Content uploaded by Zhisheng Yin
Author content
All content in this area was uploaded by Zhisheng Yin on Oct 15, 2023
Content may be subject to copyright.
1
UAV-assisted Secure Uplink Communications in
Satellite-supported IoT: Secrecy Fairness Approach
Zhisheng Yin, Member IEEE, Nan Cheng, Member IEEE, Yunchao Song, Member IEEE,
Yilong Hui, Member IEEE, Yunhan Li, Tom H. Luan, Senior Member IEEE, and Shui Yu, Fellow IEEE
Abstract—The escalating growth of the Internet of things
(IoT) has intensified the demand for dependable and efficient
communication networks to accommodate the massive data vol-
umes produced by interconnected devices. Satellite networks have
emerged as a promising alternative, particularly in remote and
underserved regions where terrestrial communication infrastruc-
tures are inadequate. Nevertheless, guaranteeing secure uplink
communications in satellite-based IoT networks is a daunting
task due to similar satellite channels and limited resources at
IoT nodes. In this paper, we explore the potential of unmanned
aerial vehicle (UAV) to improve the secrecy performance of uplink
transmissions in satellite-supported IoT networks. Specifically,
we first introduce a framework for UAV-aided secure uplink
communications, presuming a secure UAV-to-satellite connection.
To mitigate the risks of ground eavesdroppers intercepting uplink
transmissions, we develop a max-min secrecy rate optimization
problem with uplink power constraints. To address this non-
convex problem, a streamlined two-stage optimization approach
is proposed. In inner stage, we combine uplink power alloca-
tion and UAV beamforming and propose a successive convex
approximation (SCA) based joint optimization algorithm to
address them. In outer stage, we propose an synergized bisection
and coordinate descent algorithm to optimize UAV positioning.
Convergence is attained by alternating iterations between these
two stages. Particularly, the secrecy fairness among IoT users is
reached by solving the max-min problem. Additionally, we offer
a complexity analysis of the proposed algorithm and validate
the efficacy of the presented approach through comprehensive
simulation results.
Index Terms—IoT, satellite, UAV, secure uplink, secrecy rate.
I. INTRODUCTION
THe exponential growth of the Internet of Things (IoT) has
resulted in the substantial rise in interconnected devices,
producing vast amounts of data that necessitate efficient and
dependable communication networks for their transmission
[1], [2]. Satellite networks have surfaced as a feasible option
for IoT implementations, particularly in remote and under-
served regions where terrestrial communication infrastructures
This work was supported in part by the National Natural Science Foundation
of China (No. 62201432, 62071356, and 62101429), the National Natural
Science Foundation of Shaanxi Province under Grant 2022JQ-602, and the
Guangzhou Science and Technology Program under Grant 202201011732.
Z. Yin, N. Cheng, Y. Hui and Tom H. Luan are with State Key Lab. of ISN,
Xidian University, 710071, Xi’an, Shaanxi, China (e-mail: {zsyin, tom.luan,
ylhui}@xidian.edu.cn, dr.nan.cheng@ieee.org).
Y. Song is with the College of Electronic and Optical Engineering, Nanjing
University of Posts and Telecommunications, 210003, Nanjing, Jiangsu, China
(e-mail: songyc@njupt.edu.cn).
Y. Li is with the Shaanxi Transportation Holding Group Co., Ltd., 710065,
Xi’an, Shaanxi, China. (e-mail: lyh199433@126.com).
S. Yu is with the School of Computer Science, University of Technology
Sydney, Australia. E-mail: Shui.Yu@uts.edu.au.
Corresponding author: Nan Cheng.
are insufficient or absent [3]–[5]. Satellite communication sys-
tems provide several essential advantages for IoT applications,
such as extensive coverage, uninterrupted connectivity, and
resilience to disasters and infrastructure breakdowns [6], [7].
Recently, Low Earth Orbit (LEO) satellite networks have
gained significant interest in IoT domains due to their low la-
tency and improved signal quality compared to Geostationary
Earth Orbit (GEO) satellites [8]. The growing interest in LEO
satellite networks is further fueled by the ongoing deployment
of satellite constellations, e.g., SpaceX’s Starlink and OneWeb
etc., which aim to provide ubiquitous global connectivity [9].
