |bk,k + θ H ak,k |2
(P2) max f2 (W) f3 (θ) = ∑K . (13)
i=1 |bi,k + θ ai,k | + σ0
H 2 2
W k=1
∑
K
s.t. ∥wk ∥2 ≤ PT . As a result, optimizing Θ is translated to optimizing θ, which
k=1 is represented as follows:
(P2) is the multiple-ratio fractional programming problem.
(P3) max f3 (θ)
Using quadratic transform proposed in [18], (P2) can be θ
reformulated as a biconvex optimization problem with a new s.t. θn ∈ FD , ∀n = 1, · · · , N.
objective function:
∑ (P3) is also a multiple-ratio fractional programming prob-
K √ { }
f2a (W, β) = 2 α̃k Re βk∗ hH
k wk
lem, and can be translated to the following problem
k=1
(K ) (P3a) max f3a (θ, ε)
∑K ∑ 2 θ,ε
−
2
|βk | hk wi  + σ0 ,
H 2
(7)
s.t. θn ∈ FD , ∀n = 1, · · · , N,
k=1 i=1

Authorized licensed use limited to: ULAKBIM UASL - KOCAELI UNIVERSITESI. Downloaded on October 28,2020 at 11:42:01 UTC from IEEE Xplore. Restrictions apply.
 where the new objective function is Then, (P4) is equivalently represented as
µ
∑
K √ { } (P4a) max f4 (q) − ∥q − θ∥22
f3a (θ, ε) = 2 α̃k Re ε∗k θ H ak,k + εH
k bk,k
θ,q 2
k=1 s.t. q = θ, (20a)
( )
∑K
2
∑
K
θn ∈ FD , ∀n = 1, · · · , N, (20b)
− |εk | |bi,k + θ H ai,k |2 + σ02 , (14)
k=1 i=1 where µ > 0 is the penalty parameter. Then, we have the
Lagrangian of (P4c):
and ε refers to the auxiliary variable vector [ε1 , · · · , εK ]T .
Similarly, we optimize θ and ε alternatively. The optimal ∑
N
µ
εk for a given θ can be obtained by setting ∂f3a /∂εk to zero,
G(q, θ, λR , λI ) = −q H U q − 1F (θn ) − ∥q − θ∥22
2
i.e., { n=1
}
√ ( ) H
α̃k bk,k + θ H ak,k + Re 2q ν + (λR + jλI ) (q − θ) ,
H
ε◦k = ∑K . (15) (21)
i=1 |bi,k + θ ai,k | + σ0
H 2 2
where λR = [λR,1 , · · · , λR,N ]T and λI = [λI,1 , · · · , λI,N ]T
Then the remaining problem is optimizing θ for a given ε. It are the Lagrange variables for Re{q − θ} = 0 and Im{q −
is known that, |bi,k + θ H ai,k |2 in (14) can be further written θ} = 0, respectively, and 1F (·) is the indicator function of
as set F (i.e., 1F (θn ) = 0 if θn ∈ F ; otherwise, equals infinity).
( )( ) The dual problem for (P4a) is formulated as
|bi,k + θ H ai,k |2 = bi,k + θ H ai,k b∗i,k + aH
i,k θ
{ ∗ H } (P5) min L(λR , λI ) = max {G(q, θ, λR , λI )} .
i,k θ + 2Re bi,k θ ai,k + |bi,k | . (16)
= θ H ai,k aH 2
λR ,λI θ,q

Substituting (15) and (16) into (14), and dropping irrelevant The ADMM proposed in [19] aims to find the optimal
constant terms, the optimization problem for θ is represented solution for the dual problem (P5), which has the following
as follows iterative forms:

(P4) max f4 (θ) θ t+1 = arg max G(qt , θ, λ̄t ), (22)

θ
θ
s.t. θn ∈ FD , ∀n = 1, · · · , N, qt+1 = arg max G(q, θ t+1 , λ̄t ), (23)
q
( )
λ̄t+1
= λ̄ − µ qt+1 − λ̄t+1 ,
t
(24)
where the objective function is
{ } where λ̄ = λR + jλI , and t is the iteration index.
f4 (θ) = −θ H U θ + 2Re θ H ν , (17) 1) Optimizing θ: In (22), the optimal θ for fixed qt and
and λ̄t is ( )
1
θ t+1 = PjF qt − λ̄t , (25)
∑
K
2
∑
K µ
U= |εk | ai,k aH
i,k , (18)
k=1 i=1
where PjF (·) is the projection operation. Let θ̄ = qt − µ1 λ̄t .
( ) We have
∑K √ ∑
K
α̃k ε∗k ak,k b∗i,k ai,k
2
ν= − |εk | . (19) ∠θnt+1 = arg min |φn − ∠θ̄n |. (26)
2π(d−1)
k=1 i=1 φn ∈{0, 2π
d ,··· , d }

