Exploiting millimeter wave in non-orthogonal multiple access based full-duplex cooperative device-to-device communications system

Abstract

The ever increasing demand for high data rate and spectral efficiency has led to the advent of the underutilized millimeter-wave (mmWave) frequency spectrum for fifth-generation and beyond (B5G) cellular networks. mmWave facilitates high throughput for short-range technologies such as device-to-device (D2D) communications by utilizing the carrier frequencies beyond 30 GHz. This paper proposes a mmWave assisted cooperative D2D (C-D2D) framework wherein a D2D transmitter (DT) acts as a full-duplex (FD) relay for cellular uplink transmission. In addition, DT employs non-orthogonal multiple access (NOMA) to transmit the superimposed cellular and D2D signal while utilizing power domain multiplexing. Successive interference cancellation is applied at the D2D receiver to decode the D2D signal. Further, we considered that each node is equipped with directional antennas to compensate for the impact of high propagation loss, and a sectored beamforming framework is used to model the antenna gain. The analytical expressions for the achievable rates and outage probabilities for cellular and D2D users with an optimal value of the power splitting factor have been derived to characterize the system performance. Our results include the impact of various system parameters such as half-power beamwidth, sidelobe gain, and residual self-interference constants on the cellular outage probability. We have also shown that our proposed model outperforms the recent work on the NOMA-aided FD C-D2D communications system by approximately 104\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$10^4$$\end{document} times in terms of cellular outage probability.

Full text

Exploiting Millimeter Wave in
Non-Orthogonal Multiple Access Based
Full-Duplex Cooperative Device-to-Device
Communications System
Rahul Bajpai*, Nikhil Mahadev Karoti and Naveen Gupta
Department of Electrical and Electronics Engineering, BITS
Pilani K.K. Birla Goa Campus, NH 17B, Bypass, Road, Sancoale,
403726, Goa, India.
*Corresponding author(s). E-mail(s):
p20190003@goa.bits-pilani.ac.in;
Contributing authors: f20180740@goa.bits-pilani.ac.in;
naveeng@goa.bits-pilani.ac.in;
Abstract
The ever increasing demand for high data rate and spectral efficiency
has led to the advent of the underutilized millimeter-wave (mmWave)
frequency spectrum for fifth-generation and beyond (B5G) cellular net-
works. mmWave facilitates high throughput for short-range technologies
such as device-to-device (D2D) communications by utilizing the carrier
frequencies beyond 30 GHz. This paper proposes a mmWave assisted
cooperative D2D (C-D2D) framework wherein a D2D transmitter (DT)
acts as a full-duplex (FD) relay for cellular uplink transmission. In addi-
tion, DT employs non-orthogonal multiple access (NOMA) to transmit
the superimposed cellular and D2D signal while utilizing power domain
multiplexing. Successive interference cancellation is applied at the D2D
receiver to decode the D2D signal. Further, we considered that each node
is equipped with directional antennas to compensate for the impact of
high propagation loss, and a sectored beamforming framework is used
to model the antenna gain. The analytical expressions for the achiev-
able rates and outage probabilities for cellular and D2D users with
an optimal value of the power splitting factor have been derived to
characterize the system performance. Our results include the impact
of various system parameters such as half-power beamwidth, sidelobe
1

2Telecommunication Systems 2023
gain, and residual self-interference constants on the cellular outage prob-
ability. We have also shown that our proposed model outperforms the
recent work on the NOMA-aided FD C-D2D communications system
by approximately 104times in terms of cellular outage probability.
Keywords: Millimeter-wave, full-duplex relaying, device-to-device
communications, non-orthogonal multiple access, outage probability
1 Introduction
The expeditious growth in the number of connecting users and high-speed
multimedia applications have augmented the need for technologies with ultra-
high data rates, low latency, and improved reliability. Beyond fifth-generation
(B5G) cellular networks promise to provide multi-gigabit, ultra-reliable, and
ultra-low latency connectivity for mobile users [1]. One of the core technology
defined in long-term evolution-advanced Release 12-15 [2] is device-to-device
(D2D) communications which eliminate the involvement of gNodeB (gNB) to
improve the spectral efficiency. D2D offloads data from the gNB and enables
direct communications between devices in close proximity, contrary to the
traditional cellular networks, where every transmission is routed through the
gNB.
Cooperative-D2D (C-D2D) [3] is a subclass of D2D communications where
D2D users coexist with the cellular users (CUs) in a cell. In a C-D2D commu-
nications system, a single or multiple D2D users act as amplify-and-forward
(AF) or decode-and-forward (DF) relay to forward cellular uplink/downlink
signal to the base station (BS). C-D2D is superior to the other two D2D
frameworks: underlay and overlay, in terms of spectral efficiency and outage
probability. However, it has some limitations in effectively utilizing cellular
resources with low outage probability. The integration of the C-D2D com-
munications system with various promising technologies such as full-duplex
(FD) relaying, millimeter-wave (mmWave), and non-orthogonal multiple access
(NOMA) improves the wireless coverage and capacity for mobile users with
enhanced spectral efficiency.
Most of the existing literature on C-D2D framework assumes that the
relay node operates in the half-duplex (HD) mode [4,5], which is not an
optimal duplex mode in the B5G cellular networks. The inclusion of FD relay-
ing into the D2D communications system doubles the bandwidth efficiency
due to the simultaneous transmission and reception (STAR) over the same
time-frequency resource block (TFRB) [6]. However, the self-interference (SI)
introduced due to STAR is a dominant disadvantage for FD D2D systems. A
proper implementation of the SI suppression (SIS) technique reduces SI up
to the noise floor and enables FD’s symbiosis in the D2D system [7,8]. How-
ever, due to some pragmatic imperatives, the SI cannot be fully suppressed,