However, the integration of satellite networks into the IoT
ecosystem presents several challenges, particularly in terms
of ensuring secure and reliable uplink communications [10]–
[12].
Physical layer security plays a crucial role in protecting
sensitive data transmitted by IoT devices from eavesdropping
and other cyber threats [13], [14]. Whereas traditional cryp-
tographic techniques depend on both key management and
computational ability, and may not be suitable for resource-
constrained IoT devices [15], [16]. Physical layer security
techniques, as key-free security approaches, exploit the in-
herent characteristics of the wireless communication channels
to ensure data confidentiality and integrity without relying
on complex encryption algorithms [17], [18]. However, im-
plementing secure communications in satellite-supported IoT
networks is challenging due to the long propagation delays,
high mobility, and varying channel conditions associated with
satellite communications [19], [20]. Moreover, for the uplink
transmission, the resource usage of individual nodes is subject
to limitations, e.g., power and antennas. This factor compli-
cates the deployment of intricate signal processing techniques.
Consequently, this scenario poses substantial challenges to-
wards establishing secure transmission. [21]
Unmanned aerial vehicles (UAVs), as the aerial reinforce-
ment, have been the versatile and cost-effective tool to enhance
wireless communication systems, including satellite-supported
IoT networks [22], [23]. Their flexibility in deployment and
ability to provide line-of-sight (LoS) connections make them
well-suited for supporting secure uplink communications. By
acting as aerial relay nodes, UAVs can enhance the physical
layer security and extend the coverage of satellite communica-
tion systems, thereby improving the overall performance and
robustness of IoT networks [24], [25]. However, the related
work on secure uplink transmission is still relatively lacking.
Considering the inherent mobility of UAVs, the resource limi-
tations of IoT nodes, and the complexities arising from uplink
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
2
interference, achieving secure uplink transmission presents
significant challenges in the wireless security domain. These
pressing concerns have served as the catalyst for the rigorous
and scholarly investigation conducted in this paper.
Departing from existing researches on uplink physical layer
secure transmission, this study does not consider methods
reliant on interference machinery or relay selection [21], [26],
[27]. Instead, we utilize a lower-cost approach, taking into
account the similarity of line-of-sight channels in satellite-
to-ground and air-to-ground scenarios. The primary issue
addressed here is secure uplink transmission under extremely
harsh conditions, focusing on low energy consumption and
cost-effectiveness for physical network nodes. Additionally,
the problem of fairness among uplink nodes is also consid-
ered. Particularly, the potential of UAV-assisted secure uplink
communications in satellite-supported IoT networks is studied.
We propose a novel framework that leverages the inherent
advantages of UAV to realize the secure transmissions in
the uplink of from IoT users to LEO satellite. To combat
the interception of uplink signals from IoT users by ground-
based eavesdropper (Eve), we employ a UAV to securely relay
the uplink transmissions. A max-min optimization problem is
formulated to improve the secrecy rate performance among
IoT data uplinks, considering the energy constraints of IoT
users by incorporating uplink power as a limiting factor.
Then the max-min secrecy fairness is realized among uplink
transmissions through joint optimization of UAV placement,
uplink power allocation of IoT users, and UAV beamforming.
In addition, main contributions of this work are as follows:
•We establish a framework for secure uplink transmissions
in satellite-supported IoT networks using UAV-assisted
communication, assuming a secure UAV-to-satellite link.
To tackle challenges arising from potential ground Eve
intercepting uplink transmissions of IoT users, we formu-
late a max-min secrecy rate optimization problem, aiming
to improve the overall secrecy performance among simul-
taneous secure uplink transmissions while constraining
the uplink transmission power.
•To tackle the non-convex max-min uplink secrecy rate
problem, we introduce a streamlined two-stage optimiza-
tion approach. In the inner stage, we mathematically
consolidate uplink power allocation and UAV beam-
forming, and propose an SCA-based algorithm for their
joint optimization. In the outer stage, we propose the
synergized bisection and coordinate descent algorithm
to optimize UAV placement. Ultimately, convergence is
achieved through alternating iterations between these two
stages.