Since ai,k aHi,k for all i and k are positive-definite matrices, U

Note that, for continuous phase-shift setup (d → +∞),
is a positive-definite matrix, and f4 (θ) is a quadratic concave equation (26) reduces to
function of θ. Therefore, the non-convexity of (P4) is only
∠θnt+1 = ∠θ̄n . (27)
introduced by the constraint set.
2) Optimizing q: In (23), q is optimized given θ t+1 and
C. Solving (P4) by ADMM t
λ̄ , and we have
(P4) is a non-convex QCQP. In [11], this problem is −1 ( )
qt+1 = (2U + µIN ) 2v + λ̄t + µθ t+1 . (28)
solved by applying the SDR technique to relax the non-
convex constraint. However, the complexity of solving SDR It can be verified that the complexity of the ADMM
is very high (in the order of O(N 6 )). Recently, a low- algorithm is O(N 3 ), which is much smaller than that of the
complexity algorithm has been proposed in [19] to solve the SDR based method. In addition, the convergence condition of
non-convex problem with convex objective function and non- the ADMM algorithm is presented in the following lemma.
convex constraints based on the ADMM. In this subsection, Lemma 2: The ADMM algorithm presented above guar-
we adopt the ADMM algorithm to solve (P4). antees to converge, if the penalty parameter µ satisfies:
The key idea of the non-convex ADMM is introducing an µ
IN − U ≻ 0. (29)
auxiliary vector q for θ, as well as a penalty term for q ̸= θ. 2

Authorized licensed use limited to: ULAKBIM UASL - KOCAELI UNIVERSITESI. Downloaded on October 28,2020 at 11:42:01 UTC from IEEE Xplore. Restrictions apply.
 Proof: The detailed proof can be referred to [19]. 18
It is worth noting that, the ADMM algorithm needs not
16
to converge to the global or even local optimum. Thus, we
14
need to check whether the solution is better than the initial
point. Also note that, if we relax the non-convex constraint 12

in (20b) to a convex form, i.e., |θn | ≤ 1, (P4a) becomes

R (bps/Hz)
10
convex, and the ADMM converges to the global optimum.
8
This optimal objective value may serve an upper bound of
6 Upper bound
(P4a) for performance evaluation. Proposed, d=+
4 Proposed, d=4
Proposed, d=2
D. Algorithm Development 2 Baseline 2
Baseline 1

We summarize the proposed alternating optimization 0

0 1 2 3 4 5 6 7 8 9 10
method in Algorithm 1. In particular, the algorithm is ini- P T (dB)

tialized by W (0) and Θ(0) chosen from the feasible set.