Telecommunication Systems 2023 3
and the RSI is a restricting factor for FD operation as it reduces the received
signal-to-interference plus noise ratio (SINR) [9].
Similar to FD radios, NOMA is another promising technology that can
enhance spectral efficiency and provide massive connectivity in B5G cellu-
lar networks [10–13]. NOMA dominates the traditional orthogonal multiple
access (OMA) schemes as it facilitates fairness among users and has diverse
quality-of-service due to power domain multiplexing. In addition, as compared
to traditional OMA, NOMA can accommodate more users in the same time-
frequency-code resource block [14]. In the context of a C-D2D communications
system, cellular and D2D users use NOMA to access the same time and fre-
quency slot simultaneously to enhance the spectral efficiency of the system
[15]. NOMA superimposes multiple user data on the same TFRB by allocat-
ing more power to signals transmitted to users with poor channel conditions.
A multi-user detection algorithm, such as successive interference cancellation
(SIC), is applied at the receivers to decode the intended signals successfully
from the received superimposed signal [16].
The integration of mmWave with the C-D2D communications system
resolves the grave issue of maintaining higher data rates and low latency in
wireless systems. In [17,18], it has been shown that the influential addition of
NOMA in the mmWave communications system increases the overall through-
put. mmWave (30/60 GHz to 300 GHz) provides an abundant spectrum to
satisfy high capacity; however, a short wavelength results in high propagation
losses and path blockages. The high-frequency mmWave has short wavelengths
that result in high propagation losses and path blockages. However, the highly
directional antennas can be used to overcome the frequency-dependent path
loss and external interference, which results in an improvement in the data
rate and makes it suitable for short-range D2D transmission [19–21].
1.1 Related Works and Motivation
Although comprehensive literature is available for FD C-D2D communica-
tions system; however, the studies related to the integration of mmWave and
NOMA with the FD C-D2D communications system are limited. In [22], an
adaptive transmission mode selection algorithm is proposed for a C-D2D com-
munications system. Here, duplexing mode is selected based on the number
of subcarriers forwarded by the best D2D user to the BS. The performance
tradeoffs between both modes are presented by deriving the outage probabil-
ity closed-form expressions. However, the analytical derivations are tractable
due to the assumption of no direct link between CU and BS. Additionally, the
framework has not been tested for technologies such as mmWave and NOMA.
In [23], a cooperative NOMA and an orbital angular momentum (OAM)-
based multiple-input multiple-output multiplexing (MIMO) system with FD
relaying (CNOMF) is proposed, wherein the cell center user (CCU) acts as an
FD relay for the cell edge user (CEU). The sum capacity (SC) of the proposed
CNMOF system is compared with existing CNOMA-MIMO with FD relaying,
conventional OMA with the OAM-MIMO, and half-duplex relaying (known as

4Telecommunication Systems 2023
OMA-OAM-MIMO-HDR) techniques. The SC results show that the proposed
CNOMF scheme outperforms other cooperative schemes.
Further, in [24], an FD NOMA scheme for a cooperative relay sharing
(FD-NOMA-RS) network is proposed, wherein two source-destination pairs
share a dedicated FD relay utilizing NOMA. The analytical expressions of
ergodic SC, outage probability, and outage SC of FD-NOMA-RS are derived
considering both imperfect and perfect interference cancellation. Further, it
is shown that the proposed FD-NOMA-RS scheme substantially outperforms
the conventional HD-NOMA-RS scheme.
In [25], a cooperative NOMA downlink transmission scheme with improved
time switching simultaneous wireless information and power transfer (ITS
SWIPT), referred to as CNOMA-SWIPT-ITS, is proposed. In the CNOMA-
SWIPT-ITS scheme, CCU harvests the energy from the BS and acts as a
decode-and-forward (DF) relay for cellular downlink transmission towards
the CEU. Further, the analytical expressions of ergodic SC and outage
probabilities are derived, and the results are compared with conventional
CNOMA-based SWIPT schemes and OMA-based ITS SWIPT scheme. How-
ever, the proposed CNOMA-ITS-SWIPT scheme is limited to HD relaying, and
the advantages of FD relaying are not explored. Further, the proposed model
in [23–25] works on conventional cellular downlink systems, and the advan-
tages of sidelink D2D communications underlaying the cellular system are not
considered. Moreover, the proposed CNOMF and FD-NOMA-RS schemes do
not utilize the prospective benefits of the mmWave.
In [26], a downlink NOMA-based C-D2D system is proposed, wherein an
efficient power allocation scheme is provided for D2D transmission to improve
the spectral efficiency of the proposed system. Further, it has been shown that
the NOMA-based system outperforms the conventional OMA-based C-D2D
system. Similarly, a C-D2D system with multiple D2D pairs is proposed in [27],
wherein a D2D transmitter (DT) acts as an HD relay for the cellular uplink
transmission. Additionally, an optimal power and spectrum allocation scheme
which maximizes the achievable D2D rate with a cellular outage constraint is
provided for the D2D users. However, the performance in [26,27] is limited
to HD relay only, and the prospective capacity improvement due to the FD
relaying has not been explored. In [28], a NOMA aided FD C-D2D system is
proposed wherein CU acts as an FD relay for D2D transmission. In addition,
two power allocation schemes are proposed to maintain the fairness between
cellular and D2D users. The exact and asymptotic expressions of joint D2D
and cellular outage probability are derived for the proposed FD C-D2D system.
However, in [26–28], the performance improvement and the system complexity
due to mmWave have not been discussed.
In [29], a mmWave assisted FD C-D2D communication system utilizing
NOMA is proposed, wherein DT acts as an FD relay for the cellular uplink
transmission. Further, the closed-form expressions of D2D and cellular outage
probability have been derived by considering the most probable directive gain
scenarios. However, the analytical complexity of the system model has not been

Telecommunication Systems 2023 5
explored due to no direct link assumptions between CU and BS. In addition,
the proposed model is limited to a single CU and D2D pair within the cell,
and the obtained results were not compared with any benchmark frameworks.
Hence, there is a need to explore the mmWave assisted FD C-D2D system
wherein multiple D2D pairs are achieving potential diversity gain due to a
direct cellular link between CU and BS to improve the performance of the FD
C-D2D system.
To the best of our knowledge, the benefits of NOMA and mmWave for
multiuser FD C-D2D communications system have not been explored so
far. Motivated by the above, this paper proposes the multiuser NOMA and
mmWave aided FD C-D2D communications system wherein FD node DT acts
as a DF relay for cellular uplink transmission. Since we consider the transmis-
sions over mmWave spectrum, directional antennas are used at each node to
avoid the high propagation losses. Analytical expressions of achievable rate and
outage probability for D2D and cellular transmissions are derived to analyze
the performance of the proposed system. The derived expressions are validated
through simulations, and the outage performance of the proposed framework
is analyzed by varying different system parameters to gain insights into the
behaviour of the mmWave and NOMA in multiuser FD C-D2D system.
1.2 Major Contributions
The major contributions of this paper can be summarised as follows:
•Integration of mmWave and NOMA: A mmWave and NOMA
assisted FD C-D2D communications system is proposed, wherein DT
works as an FD relay for cellular uplink transmission. The analytical
expressions of achievable rates and outage probabilities have been derived
for both cellular and D2D users to evaluate the performance of the pro-
posed model. A rigorous mathematical analysis involving a combination of
multiple continuous and discrete random variables (RVs) is performed for
the derivations of outage probabilities expressions. In addition, a sectored
beamforming model for the directional antennas has been considered to
overcome the high propagation losses due to mmWave.
•Optimal power allocation: An optimal value of the power splitting
factor (α∗)that minimizes the D2D outage probability is determined. α∗
decides the optimal power allocated to NOMA-strong and NOMA-weak
users while multiplexing cellular uplink and D2D signals in the power
domain.
•Impact of system parameters: The impact of various system parame-
ters such as power splitting factor, half-power beamwidth (HPBW) angle,
RSI parameters, antenna lobe gain, and transmission distances on the
D2D and cellular outage probabilities are analyzed.
•Performance comparison with the conventional systems: We have
compared the performance of our proposed mmWave and NOMA assisted
multiuser FD C-D2D systems in terms of the outage performance with
the following three conventional frameworks:

6Telecommunication Systems 2023
Fig. 1 System model
– mmWave and NOMA assisted HD C-D2D systems
– mmWave assisted HD C-D2D systems without NOMA
– mmWave and NOMA assisted FD C-D2D system with no direct
cellular link
It has been shown that the proposed system outperforms all three
frameworks mentioned above.
1.3 Article Outline
The remaining part of the paper is structured as follows. Section 2 describes
the system model consisting of BS, CU, and D2D pairs. In Section 3, achiev-
able rates and outage probabilities have been derived for cellular and D2D
users. Section 4 provides the analytical and simulation results for the proposed
systems. Section 5 concludes the paper.
2 System Model
Fig. 1shows the proposed FD C-D2D communications system consisting of a
BS, PCUs, and MD2D pairs. Please note that a D2D pair refers to a DT and
corresponding D2D receiver (DR). For a specific TFRB, each CU maps with a
single D2D pair. BS is responsible for the time-frequency synchronization, and
peer discovery using primary and secondary synchronization signals [30]. Since
we consider the mmWave spectrum for the transmission, directional antennas
are used at the transmitting and receiving nodes to combat high-frequency
losses. We consider a sectored beamforming pattern to model the antenna gain
at each node.
In our proposed FD C-D2D system, DT acts as an FD relay for cellular
uplink transmission, which utilizes DF relaying protocol to forward the cellular
uplink signal to BS. The noise at the receiving nodes DT, DR, and BS is
considered as additive white Gaussian noise (AWGN) and is denoted by n∼
CN(0, σ2). The channels between node iand node j, where, i∈{CU, DT}

Telecommunication Systems 2023 7
Table 1 List of variables and their description
Variables Description
MNumber of D2D pairs
PNumber of CUs
pcu Power transmitted by CU
pdt Power transmitted by DT
σ2Noise variance
RTCellular target rate
hi,j Channel coefficient corresponds to a transmitter
receiver pair
Gi,j Antenna gain corresponds to a transmitter receiver
pair
ρi,j Distance between any transmitter receiver pair
xc(t)Cellular user signal
xdt(t)D2D user signal
lPath loss exponent
αPower splitting factor
θHAngle of half-power beamwidth
β,ζRSI parameters
νProbability that beam covers an angle θH
KMain lobe gain
kSide lobe gain
PCU
Out Outage probability for CU Case
PD2D
Out Outage probability for D2D Case
Pout
BS,1Outage probability due to CU-DT link
Pout
BS,2Outage probability due to CU-BS and DT-BS link
Pout
BS,3Outage probability due to CU-BS link
Pout
DR,1Outage probability in SIC while decoding CU
signal
Pout
DR,2Outage probability in SIC while decoding D2D sig-
nal
and j∈{DT, DR, BS} is denoted by hi,j. All the channels are assumed to
be Rayleigh flat fading with hi,j ∼ CN(0, λi,j ), where, λi,j =ρ−l
i,j .ρi,j is the
distance between two nodes defined separately for each transmitter-receiver
pair, and the path-loss exponent is denoted as l. The channel gain between any
two nodes i,jis exponentially distributed, i.e., |hi,j |2∼exp (λi,j). Kindly
note that xcand xdt have zero mean and E{x∗
cxc}=E{x∗
dtxdt }= 1.

8Telecommunication Systems 2023
2.1 Sectored Beamforming Framework
θ
(a) Case 1
(c) Case 3 (d) Case 4
(b) Case 2
Kt
kt
θ
Kr
kr
kr
kt
kr
Kr
Kr
Kr
Kt
Kt
Kt
kt
kt
kr
θ
θ
θ
θ
θ
θ
Fig. 2 Beamforming possibilities
In our proposed model, we consider that each node is transmitting over the
mmWave spectrum, and hence there will be high propagation losses. To over-
come these losses, we consider a sectored beamforming framework with four
possible cases, as shown in Fig. 2. The antennas have a constant gain Kin the
main lobe spread over a beam of angle θ, and a constant lobe gain kfor the
side lobe spread over (2π−θ). The lobe gain for a specific node sis defined as,
Ms(θ) = (Ks,0◦≤θ < θH
ks, θH≤θ < 180◦.(1)
where, θHis the angle of HPBW. The antenna gain between node iand jis
Gij =Mi(θij )Mj(θji ); where, θij represents the departure angle from node i
to node jand, θji denotes the angle of arrival from node jto node i.
Table 2 DIRECTIVE GAIN CASES OF ANTENNAS
Case Antenna Gain (Gij) Probability
1KtKrν2
2Ktkrν(1 −ν)
3ktKr(1 −ν)ν
4ktkr(1 −ν)2
The antenna gain Gij is a discrete random variable with probability mass
function (PMF) as shown in Table 2. Here, subscripts t, and rdenotes trans-
mitter and receiver, respectively while νdenotes the probability that beam
covers an angle equal to θH. From [31,32], ν=θH
2π.

Telecommunication Systems 2023 9
2.2 Residual Self-Interference
RSI at DT is modeled as an additive Gaussian RV, which is denoted by ϕs,
ϕs∼CN 0, β pζ
dt,(2)
where, pdt denotes the transmit power of DT and pdt ∈(0, Pmax].Pmax is the
maximum value of pdt whereas RSI parameters β , ζ are non-negative constant
reflecting the performance of SIS.
2.3 Protocol Description
In our proposed FD C-D2D communications system, BS allocates the resources
between the cellular and D2D users and supervises the signaling and control
information [33]. Initially, CU broadcasts a cellular uplink signal over mmWave
frequency, which is received by the BS and the FD node DT. DT utilizes DF
relaying protocol and decodes the cellular uplink signal. Further, DT employs
NOMA to superimpose the decoded cellular uplink signal and D2D signal with
different power levels before transmitting it to the BS and DR over mmWave
frequency. In our proposed model, we have considered that DR is closer to DT
than BS1; hence the effective channel gain of the DT-DR link is more than
that of the DT-BS link. Thus, BS is termed as a NOMA-weak user, while
DR is termed as a NOMA-strong user. According to the principle of NOMA,
DT employs power domain multiplexing to allocate more power to the cellular
uplink signal than the D2D signal. DR decodes the D2D signal by assuming
cellular uplink signal as noise and cancels it using SIC to decode the D2D
signal. The FD node DT transmits and receives on the same TFRB; hence it
will be affected by SI. However, after active and passive SI cancellation, the
RSI can be modeled as (2). The BS combines the signals received from two
different paths using maximal ratio combining (MRC).
3 Rate and Outage Probability
This section derives the analytical expressions of the achievable rates and
outage probabilities of CU and D2D user.
3.1 Outage analysis for cellular uplink transmission
CU broadcasts xc(t)with power pcu, which is received by BS, DT and DR.
The cellular signal received at DT is given as:
ydt(t) = √pcu hcu,dt xc(t) + ϕs(t) + ndt .(3)
1This is the most probable scenario since DT-DR exists as a proximate pair. In an environment
such as inside a building or a playground, there are fairly high chances of DR being closer to DT
as compared to BS.