•The impact of uplink transmission power from IoT users
on secrecy rate is analyzed, and it is revealed that the
max-min problem attains its solution when the secrecy
fairness among uplink transmissions is realized. Also
we provide complexity analysis of proposed algorithms.
Moreover, effectiveness of the proposed approach is sub-
stantiated through extensive simulation results.
The remainder of this paper is organized as follows. In
Section II, we provide an overview of the satellite networks for
IoT applications and discuss the challenges and opportunities
associated with implementing secure uplink transmissions
in satellite-supported IoT networks. We also introduce the
concept of UAV-assisted secure uplink communications and
explain its potential benefits for IoT networks. Section III
presents system model and formulates the max-min problem
by comprehensive considering the uplink power allocation,
UAV beamforming, and UAV placement. In Section IV, we
propose the algorithm of joint optimization of UAV placement,
UAV beamforming, and uplink power allocation to solve the
max-min problem and provide some discussions. Section V
presents the simulation results and performance evaluation,
followed by the conclusion and future research directions in
Section VI.
II. RE LATE D WORK
In this section, we discuss the current research progress
in secure uplink communication for satellite-supported IoT
networks. Satellite networks have attracted growing interest
due to their potential for IoT communication, particularly in
remote and underserved areas where terrestrial communica-
tion infrastructures are insufficient [28]. Various aspects of
satellite-based IoT networks, including network architecture,
resource allocation, and protocol design, have been explored
by researchers [28], [29]. Among these aspects, the security of
data transmission has become a vital concern. Recent research
has emphasized cryptographic techniques, key management,
and secure routing to protect the confidentiality and integrity
of data transmitted over satellite networks [29], [30]. However,
physical layer security, especially in uplink communication,
has not been extensively investigated and remains a relatively
unexplored area.
The incorporation of UAVs into satellite-supported IoT
networks has shown promising results in enhancing the se-
curity of satellite-to-ground communication [31]. Stochastic
analysis has been utilized to examine cooperative satellite-
UAV communications, considering aerial relays to ensure a
secure satellite-UAV link [32]. In [33], a two-layer Stackel-
berg game model has been suggested to counter full-duplex
(FD) eavesdropping and jamming attacks, where malicious
eavesdropping attacks are resisted by an optimal cooperative
UAV transmits jamming signals. Muli-beam satellite is con-
sidered in [18] which addresses the challenge of enhancing
the legitimate user’s secrecy rate within a designated beam
while ensuring the common communication performance for
users in surrounding beams. Besides, a UAV is introduced to
leveraged to act as the relay to reinforce secure satellite beam
and to serve as a jammer that purposefully creates artificial
noise (AN) to thwart eavesdropping attempts. Considering
energy consumption limitations at UAV, analysis of ergodic
capacity and achievable secrecy rate have been given in [34]
for the downlink of satellite-terrestrial communications, with
the UAV employing maximum-ratio combining (MRC) to
receive satellite signals and enhance transmission capacity
while simultaneously combating eavesdropping.
Considering computation capability and secure transmis-
sion, a double-edge secure offloading approach has been
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
3
Internet
User Eve
User
Legitimate
uplink
Wiretap
uplink
Fig. 1. UAV-assisted secure uplink communications in satellite-supported IoT.