Then, given a fixed solution {W (i) , Θ(i) } in the i-th iter- Fig. 3. The average sum rate versus transmit power for K = 4, N = 10
and η = 0.1.
ation, α(i+1) , W (i+1) , and Θ(i+1) are updated alternatively
for the (i + 1)-th iteration. Since the objective function is
monotonically nondecreasing after each iteration, Algorithm
• Baseline 3 (Upper bound): We relax the constraint in (3a)
1 is guaranteed to converge.
for (P1) to |θn | ≤ 1.1 Then the performance upper bound
of the proposed algorithm is obtained. Moreover, we
Algorithm 1 The alternating optimization for solving (P1′ ). adopt the solution in this case to assist the initialization
1: Set i = 0, and initialize W (0) and Θ(0) to feasible values. of the proposed algorithm under non-convex constraints.
2: repeat In Fig. 3, we show the average sum rate of different
3: i = i + 1; schemes with respect to the transmit power PT , where the IRS
4: Update the nominal SINR α(i) ; has 10 reflector elements with reflection efficiency η = 0.1.
5: Update β (i) by (10); It is known that, the performance gain of the IRS comes
6: Update transmit beamforming W (i) by (8); from two aspects. One is the passive beamforming gain, and
7: Update ε(i) by (15); the other is the inherent aperture gain by relaying all signal
8: Get Θ⋆ by solving (P4) based on ADMM; power in the BS-IRS link. Therefore, we observe that baseline
9: if f3u (Θ⋆ ) > f3u (Θ(i−1) ) then 2 outperforms baseline 1, although the phase-shift matrix is
10: Θ(i) = Θ⋆ ; randomly chosen. In addition, the proposed algorithm with
11: else continuous phase shifter achieves almost the same perfor-
12: Θ(i) = Θ(i−1) ; mance as the upper bound. Thus, the ADMM obtains nearly
13: end if optimal solution for the non-convex passive beamforming
14: until The value of function f1a in (4) converges. subproblem for the continuous phase-shift setup. Moreover,
the “1-bit” phase-shift setup achieves more than 2-dB gain
compared with baseline 2, and the performance gap between
the “2-bit” phase-shift setup and the upper bound is smaller
IV. N UMERICAL R ESULTS
than 1 dB.
In this section, numerical examples are provided to validate In Fig. 4, we compare the average sum rate with the number
the effectiveness of the proposed algorithm for the IRS-aid of reflector elements. The transmission power is fixed at
multi-user MISO system. The system has 4 users, in which PT = 5 dB, and the reflection efficiency is η = 0.1. It is
the BS is equipped with 4 antennas. Simulation results are shown that the performance of all the schemes compared with
obtained by averaging over 104 channel realizations. All the baseline 1 increases with the increase of N , since the signal
channels involved are assumed to be Rayleigh fading, and the strength of the reflection links (from BS-to-IRS-to-users)
channel coefficients (i.e., the elements in G, hd,k , and hr,k , becomes stronger. We also observe that, the performance gain
for all k) are normalized with zero mean and unit variance. of the proposed algorithm compared with baseline 2 increases
We compare the performance of the proposed algorithm with as N increases.
the following 3 baselines: In Fig. 5, we continue comparing the average sum rate
• Baseline 1 (Without IRS): Let N = 0, and then the active with N . The only difference with Fig. 4 is that the reflection
beamforming is optimized via the WMMSE in [17]. efficiency is normalized by the number of reflector elements,
• Baseline 2 (Random θ): the phase-shift matrix of the 1 Note that, since both the amplitude and the phase of θ are adjustable, the
n
IRS is not optimized, in which the RCs are chosen by performance under this assumption will better than that under the continuous
random values. phase-shift assumption.

Authorized licensed use limited to: ULAKBIM UASL - KOCAELI UNIVERSITESI. Downloaded on October 28,2020 at 11:42:01 UTC from IEEE Xplore. Restrictions apply.
 25
ACKNOWLEDGEMENT
Upper bound
Proposed, d=+
This work was supported in part by the National Natu-
Proposed, d=4 ral Science Foundation of China under Grants U1801261,
Proposed, d=2
20
Baseline 2 61631005, and 61571100.
Baseline 1
R EFERENCES
R (bps/Hz)

15 [1] N. V. Huynh, D. T. Hoang, X. Lu, D. Niyato, P. Wang, and D. I.

Kim, “Ambient backscatter communications: A contemporary survey,”
Commun. Surveys Tuts., vol. 20, no. 4, pp. 2889–2922, 2018.
[2] H. Guo, Q. Zhang, S. Xiao, and Y.-C. Liang, “Exploiting multiple anten-
10 nas for cognitive ambient backscatter communication,” IEEE Internet
Things J., vol. 6, no. 1, pp. 765–775, Feb 2019.
[3] H. Guo, Y.-C. Liang, R. Long, S. Xiao, and Q. Zhang, “Resource
allocation for symbiotic radio system with fading channels,” IEEE
5
5 10 15 20 25 30 35 40 45 50 Access, vol. 7, pp. 34 333–34 347, 2019.
N [4] Q. Zhang, H. Guo, Y.-C. Liang, and X. Yuan, “Constellation learning
based signal detection for ambient backscatter communication systems,”
IEEE J. Sel. Areas Commun., vol. 37, no. 2, pp. 452–463, 2019.
Fig. 4. The average sum rate versus N for η = 0.1.
[5] X. Tan, Z. Sun, D. Koutsonikolas, and J. M. Jornet, “Enabling indoor
mobile millimeter-wave networks based on smart reflect-arrays,” in
Proc. IEEE INFOCOM, Apr. 2018, pp. 270–278.
20 [6] F. Liu, O. Tsilipakos, A. Pitilakis, A. C. Tasolamprou, M. S. Mirmoosa,
Upper bound N. V. Kantartzis, D.-H. Kwon, M. Kafesaki, C. M. Soukoulis, and
Proposed, d=+
S. A. Tretyakov, “Intelligent metasurfaces with continuously tunable
Proposed, d=4
Proposed, d=2
local surface impedance for multiple reconfigurable functions,” Physical
Baseline 2 Review Applied, vol. 11, no. 4, p. 044024, 2019.
15 Baseline 1 [7] S. V. Hum and J. Perruisseau-Carrier, “Reconfigurable reflectarrays and
array lenses for dynamic antenna beam control: A review,” IEEE Trans.
R (bps/Hz)

Antennas Propag., vol. 62, no. 1, pp. 183–198, 2014.