10 Telecommunication Systems 2023
The instantaneous rate at DT is given as
Rcu,dt = log2 1 + Gcu,dt pcu |hcu,dt |2
βpζ
dt +σ2!,(4)
where, βP ζ
DT denotes RSI at DT. When RTdenotes the target rate, the outage
probability for CU-DT link is defined as,
Pout
BS,1= Pr[Rcu,dt < RT]
= Pr "log2 1 + Gcu,dt pcu |hcu,dt |2
βpζ
dt +σ2!< RT#
= Pr "Gcu,dt |hcu,dt |2<(2RT−1) βpζ
dt +σ2
pcu #.(5)
Gcu,dt is a discrete random variable with four possible values as defined in
Table 2. On further soling (5),
Pout
BS,1=
4
X
m=1 Pr |hcu,dt |2<ρ1
xm×(Pr[Gcu,dt =xm]),(6)
where,
ρ1= (2RT−1) βpζ
dt +σ2
pcu !
xm∈[KtKr, Ktkr, ktKr, ktkr].(7)
Since, |hcu,dt |2∼exp(λcu,dt), and ρ1
xm>0, using (6), and (7),
Pout
BS,1=
4
X
m=1
[1 −e
−ρ1
xmλcu,dt ]×Pr[Gcu,dt =xm].(8)
After decoding signal ydt(t), DT superimposes it with the D2D signal using
NOMA, and the superimposed signal is represented as,
Xdt(t) = √αpdt xc(t−to) + p(1 −α)pdt xdt (t),(9)
where, xc(t−to)is the delayed version of xc(t),tois the delay in decoding
and processing, and α∈(0.5,1). The NOMA-weak user BS receives both the
uplink signal from CU and the superimposed signal from DT, which is denoted
as:
ybs(t) = √pcu hcu,bs xc(t) + hdt,bs Xdt (t) + nbs(t).(10)

Telecommunication Systems 2023 11
The instantaneous rate at BS is given as
Rcu,bs = log21 + Gcu,bs pcu |hcu,bs |2
σ2.(11)
The outage probability for CU-BS link is given as
Pout
BS,3= Pr[Rcu,bs < RT](12)
= Pr Gcu,bs |hcu,bs |2< ρ2,
where, ρ2= (2RT−1) σ2
pcu , and Gcu,bs is a discrete RV with same PMF as
defined in Table 2. Further,
Pout
BS,3=
4
X
m=1 Pr |hcu,bs |2<ρ2
xm×Pr(Gcu,bs =xm)
=
4
X
m=1 1−e
−ρ2
xmλcu,bs ×Pr(Gcu,bs =xm).(13)
From (10), it can be seen that both the signals xc(t)and xDT (t)are fully
resolvable at BS, and hence by utilizing MRC at BS, the combined SINR can
be represented as
γMRC =γbs,1+γbs,2,(14)
where,
γbs,1=Gcu,bspcu |hcu,bs |2
σ2,(15)
γbs,2=αGdt,bspdt |hdt,bs |2
σ2+ (1 −α)Gdt,bspdt |hdt,bs |2,(16)
where, |hdt,bs |2,|hcu,bs |2are exponentially distributed RVs with mean λdt,bs
and λcu,bs, respectively. The instantaneous rate at BS
RMRC
bs = log2(1 + γM RC ).(17)
Substituting (14), (15) and (16) into (17),
RMRC
bs = log21 + Gcu,bs pcu |hcu,bs |2
σ2
+αGdt,bspdt |hdt,bs |2
σ2+ (1 −α)Gdt,bspdt |hdt,bs |2.(18)

12 Telecommunication Systems 2023
The cellular outage probability is given as,
Pout
BS,2= Pr[RM RC
bs < RT]
= Pr[γbs,1< γth −γbs,2].(19)
where, γth = 2RT−1.(19) involves two continuous and two discrete RVs.
Considering, Gcu,bs =xm, and Gdt,bs =xn,
Pout
BS,2=
4
X
m=1
4
X
n=1
Pmn Zy0
0fγbs,2(y)×Fγbs,1(ω)dy, (20)
where, fγbs,2(y)represents the probability density function (PDF) of γbs,2, and
Fγbs,1(ω)represents the cumulative distribution function (CDF) of γbs,1.Pmn
is the joint probability defined as
Pmn = Pr[Gcu,bs =xm, Gdt,bs =xn].(21)
Since, Gcu,bs and Gdt,bs are independent discrete RVs with PMF defined as
given in Table 2,Pmn is further simplified as
Pmn = Pr[Gcu,bs =xm]×Pr[Gdt,bs =xn],(22)
where, xmand xnare antenna gain defined similar to (7). Further, from (16),
γbs,2can be represented as
γbs,2=α pdtxn|hdt,bs |2
(1 −α)pdtxn|hdt,bs |2+σ2.(23)
The CDF of γbs,2can be written as
Fγbs,2(y) = Pr[γbs,2< y]
= Pr |hdt,bs |2<σ2y
α pdtxn−(1 −α)pdt xny
= 1 −e−nσ2y
pdtxnλdt,bs (α−(1−α)y)o.(24)
Using (24), PDF of γbs,2
fγbs,2(y) = α σ2
pdtxnλdt,bs (α−(1 −α)y)2en−σ2y
pdtxnλdt,bs (α−(1−α)y)o.(25)
The CDF of γbs,1is given by,

Telecommunication Systems 2023 13
Fγbs,1(ω) = Pr[γbs,1< ω]=1−e
−σ2ω
pcuxnλcu,bs ,
= 1 −e
−σ2
pcuxnλcu,bs γth −α pdtxny
σ2+(1−α)pdtxny,(26)
Further, from (19),
γth −γbs,2>0.(27)
Using (23) and (27),
γth >α pdtxn|hdt,bs |2
(1 −α)pdtxn|hdt,bs |2+σ2,
γthσ2>|hdt,bs |2pdt xn[α−γth(1 −α) ],
or,|hdt,bs |2< y0,(28)
where,
y0=γthσ2
pdtxn(α−γth (1 −α)) .(29)
Using (20), (25), (26) and (29),
Pout
BS,2=
4
X
m=1
4
X
n=1
Pmn Zy0
0α σ2
pdtxnλdt,bs (α−(1 −α)y)2
×e−σ2y
pdtxnλdt,bs (α−(1−α)y)
1−e
−σ2
pcuxnλcu,bs γth −α pdtxny
σ2+(1−α)pdtxnydy. (30)
Solving further,
Pout
BS,2=
4
X
m=1
4
X
n=1
Pmn Zy0
0{I1(y)−I2(y)}dy, (31)
where,
I1(y) =α σ2e−σ2y
pdtxnλdt,bs (α−(1−α)y)
pdtxnλdt,bs (α−(1 −α)y)2,