presented in [35] for space-air-aqua integrated networks. This
scheme involves UAVs securely relaying offloading for mar-
itime mobile users and deploying jamming UAVs to pro-
tect the offloading process by determining transmit power
and UAV trajectories. For power-limited or battery-free IoT
devices, a secure structure has been proposed in [36] to
support UAV-assisted IoT networks. This strategy includes
trajectory planning for UAVs to minimize energy consumption
across multiple clusters to maximize secrecy performance. In
scenarios where confidential messages are transmitted to a
mobile user by a UAV and AN is emitted by a cooperative
UAV to deter eavesdroppers, the challenge of maximizing
secrecy rates is tackled in [37]. A joint design for UAVs’ 3D
trajectories and time allocation is employed, taking into ac-
count practical constraints such as speed, collision avoidance,
positioning error, and energy harvesting. To maximize average
secrecy rates for both uplink and downlink in air-to-ground
transmissions, a joint optimization framework is proposed in
[38], incorporating the UAV trajectory and the transmission
power of legitimate user. Besides, to maximize the average
worst-case secrecy rate among UAV downlink transmissions,
a joint optimization of the UAV trajectory, beamforming
of intelligent reflective surface, and transmission power of
legitimate users, is proposed in [39]. By employing the UAV
as a relay between cluster users and terrestrial base stations,
the secrecy energy efficiency is maximized by jointly adjusting
the uplink transmission powers and the UAV’s position [40].
Notations: Tr (·)denotes the trace of a matrix. rank (·)
denotes the rank of a matrix. ∆
=means the redefinition. Ca×b
denotes a complex space of a×b.(·)†denotes the Hermitian
transpose. Nµ, δ2denotes the normal distribution with
mean µand variance δ2.k·k stands for the Euclidean norm of
a vector. Other notations can be found in Table I.
TABLE I
NOTATIO NS AN D DEFI NIT IO NS
Notation Definition
MNumber of IoT users within the UVA’s coverage area
MSet of IoT user index
IUmThe mth IoT user within the UAV coverage
g†
m,u ∈C1×NThe legitimate wiretap channel from IUmto UAV
g†
m,e ∈C1×NThe wiretap channel from IUmto UAV
pmThe transmission power of IUm
Dm→uThe distance from IUmto UAV
wm∈CN×1The beamforming vector at UAV receiver
RmThe secrecy rate of transmission from IUmto UAV
γm,u The received SINR of transmission from IUmto UAV
γe,m The received SINR at for wiretapping IUm
pLos The LoS probability
pNLos The non-LoS probability
(xu, yu, hu) The 3D coordinates of UAV
(xm, ym) The 2D horizontal coordinates of IoT user
gLoS ∈CN×1The LoS component of ground-to-UAV channel
gRay ∈CN×1The NLoS Rayleigh fading component
WmUAV beamforming matrix (Wm=w†
mwm)
III. SYS TE M MOD EL A ND PRO BL EM FO RM UL ATIO N
We investigate UAV-assisted secure uplink communications
in satellite-supported IoT networks, as illustrated in Fig. 1.
We focus on a remote area where multiple IoT devices are lo-
cated within the coverage of a satellite communication, which
provides backhaul connectivity to the Internet. In the uplink,
IoT devices transmit confidential information to the satellite,
which subsequently relays the data to Internet servers via
the backhaul. To enhance the secrecy performance of uplink
transmissions from IoT devices to the satellite, we propose
utilizing a UAV as an aerial relay to assist the implement
of physical layer security. Within the UAV’s coverage area,
we assume that MIoT devices are distributed following a 2D
Poisson Point Process (PPP) with intensity λband are denoted
by the set Φb[41]. These IoT devices are subject to random
activation, while an eavesdropper (Eve) within the same cover-
age area which aims to intercept the uplink transmissions. The
proposed UAV-assisted secure uplink communication system
seeks to protect the data transmitted by IoT devices from
being compromised by the eavesdropper. By leveraging the
inherent advantages of the UAV in terms of mobility and line-
of-sight (LoS) connections, the system aims to improve the
physical layer security of uplink communications, ensuring
data confidentiality and integrity for IoT devices in satellite-
supported networks.
A. Channel Models
In this paper, we follow the empirical uplink channel of
satellite [42], [43], and the channel power gain at satellite is
modeled as
H=βEI RP lairClζ, (1)
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
4
where the βEI RP is the uplink EIRP, lair denotes the air
absorption attenuation induced by the resonance of gas and
water vapor in satellite-to-ground links, and ldenotes the free-
space path gain which is calculated as
l(θ) = l0
d2=l0
r2+ (r+hso)2−2r(r+hso ) cos θ,(2)
where l0=c2(4πf )2denotes the path gain with cbeing
the speed of light and fbeing the center carrier frequency.