10
5 urable meta-surfaces: An idea whose time has come,” arXiv preprint
5 10 15 20 25 30 35 40
N
Fig. 5. The average sum rate versus N for η =
i.e., η = N1 . As a result, the aperture gain of the IRS changes
little as N increases, which is verified by baseline 2. However,
we can still observe that the performance gain of the proposed
algorithm increases with the increase of N .
V. C ONCLUSION
In this paper, we investigate the IRS-aided multiuser down-
link MISO system while considering practical discrete phase-
shift constraints for the IRS. An iterative low-complexity algo-
rithm with closed-form solutions is proposed to maximize the
WSR of all users by jointly designing the active beamforming
at the BS and the passive beamforming at the IRS. The pro-
posed algorithm is applicable not only to the practical discrete
phase-shift IRS but also the popular continuous phase-shift
IRS. Extensive simulation results have demonstrated that the
proposed joint beamforming algorithm achieves significant
capacity gain compared with the conventional system without
the IRS and the IRS-aided system employing random passive
beamforming. Moreover, it is also shown that the discrete-
phase-shifting IRS with 2-bit quantizer may achieve sufficient
capacity gain with only a small performance degradation.
[8] C. Liaskos, S. Nie, A. Tsioliaridou, A. Pitsillides, S. Ioannidis,
and I. Akyildiz, “A new wireless communication paradigm through
software-controlled metasurfaces,” IEEE Commun. Mag., vol. 56, no. 9,
pp. 162–169, 2018.
[9] M. Di Renzo, M. Debbah, D.-T. Phan-Huy, A. Zappone, M.-S. Alouini,
C. Yuen, V. Sciancalepore, G. C. Alexandropoulos, J. Hoydis, and
H. Gacanin, “Smart radio environments empowered by AI reconfig-
45 50 arXiv:1903.08925, 2019.
[10] Y.-C. Liang, R. Long, Q. Zhang, J. Chen, H. V. Cheng, and H. Guo,
“Large intelligent surface/antennas (LISA): Making reflective radios
1
N
. smart,” Journal of Communications and Information Networks, vol. 4,
no. 2, pp. 40–50, 2019.
[11] Q. Wu and R. Zhang, “Intelligent reflecting surface enhanced wireless
network: Joint active and passive beamforming design,” in Proc. IEEE
Globecom, Dec. 2018, pp. 1–6.
[12] C. Huang, A. Zappone, G. C. Alexandropoulos, M. Debbah, and
C. Yuen, “Reconfigurable intelligent surfaces for energy efficiency
in wireless communication,” IEEE Trans. Wireless Commun., Early
Access.
[13] Q.-U.-A. Nadeem, A. Kammoun, A. Chaaban, M. Debbah, and M.-S.
Alouini, “Large intelligent surface assisted MIMO communications,”
arXiv:1903.08127, 2019.
[14] C. Huang, G. C. Alexandropoulos, A. Zappone, M. Debbah, and
C. Yuen, “Energy efficient multi-user MISO communication using
low resolution large intelligent surfaces,” in Proc. IEEE Globecom
Workshops, Dec. 2018, pp. 1–6.
[15] J. Chen, Y. Liang, Y. Pei, and H. Guo, “Intelligent reflecting surface: A
programmable wireless environment for physical layer security,” IEEE
Access, vol. 7, pp. 82 599–82 612, 2019.
[16] Q. Wu and R. Zhang, “Beamforming optimization for intelligent reflect-
ing surface with discrete phase shifts,” in Proc. IEEE ICASSP, 2019,
pp. 7830–7833.
[17] Q. Shi, M. Razaviyayn, Z. Luo, and C. He, “An iteratively weighted
MMSE approach to distributed sum-utility maximization for a MIMO
interfering broadcast channel,” IEEE Trans. Signal Process., vol. 59,
no. 9, pp. 4331–4340, 2011.
[18] K. Shen and W. Yu, “Fractional programming for communication
systems–Part I: Power control and beamforming,” IEEE Trans. Signal
Process., vol. 66, no. 10, pp. 2616–2630, 2018.
[19] Y. Wang, W. Yin, and J. Zeng, “Global convergence of ADMM in
nonconvex nonsmooth optimization,” Journal of Scientific Computing,
vol. 78, no. 1, pp. 29–63, 2019.

Authorized licensed use limited to: ULAKBIM UASL - KOCAELI UNIVERSITESI. Downloaded on October 28,2020 at 11:42:01 UTC from IEEE Xplore. Restrictions apply.