14 Telecommunication Systems 2023
I2(y) =α σ2e−σ2y
pdtxnλdt,bs (α−(1−α)y)
pdtxnλdt,bs (α−(1 −α)y)2
×e
−σ2
pcuλcu,bs xnγth −α pdtxny
σ2+(1−α)pdtxny.(32)
Using substitution, let
u=−σ2y
pdtxnλdt,bs (α−(1 −α)y)=⇒
dy =1
−σ2y
pdtxnλdt,bs (α−(1−α)y)−σ2(1−α)y
pdtxnλdt,bs (α−(1−α)y)2
du. (33)
Using (33), integration of I1can be given as,
Zy0
0
I1(y)dy =−Zσ2y0
pdtxnλdt,bs ((1−α)y0−α)
0
eudu
=1 −e
σ2y0
pdtxnλdt,bs ((1−α)y0−α).(34)
Without loss of generality, the closed-form of I2(y)can be obtained using some
approximations. The maximum value of (1 −α)pdtxnyis (1 −α)pdt xny0, and
can be given as,
(1 −α)pdtxny0=(1 −α)σ2
2α−1.(35)
It is evident from (35), (1−α)pdt xny < σ2. Similarly, (1−α)pdt xny < α pdt xn
and pcuλcu,bs xn≈λdt,bs , and hence the integration of I2(y)can be obtained
as,
Zy0
0
I2(y)dy
=Zy0
0
σ2
α pdtxnλdt,bs
e−σ2
λdt,bs y
α pdtxn+γth −α pdtxny
σ2dy,
=σ2e−σ2γth
λdt,bs
σ2−(α pdtxn)21−e−σ2y0
λdt,bs 1
α pdtxn−α pdt xn
σ2.(36)
The final outage probability for CU is given as
PCU
Out =Pout
BS,1×Pout
BS,3+ (1 −Pout
BS,1)×Pout
BS,2
=
4
X
m=1 1−e
−ρ1
xmλcu,dt Pr (Gcu,dt =xm)

Telecommunication Systems 2023 15
×
4
X
m=1 1−e
−ρ2
xmλcu,bs Pr(Gcu,bs =xm)
+"1−
4
X
m=1 1−e
−ρ1
xmλcu,dt Pr (Gcu,dt =xm)#
×
4
X
m=1
4
X
n=1
Pmn 1−e
σ2y0
pdtxnλdt,bs ((1−α)y0−α)
−σ2e−σ2γth
λdt,bs
σ2−(α pdtxn)21−e−σ2y0
λdt,bs 1
α pdtxn−α pdt xn
σ2
(37)
3.2 Achievable D2D rate and outage probability
This subsection includes the derivation of achievable rate and outage proba-
bility expressions for D2D user. The superimposed signal transmitted by FD
node DT will be received at DR as
ydr(t) = hdt,dr Xdt (t) + hcu,dr √pcu xc(t) + ndr (t).(38)
DR receives interference from CU and DT-BS signal. The instantaneous rate
at DR is:
Rdr,1= log21 + αGdt,dr pdt |hdt,dr |2
σ2+δ1+δ2,(39)
where,
δ1= (1 −α)Gdt,dr pdt |hdt,dr |2,
δ2=Gcu,drpcu |hcu,dr |2.
A multi-user detection technique SIC is employed at DR to decode the D2D
signal. As per SIC, DR initially decodes the signal xc(t)assuming xDT (t)as
an interference. The probability that DR is unable to decode xc(t)is given as
Pout
DR,1= Pr[Rdr,1< RT]
= Pr{| hdt,dr |2<Υ1+ Υ2|hcu,dr |2},(40)
where,
Υ1=(2RT−1)σ2
pdtGdt,dr [α−(1 −α)(2RT−1)] ,
Υ2=(2RT−1)pcuGcu,dr
pdtGdt,dr [α−(1 −α)(2RT−1)] .
The joint PDF of two independent and exponential RV |hdt,dr |2and
|hcu,dr |2is given as,

16 Telecommunication Systems 2023
1
λcu,dr
.e
−|hcu,dr|2
λcu,dr ×1
λdt,dr
.e
−|hdt,dr|2
λdt,dr .(41)
Further, from (40),
Pout
DR,1= Pr{| hdt,dr |2<Υ1+ Υ2× | hcu,dr |2}
=
4
X
p=1
4
X
q=1
Ppq Z+∞
0ZΥ1+Υ2|hcu,dr|2
0
1
λcu,dr
.e
−|hcu,dr|2
λcu,dr
×1
λdt,dr
.e
−|hdt,dr|2
λdt,dr d|hdt,dr |2d|hcu,dr |2
=
4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ1
λdt,dr
λdt,dr + Υ2λcu,dr 
,(42)
where, Ppq is defined as,
Ppq = Pr[Gcu,dr =xp]×Pr[Gdt,dr =xq].
After successful decoding of xc(t), DR subtracts xc(t)from the received
signal ydr(t)to decode D2D signal xDT (t). The instantaneous rate at DR after
subtracting xc(t)is given as,
Rdr,2= log21 + (1 −α)pdtGdt,dr |hdt,dr |2
σ2+pcuGcu,dr |hdt,dr |2.(43)
Probability that D2D signal could not be decoded at DR is given as,
Pout
DR,2= Pr[Rdr,2< RT]
Pout
DR,2= Pr[(1 −α)pdtGdt,dr |hdt,dr |2<(2RT−1)]
×[(σ2+pcuGcu,dr |hdt,dr |2)]
= Pr{| hdt,dr |2<Υ3+ Υ4|hdt,dr |2},(44)
where,
Υ3=(2RT−1)σ2
(1 −α)pdtGdt,dr
,(45)
Υ4=(2RT−1)pcuGcu,dr
(1 −α)pdtGdt,dr
,(46)
From (44),
Pout
DR,2=
4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ3
λdt,dr
λdt,dr + Υ4λcu,dr 
.(47)

Telecommunication Systems 2023 17
Hence, the overall outage probability expression for D2D User is given as
PD2D
Out =Pout
DR,1+ (1 −Pout
DR,1)×Pout
DR,2
=
4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ1
λdt,dr
λdt,dr + Υ2λcu,dr 

+


1−
4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ1
λdt,dr
λdt,dr + Υ2λcu,dr 



×
4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ3
λdt,dr
λdt,dr + Υ4λcu,dr 
.(48)
3.3 Corollary
From (48), we can infer that D2D outage probability depends on α. Hence, an
optimal value of αthat minimizes the D2D outage probability can be obtained
as
α∗= argmin
α
PD2D
out
subject to α∈(0.5,1),(49)
where, α∗is the optimal value of α, and PD2D
out is given as,
PD2D
out =


4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ1
λdt,dr
λdt,dr + Υ2λcu,dr 



+



1−
4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ1
λdt,dr
λdt,dr + Υ2λcu,dr 



×
4
X
p=1
4
X
q=1
Ppq 
1−λdt,dre
−Υ3
λdt,dr
λdt,dr + Υ4λcu,dr 
.(50)
The optimization problem in (49) has been solved using one-dimensional min-
imizer function fminbnd of MATLAB. Solving (49) gives α∗= 0.7071, which
minimizes the D2D outage probability2.
2This value is obtained by considering the most probable directive gain scenarios of the
interfering nodes involved.