Besides, ζdenotes small-scale fading which experiences a
mixed Gaussian distribution, which is written as
ζ[dB] ∼pLosN−µLos , δ2
Los+pN LosN−µN Los , δ2
NLos ,
(3)
where pLos and pNLos are the LoS and non-LoS probabilities
respectively, and N−µLos, δ2
Losand N−µN Los, δ 2
NLos
are normal distributions with mean µLos,µN Los and variance
δ2
Los,δ2
NLos , respectively. Particularly, pLos +pNLos = 1 and
the probability of the LoS in (3) can be calculated as
pLos = exp (−`cot φ) = exp −`sin θ
cos θ−θ0,(4)
where `denotes the propagation environment parameter and
φis the elevation angle between satellite and user.
We focus primarily on the secure transmission from IoT
nodes to the UAV relay in the uplink. Additionally, we
assume that the security of the link between the UAV relay
and the satellite, which is responsible for forwarding secure
information, can be guaranteed. Particularly, the channel from
IoT user m(IUm) to the UAV can be modeled as [44]
gm=√g0
Dm→u rK
K+ 1gLoS +r1
K+ 1gRay!,(5)
where g0represents the channel power gain from the ground
source to the aerial destination at a reference distance of 1
meter, and Dm→udenotes the distance between IUmand UAV,
defining as
Dm→u=q(xu−xm)2+ (yu−ym)2+h2
u,(6)
where (xu, yu, hu) is the 3D coordinates of UAV and (xm, ym)
is the 2D horizontal coordinates of IoT user. Whereas, the
small-scale fading adopts Rician channel model, where Kis
the Rician factor (KB= 10log10(K)in dB), gLoS ∈CN×1
denotes LoS component, and gRay ∈CN×1represents the
NLoS Rayleigh fading component.
Besides, the channel between IoT users and ground Eve is
modeled as a Nakagami-mfading channel. This channel model
characterizes the signal propagation through various fading
environments and captures the fluctuations in signal strength
due to multipath propagation, shadowing, and other factors
[45]. The Nakagami-mfading channel model is versatile,
as it can represent different fading conditions by adjusting
the mparameter. A higher value of mindicates less severe
fading, whereas a lower value represents a more severe fading
environment.
B. Signal Models
In the uplink, we consider IoT nodes operating in the same
frequency band, and the signal received by the UAV from IoT
users can be represented as
yu=X
M
g†
m,uwm√pmsm+nm,(7)
where pmdenotes the uplink transmission power of IUm,
g†
m,u ∈C1×Ndenotes the channel sate information (CSI)
from IUmto UAV, wm∈CN×1is the beamforming vector
at UAV for shaping the signal from IUm,smcontains the
confident information expected to be delivered to the satellite
backhaul network, nmis the noise received by UAV.
Based on our considered eavesdropping model, due to the
openness of wireless channel and the ground Eve operating in
the same frequency band as the IoT nodes, the received signal
by the Eve can be represented as
ye=X
M
g†
m,ew0m√pmsm+ne,(8)
where g†
m,e ∈C1×Ndenotes the wiretap channel from IUm
to Eve, w0∈CN×1is the beamforming vector at Eve, and ne
denotes the noise received by UAV and Eve, respectively.
From (7–8), it can be seen that the uplink signal of the IoT
user has co-channel interference, and the signal received by
the Eve also experiences co-channel interference among users.
Based on this, we calculate the uplink SINRs of IUmat UAV
and Eve, respectively, which are obtained as
γm,u =pm
g†
m,uwm
2
P
i6=m,i∈M
pi
g†
i,uwi
2
+δ2
m
,(9)
γe,m =pm
G†
m,ew0m
2
P
i6=m,i∈M
pi
g†
i,ew0i
2
+δ2
e
,(10)
where the δ2
mand δ2
edenote the noise power received by IUm
and Eve.
C. Problem Formulation
Using (9) and (10), the secrecy rate of transmission from
IUmto UAV is obtained as
Rm= [log2(1 + γm,u)−log2(1 + γe,m)]+.(11)
To enhance the uplink secrecy rate from IoT users to
UAV and guarantee the secrecy fairness, we devise a problem
formulation aimed at maximizing the minimum secrecy rate
across uplink transmissions, which can be mathematically
expressed as
P1 : MaxMin
{xu,yu,pm,wm}{Rm}(12)
s.t.: X
m∈M
pm≤ Q,(12a)
0≤pm≤Pmax,(12b)
kwmk= 1,(12c)
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
5
where (12a) constrains the sum power of IoT users with
a predefined Q, (12b) represents the constrained-power of
IUm, and (12c) constrains the beamforming at UAV. The
formulated problem P1aims at uplink secure fairness by
jointly optimizing the uplink power allocation, the UAV’s
position and the beamforming at the UAV’s receiving end.
In addition, the formulated problem P1exhibits non-
convexity due to three key factors: (i) The secrecy rate (11),
calculating the difference between two logarithmic functions,
inherently creates non-convexity; (ii) The Max-Min objective,
targeting to maximize the minimum secrecy rate across uplink
transmissions, adds complexity to the problem due to the
need for multi-objective optimization; (iii) Power constraints
(12a and 12b) and the unit-norm beamforming constraint
(12c) further enhance non-convexity. Due to the problem’s
non-convexity, conventional convex optimization falls short.
Thus, advanced mathematical simplifications and optimization
techniques are imperative, warranting meticulous scrutiny of
potential local optima.
IV. JOI NT OPTIMIZATION OF UAV PL ACE ME NT, UAV
BEAMFORMING,AN D UPLINK POWE R ALL OC ATIO N
In this section, we propose an approach that jointly optimize
the UAV placement and the power allocation of IoT users to
improve the secrecy rate performance of IoT users for the
uplink transmissions. Since the formulated problem in (12) is
non-convex and entails situations where multiple optimization
variables are multiplied, in addition to the presence of numer-
ous quadratic optimization variables, we initially undertake
a transformation and simplification of the original problem.
Specifically, we recast the original problem as a two-stage
solution problem. In the first stage, we jointly optimize the
uplink power allocation for IoT nodes and UAV beamforming.
Subsequently, in the second stage, the placement of UAV is
also optimized. Finally, the optimization of these two stages
is iteratively alternated until performance converges.
For facilitating the derivation of formulas, the secrecy rate
in (11) is further represented as shown in (13), where the
following replacements are adopted ,
Gm,u =D2
m→ugm,ug†
m,u,(14)
Gm,e =D2
m→uGm,eG†
m,e.(15)
Wm=w†
mwm.(16)
Based on (13), the inter-user interference can have a sig-
nificant impact on the achievable secrecy rate for each user
in a multi-user network with an eavesdropper. This is because
the interference power increases as the transmission power of
each IoT user increases, and also as the number of users in the
network increases. In addition, changes in the position of the
UAV can also have an impact on the security rate performance.
A. Uplink Power Allocation and UAV Beamforming
We assume that the UAV updates its beamforming once after
its position changes, and the uplink power allocation of the
IoT users is optimized at the same time. Therefore, in order
to solve the original problem, in the first stage, we assume
that when the UAV placement is fixed at a certain position,
the original problem is simplified into a joint power allocation
and beamforming optimization problem. Particularly, we first
define an arbitrary variable ϕwith ϕ≤Rm, m ∈ M to
reformulate P1as
P2 : Max
{pm,Wm}ϕ(17)
s.t.: (12a),(12b),(17a)
ϕ≤Rm, m ∈ M,(17b)
Tr (Wm) = 1,(17c)
Wm0.(17d)
In P2, constraints (12a) and (12b) are retained from the
original problem. The constraint condition (17b) is satisfied to
simplify the max-min problem present in the single-objective
function. Moreover, combining constraints (17c) and (17d)
together is equivalent to constraint (12c). Based on (16),
rank (Wm)=1is held. To address non-convexity, P2
applies Semi-Definite Relaxation (SDR) to handle the rank-1
constraint of UAV beamforming matrix [46], easing problem-
solving.