18 Telecommunication Systems 2023
4 Results
This section presents the analytical and simulation results for the proposed
NOMA-aided, mmWave assisted FD C-D2D communication system. The cellu-
lar and D2D outage probabilities are plotted with respect to various parameters
such as the distance between nodes and power allocation factor α. Path loss
component is taken as l= 4 (urban areas). The target rate for both the cel-
lular and D2D communication is set to be RT=1bit/sec/Hz. The simulation
parameters are listed in Table 3.
Table 3 Simulation Parameters
Parameters Values Parameters Values
Kt10 dB Kr10 dB
kt−1.5 dB kr−1.5 dB
Rt1bit/sec/Hz l4
pdt 10 dBm pcu 0 dBm
σ2
−120 dBm θH10◦
ζ0.8β10−16
4.1 Optimal Power Allocation Analysis
Fig. 3represents the D2D outage probability PD2D
Out with respect to αranging
from 0.5 to 1 for dDT ,DR =30m, 65m and 100m keeping dCU,DR =400m.
It can be observed from Fig. 3that PD2D
Out initially decreases with increase
in αup to α∗=0.7071 and beyond this PD2D
Out increases. For instance, when
dDT ,DR =65m, PD2D
Out =9.47×10−4for α=0.6. Subsequently, probability
decreases to 7.38×10−4for α= 0.7and again PD2D
Out increases to 1.41×10−3
for α= 0.9. The reason for this behavior of D2D outage probability can be
understood from (48) which consist of PDR,1and PDR,2. With an increase in
αup to the value α∗=0.7071, Pout
DR,1in (48) dominates thereby decreasing
PD2D
Out , thereafter, Pout
DR,2dominates which results in an increase in PD2D
Out .
4.2 Parametric sweep analysis
Fig. 4represents the cellular outage probability versus ζfor different values of
dCU,DT =200m, 400m and 600m. The value of βis set to 10−12.5. It is observed
from Fig. 4that as ζincreases from 0.4 to 1.6, the cellular outage probability
decreases for dCU,DT . For instance, when dCU,DT =400m, PC U
out =7.56×10−2,
3.95×10−2and 2.24×10−2for ζ=0.8, 1 and 1.2, respectively. However, it can
be seen that for a further increase in value of ζ, outage probability is almost
stagnant. This can be explained by the fact that as ζincreases beyond 1.6,
the RSI value falls below the noise floor, and hence has negligible impact on
the outage probability. In addition, we can also infer from Fig. 4that PCU
out
increases with distance for all values of ζ.

Telecommunication Systems 2023 19
0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95
10-5
10-4
10-3
10-2
10-1
Fig. 3 D2D Outage probability vs α
0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
10-4
10-3
10-2
10-1
100
Fig. 4 Parametric analysis of ζ
In Fig. 5the cellular outage probability versus HPBW angle has been
plotted for dCU,DT =200m, 400m, 600m. The value of HPBW angle on the
X−axis ranges from 10◦to 90◦[34],[35],[36]. It can be inferred from the plot
that as value of θHincreases the outage probability decreases. Specifically,
PCU
out takes the value 1.03×10−2for θH=10◦and reduces to 8.29×10−3and
5.89×10−3when θH=30◦and 60◦, respectively. Table 2, and (1) show that
the probability of Case 1 is directly proportional to HPBW angle (θH). Hence,
as θHincreases, probability for high antenna gain increases resulting in lower
outage probability values. Similar to other plots, for a specified value of θH,
the cellular outage probability increases with increase in dCU,DT .

20 Telecommunication Systems 2023
10 20 30 40 50 60 70 80 90
10-4
10-3
10-2
10-1
Lines : Analytical
Markers : Simulation
Fig. 5 Parametric analysis of θH
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
10-5
10-4
10-3
10-2
10-1
100
Fig. 6 Parametric analysis of side lobe gain
Fig. 6shows the cellular outage probability versus the side lobe gain (ki) for
three different values of dC U,DT . The main lobe gain is set as Ki= 10dB. The
value of θHis set as 10◦, and accordingly, the antenna gain probabilities are
computed for the four cases as shown in Table I. It can be observed that as the
value of side lobe gain increases, the cellular outage probability decreases which
can be explained by the probability values of the four cases. Since the value of
θH=10◦is small, the corresponding probability of Case 4 is large (0.945). As
a result, the beamforming pattern follows Case 4 with high probability, where
the side lobes of the transmitter and receiver are aligned. Hence, as the side
lobe gain product increases, the outage probability decreases. For instance,
PCU
out =6.50×10−3, 2.78×10−3, 1.37×10−3for ki= 0.8, 1, 1.2 while keeping

Telecommunication Systems 2023 21
dCU,DT =400m. Further, Fig. 6also shows that, cellular outage probability
increases with dCU,DT for all values of ki.
4.3 Comparison with conventional HD C-D2D system
300 350 400 450 500 550 600 650 700
10-6
10-5
10-4
10-3
10-2
10-1
100
Fig. 7 Comparison of the proposed model with mmWave based HD C-D2D system without
NOMA
Fig. 7, shows a comparison of the proposed model with the conventional
HD C-D2D sytems without NOMA in terms of cellular outage probability.
The cellular outage probability for dCU,BS ranging from 300m to 700m has
been plotted with dCU,DT =200m, 300m and 400m. It can be clearly observed
from Fig. 7that our proposed system outperforms the conventional HD C-
D2D systems without NOMA. Particularly, when dCU,B S =500m the outage
probability for HD relaying without NOMA for dCU,DT =200m, 300m, and
400m is 1.16×10−2, 5.64×10−2, and 1.58×10−1, repectively. In contrast, for
the proposed FD C-D2D system with NOMA, the outage probability is as low
as 3.39×10−4, 1.7×10−3, and 5.29×10−3, respectively for the same values of
dCU,DT .
Fig. 8illustrates the cellular outage probability comparison when the DF
relay DT, works in HD and FD mode. The outage probability is plotted against
dCU,BS from 300m to 700m for varying values of dCU,DT . The value of α= 0.85
has been considered for plotting since HD mode limits the range of α∈(0.75,1)
as compared to FD where αoperates in the range 0.5 to 1. A similar trend
for both FD and HD relaying can be observed in terms of outage probability
with an increase in dCU,DT and dC U,BS . However, it is evident from Fig. 8
that the proposed FD C-D2D systems outperforms the conventional HD C-
D2D systems for all values of dCU,DT =200m, 300m and 400m. Specifically,
for dCU,BS =500m and dC U,DT =400m , the cellular outage probability is