Particularly, based on P2, we can draw some interesting
findings as follows.
Theorem 1. Considering spectrum sharing among IoT nodes,
when the maximum transmission power for the uplink is
specified, given a specific receiver position and receiver beam
direction, the uplink secrecy rate of IoT user monotonously
increases as its power allocation.
Proof. we take derivative of the secrecy rate in (13) is calcu-
lated as
∂Rm
∂pm
= log2eαAm
pmAm+α−βBm
pmBm+β,(18)
Rm= log2
1 + pm
g†
m,uwm
2
P
i6=m,i∈M
pi
g†
i,uwi
2
+δ2
m
−log2
1 + pm
G†
m,ew0m
2
P
i6=m,i∈M
pi
g†
e,iw0i
2
+δ2
e
= log2
1 + pmTr (Gm,uWm)
P
i6=m,i∈M
piTr (Gi,uWi) + D2
m→u
−log2
1 + pmTr (Gm,eW0m)
P
i6=m,i∈M
piTr (Ge,iW0i) + D2
m→u
.(13)
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
6
where Am,Bm,α, and βare defined as follows
Am= Tr (Gm,uWm),(19)
Bm= Tr (Gm,eW0m),(20)
α=X
i6=m,i∈M
piTr (Gi,uW) + D2
m→u,(21)
β=X
i6=m,i∈M
piTr (Ge,iW0i) + D2
m→u.(22)
We note that Am≥Bmand α≥βdue to the fact that only
positive secrecy rates are feasible in our formulated problem.
Therefore, we have
αpmAmBm+αβAm≥βpmAmBm+αβBm,(23)
which implies that ∂Rm
∂pm≥0. This indicates the secrecy rate
of IUmmonotonically increases as its power allocation.
This concludes the proof of the theorem.
With the constraint in (12a), the power allocation of IoT
user restricts that of other users. Thus we have the following
Theorem to analyze the secrecy fairness among the IoT users
for secure uplink transmissions.
Theorem 2. Secrecy fairness: For feasible power allocations
of IoT users, the secrecy rates of IoT users achieve the same
values, Rm=Ri(m6=i, m ∈ M, i ∈ M).
Proof. Consider the optimization problem in P2, where the
secrecy goal is reformulated to maximize ϕsubject to con-
straints (14a) through (14d).
Observe that constraint (14b) requires ϕ≤Rm,∀m∈ M.
In the optimal solution, the goal is to maximize ϕ, which en-
tails enlarging ϕto the greatest possible extent while adhering
to this constraint. Consequently, in the optimal solution, at
least one user m∈ M must satisfy ϕ=Rm.
Assume there exists at least one user i∈ M such that Rm<
Ri. Under this condition, ϕcan be augmented by reallocating
power from user ito user m. This reallocation can be executed
without infringing constraint (14a) and will result in a larger
ϕ, as Rmincreases and Ridecreases. This procedure can be
iterated until Rm=Ri∀m6=i, m ∈ M, i ∈ M.
Thus, the secrecy rates of IoT users achieve the same
values, Rm=Ri(m6=i, m ∈ M, i ∈ M), for feasible power
allocations of IoT users.
This concludes the proof of the theorem.
Based on the aforementioned theorem, an iterative binary
search method can be employed to determine the final power
allocation. However, the objective is to jointly optimize the
uplink power allocation of IoT users and UAV beamforming.
Therefore, we further simplify problem P2and utilize a
convex approximation algorithm for its resolution. To elab-
orate, the original problem P2aims to maximize the worst-
case secrecy rate ϕamong IoT uplink transmissions while
optimizing both the uplink power allocation of IoT users
and the UAV beamforming. By employing the proof in the
theorem, we have established that in the optimal solution, the
secrecy rates of all IoT users achieve the same values, i.e.,
Rm=Rifor all m6=i, m ∈ M, i ∈ M, given feasible power
allocations. Specifically, with loss of generality, we fist assume
a group of initialized power allocations, p0
m, m ∈ M. Then,
we simplify problem P2by reformulating it into a more
tractable problem, which focuses on the joint optimization of
uplink power allocation of IoT users and UAV beamforming.