22 Telecommunication Systems 2023
300 350 400 450 500 550 600 650 700
10-5
10-4
10-3
10-2
10-1
100
Fig. 8 Comparison of the proposed model with the conventional HD C-D2D system
4.01×10−2for HD and 5.29×10−3for FD. Similarly when dCU,DT =200m and
300m, the outage probability is 2.69×10−3and 1.33×10−2for HD. In contrast,
outage probability is 3.39×10−4and 1.7×10−3for FD which are significantly
less as compared to the HD mode.
4.4 Comparison with recent literature
100 150 200 250 300 350 400 450 500 550 600
10-10
10-8
10-6
10-4
Fig. 9 Comparison with [29] in terms of cellular outage probability vs dCU,DT
Fig. 9shows a comparison of the cellular outage probability of the proposed
system model with [29] where the direct link between CU-BS is not available.
In [29], the cellular outage probability with respect to dCU,DT has been plotted

Telecommunication Systems 2023 23
for dDT ,BS =200m, 400m and 600m. In our proposed framework, due to
consideration of the direct CU-BS link, there is a significant decrease in cellular
outage probability as shown in Fig. 9. A direct CU-BS link leads to a stagnant
cellular outage probability with respect to dDT ,BS . Due to MRC at BS, the
outage probability Pout
BS,2is very low and hence has negligible contribution to
the final outage probability in (37) as compared to Pout
BS,1×Pout
BS,3which is
almost stagnant with respect to dDT ,BS . Hence, in our proposed framework,
we obtain almost similar outage probability for dDT,BS =200m, 400m, and
600m which has been represented by a green colored line in Fig. 9.
5 Conclusion
In this paper, a mmWave aided FD C-D2D communications system is pro-
posed, wherein FD node DT employs NOMA to transmit cellular uplink and
D2D signals in the same TFRB with different power levels. SIC is employed
at the DR, which decodes D2D signal considering cellular uplink signal as
an interference. The BS uses MRC to combine the signals received from two
different paths. Further, analytical expressions of cellular and D2D outage
probabilities are derived and verified by the simulations. Additionally, the
optimal value of αhas been evaluated, which minimizes the D2D outage prob-
ability. The impact of various system parameters such as λ,θH, side lobe
gain on cellular outage probability has been analyzed for different values of
dCU,DT . Finally, our proposed system model is compared with recent works
on mmWave aided FD/HD C-D2D communications systems with and without
NOMA, and it has been shown that our proposed framework outperforms all
the other conventional C-D2D systems.
References
[1] Samdanis, K., Taleb, T.: The road beyond 5g: A vision and insight of the
key technologies. IEEE Network 34(2), 135–141 (2020). https://doi.org/
10.1109/MNET.001.1900228
[2] Ansari, R.I., Chrysostomou, C., Hassan, S.A., Guizani, M., Mumtaz, S.,
Rodriguez, J., Rodrigues, J.J.P.C.: 5G D2D networks: Techniques, chal-
lenges, and future prospects. IEEE Systems Journal 12(4), 3970–3984
(2018). https://doi.org/10.1109/JSYST.2017.2773633
[3] Cao, Y., Jiang, T., Wang, C.: Cooperative device-to-device communica-
tions in cellular networks. IEEE Wireless Communications 22(3), 124–129
(2015). https://doi.org/10.1109/MWC.2015.7143335
[4] Gupta, N., Ashok Bohara, V.: Rate and outage trade-offs for OFDMA
based device to device communication frameworks. IEEE Access 5, 14095–
14106 (2017). https://doi.org/10.1109/ACCESS.2017.2727055

24 Telecommunication Systems 2023
[5] Gupta, N., Kumar, D., Bohara, V.A.: A novel user selection and resource
allocation framework for cooperative D2D communication. In: 2018 IEEE
Global Communications Conference (GLOBECOM), pp. 1–7 (2018).
https://doi.org/10.1109/GLOCOM.2018.8647818
[6] Zhang, Z., Long, K., Vasilakos, A.V., Hanzo, L.: Full-duplex wireless
communications: Challenges, solutions, and future research directions.
Proceedings of the IEEE 104(7), 1369–1409 (2016). https://doi.org/10.
1109/JPROC.2015.2497203
[7] Nwankwo, C.D., Zhang, L., Quddus, A., Imran, M.A., Tafazolli, R.: A
survey of self-interference management techniques for single frequency full
duplex systems. IEEE Access 6, 30242–30268 (2018). https://doi.org/10.
1109/ACCESS.2017.2774143
[8] Lee, Y., Huang, B.-Y.: Active interference cancellation for full-duplex mul-
tiuser networks with or without existence of self-interference. IEEE Access
7, 15056–15068 (2019). https://doi.org/10.1109/ACCESS.2019.2895160
[9] Duarte, M., Dick, C., Sabharwal, A.: Experiment-driven characterization
of full-duplex wireless systems. IEEE Transactions on Wireless Commu-
nications 11(12), 4296–4307 (2012). https://doi.org/10.1109/TWC.2012.
102612.111278
[10] Benjebbour, A., Saito, Y., Kishiyama, Y., Li, A., Harada, A., Naka-
mura, T.: Concept and practical considerations of non-orthogonal multiple
access (NOMA) for future radio access. In: 2013 International Sympo-
sium on Intelligent Signal Processing and Communication Systems, pp.
770–774 (2013). https://doi.org/10.1109/ISPACS.2013.6704653
[11] Chen, Z., Ding, Z., Dai, X., Zhang, R.: An optimization perspective of the
superiority of NOMA compared to conventional OMA. IEEE Transactions
on Signal Processing 65(19), 5191–5202 (2017). https://doi.org/10.1109/
TSP.2017.2725223
[12] Islam, S.R., Zeng, M., Dobre, O.A., Kwak, K.-S.: Nonorthogonal multiple
access (noma): How it meets 5G and beyond. Wiley 5G Ref: The Essential
5G Reference Online, 1–28 (2019)
[13] Zhang, Z., Ma, Z., Xiao, M., Ding, Z., Fan, P.: Full-duplex device-to-
device-aided cooperative nonorthogonal multiple access. IEEE Transac-
tions on Vehicular Technology 66(5), 4467–4471 (2017). https://doi.org/
10.1109/TVT.2016.2600102
[14] Islam, S.M.R., Avazov, N., Dobre, O.A., Kwak, K.-s.: Power-domain
non-orthogonal multiple access (NOMA) in 5G systems: Potentials and
challenges. IEEE Communications Surveys Tutorials 19(2), 721–742

Telecommunication Systems 2023 25
(2017). https://doi.org/10.1109/COMST.2016.2621116
[15] Bajpai, R., Kulkarni, A., Malhotra, G., Gupta, N.: Outage analysis of
ofdma based noma aided full-duplex cooperative d2d system. In: 2020 27th
International Conference on Telecommunications (ICT), pp. 1–5 (2020).
https://doi.org/10.1109/ICT49546.2020.9239456
[16] Zhang, X., Haenggi, M.: The performance of successive interference
cancellation in random wireless networks. IEEE Transactions on Infor-
mation Theory 60(10), 6368–6388 (2014). https://doi.org/10.1109/TIT.
2014.2341248
[17] Xiao, Z., Zhu, L., Choi, J., Xia, P., Xia, X.-G.: Joint power allo-
cation and beamforming for non-orthogonal multiple access (NOMA)
in 5G millimeter wave communications. IEEE Transactions on Wire-
less Communications 17(5), 2961–2974 (2018). https://doi.org/10.1109/
TWC.2018.2804953
[18] Zhu, L., Zhang, J., Xiao, Z., Cao, X., Wu, D.O., Xia, X.-G.: Joint
power control and beamforming for uplink non-orthogonal multiple access
in 5G millimeter-wave communications. IEEE Transactions on Wire-
less Communications 17(9), 6177–6189 (2018). https://doi.org/10.1109/
TWC.2018.2855151
[19] Sim, G.H., Loch, A., Asadi, A., Mancuso, V., Widmer, J.: 5G millimeter-
wave and D2D symbiosis: 60 GHz for proximity-based services. IEEE
Wireless Communications 24(4), 140–145 (2017). https://doi.org/10.
1109/MWC.2017.1600098
[20] Wang, X., Kong, L., Kong, F., Qiu, F., Xia, M., Arnon, S., Chen, G.:
Millimeter wave communication: A comprehensive survey. IEEE Commu-
nications Surveys Tutorials 20(3), 1616–1653 (2018). https://doi.org/10.
1109/COMST.2018.2844322
[21] Xiao, M., Mumtaz, S., Huang, Y., Dai, L., Li, Y., Matthaiou, M.,
Karagiannidis, G.K., Björnson, E., Yang, K., I, C.-L., Ghosh, A.: Mil-
limeter wave communications for future mobile networks. IEEE Journal
on Selected Areas in Communications 35(9), 1909–1935 (2017). https:
//doi.org/10.1109/JSAC.2017.2719924
[22] Bajpai, R., Gupta, N., Bohara, V.A.: An adaptive full-duplex/half-duplex
multiuser cooperative D2D communications system with best user selec-
tion. IEEE Open Journal of the Communications Society 2, 1445–1457
(2021). https://doi.org/10.1109/OJCOMS.2021.3091905

26 Telecommunication Systems 2023
[23] Amin, A.A., Shin, S.Y.: Capacity analysis of cooperative NOMA-OAM-
MIMO based full-duplex relaying for 6G. IEEE Wireless Communi-
cations Letters 10(7), 1395–1399 (2021). https://doi.org/10.1109/LWC.
2021.3068654
[24] Kader, M.F., Shin, S.Y., Leung, V.C.M.: Full-duplex non-orthogonal
multiple access in cooperative relay sharing for 5G systems. IEEE Trans-
actions on Vehicular Technology 67(7), 5831–5840 (2018). https://doi.
org/10.1109/TVT.2018.2799939
[25] Al Amin, A., Young Shin, S.: Performance analysis of cooperative
nonorthogonal multiple access with improved time switching simultane-
ous wireless information and power transfer protocol. Transactions on
Emerging Telecommunications Technologies 31(11), 4077 (2020). https:
//doi.org/10.1002/ett.4077
[26] Kim, J.-B., Lee, I.-H., Lee, J.: Capacity scaling for D2D aided coopera-
tive relaying systems using noma. IEEE Wireless Communications Letters
7(1), 42–45 (2018). https://doi.org/10.1109/LWC.2017.2752162
[27] Lee, J., Lee, J.H.: Performance analysis and resource allocation for coop-
erative D2D communication in cellular networks with multiple D2D pairs.
IEEE Communications Letters 23(5), 909–912 (2019). https://doi.org/
10.1109/LCOMM.2019.2907252
[28] Liu, G., Feng, W., Han, Z., Jiang, W.: Performance analysis and opti-
mization of cooperative full-duplex D2D communication underlaying cel-
lular networks. IEEE Transactions on Wireless Communications 18(11),
5113–5127 (2019). https://doi.org/10.1109/TWC.2019.2932982
[29] Bajpai, R., Nawandar, K., Nag, S., Gupta, N.: Outage analysis of millime-
ter wave assisted full-duplex cooperative D2D communications system
with non-orthogonal multiple access. In: 2020 IEEE International Confer-
ence on Advanced Networks and Telecommunications Systems (ANTS),
pp. 1–6 (2020). https://doi.org/10.1109/ANTS50601.2020.9342774
[30] Omri, A., Shaqfeh, M., Ali, A., Alnuweiri, H.: Synchronization procedure
in 5g nr systems. IEEE Access 7, 41286–41295 (2019). https://doi.org/
10.1109/ACCESS.2019.2907970
[31] Turgut, E., Gursoy, M.C.: Uplink performance analysis in D2D-enabled
millimeter-wave cellular networks with clustered users. IEEE Transactions
on Wireless Communications 18(2), 1085–1100 (2019). https://doi.org/
10.1109/TWC.2018.2889755
[32] Govenker, R.D., Phatak, A.Y., Bajpai, R., Gupta, N.: Outage analy-
sis of mmWave integrated device-to-device communication system under

Telecommunication Systems 2023 27
nakagami fading channel. In: 2020 National Conference on Communica-
tions (NCC), pp. 1–6 (2020). https://doi.org/10.1109/NCC48643.2020.
9056087
[33] Zhang, Z., Wu, Y., Chu, X., Zhang, J.: Resource allocation and power
control for D2D communications to prolong the overall system survival
time of mobile cells. IEEE Access 7, 17111–17124 (2019). https://doi.org/
10.1109/ACCESS.2019.2893378
[34] MacCartney, G.R., Rappaport, T.S., Rangan, S.: Rapid fading due to
human blockage in pedestrian crowds at 5G millimeter-wave frequencies.
In: GLOBECOM 2017 - 2017 IEEE Global Communications Conference,
pp. 1–7 (2017). https://doi.org/10.1109/GLOCOM.2017.8254900
[35] Zekri, A.B., Ajgou, R., Hettiri, M.: Impact of azimuth and elevation half
power beam width on human blockage scenarios in mmwave channels. In:
2020 1st International Conference on Communications, Control Systems
and Signal Processing (CCSSP), pp. 41–45 (2020). https://doi.org/10.
1109/CCSSP49278.2020.9151811
[36] El-Halwagy, W., Mirzavand, R., Melzer, J., Hossain, M., Mousavi,
P.: Investigation of wideband substrate-integrated vertically-polarized
electric dipole antenna and arrays for mm-wave 5G mobile devices.
IEEE Access 6, 2145–2157 (2018). https://doi.org/10.1109/ACCESS.
2017.2782083