The constraint (17b) can be reformulated as
ϕ≤log2e(u−mu −t+mt),(24)
where the new introduced variables u, mu, t, and mt satisfy
the following definitions
eu=X
i∈M
piTr (Gi,uWm) + D2
m→u,(25)
et=X
i∈M
piTr (Ge,iW0i) + D2
m→e,(26)
emu =X
i6=m,i∈M
piTr (Gi,uWm) + D2
m→u,(27)
emt =X
i6=m,i∈M
piTr (Ge,iW0i) + D2
m→e.(28)
Besides, by defining ωm=pmWm, the joint uplink power
allocation and UAV beamforming problem is reformulated as
P3 : Max
ωm
ϕ(29)
s.t.: X
m∈M
Tr (ωm)≤ Q,(29a)
ϕ≤log2e(u−mu −t+mt), m ∈ M,(29b)
eu≤X
i∈M
piTr (Gi,uWi) + D2
m→u,(29c)
e˜
t˜
t−t+ 1≥X
i∈M
piTr (Ge,iW0i) + D2
m→e,(29d)
e˜mu ( ˜mu −mu + 1) ≥X
i6=m,i∈M
piTr (Gi,uWi) + D2
m→u,
(29e)
emt ≤X
i6=m,i∈M
piTr (Ge,iW0i) + D2
m→e,(29f)
ωm0.(29g)
To address P3, we propose a successive convex approxi-
mation (SCA) based joint optimization algorithm for uplink
power allocation and UAV beamforming, as illustrated in
Algorithm 1.
Complexity analysis: The complexity of the algorithm
mainly comes from two aspects: the semi-definite program-
ming (SDP) problem solved using the CVX tool and the
convergence criterion.
SDP: The SDP problem in Step 2 of the algorithm is solved
using the CVX tool. SDP problems have a complexity that is
polynomial in the number of variables and constraints. Let n
denote the number of variables and mdenote the number of
constraints in the SDP problem. The complexity of solving an
SDP problem is generally in the order of O(n2m+n3). In
this specific problem, the number of variables and constraints
will depend on the dimensions of the matrices involved and
the size of the optimization problem.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
7
Algorithm 1: SCA-based Joint Optimization Algo-
rithm for the Uplink Power Allocation of IoT users
and the UAV Beamforming
Require: Initial values: ˜
tand ˜mu. Input channel state
information (CSI) of uplink channels, estimated CSI of
Eve, and location information of the UAV.
Ensure: Optimized uplink power allocation pmand UAV
beamforming matrix Wm.
1: Initialize ˜
tand ˜mu;
2: repeat
3: Execute semi-definite programming (SDP) using the
CVX tool.
4: Update ˜
tand ˜mu;
5: Compute the objective function ϕusing (29);
6: until convergence criterion is met
7: Compute pm=|ωm|2,Wm=ωm/pm
8: return pmand Wm;
Convergence criterion: The algorithm iterates until a con-
vergence criterion is met. Let Niter denote the maximum
number of iterations required for the algorithm to converge.
The complexity of this part depends on how quickly the
algorithm converges to the optimal solution, which can be
influenced by factors such as the initial values of ˜
tand ˜mu,
the channel state information (CSI), and the location of the
UAV.
Considering both aspects, the overall complexity of the
alternating optimization algorithm is O(Niter(n2m+n3)).
Since the algorithm employs the CVX tool to solve the SDP
problem.
B. Optimization of UAV Placement
Based on the uplink power allocation parameters and UAV
beamforming vectors obtained from Algorithm 1 in the first-
stage, the UAV placement optimization problem inherited from
the original max-min problem can be reformulated as:
P4 : Max
{xu,yu}ϕ(30)
s.t.: ϕ≤Rm,(30a)
where the constraint in (30a) can be reformulated as
log2α0+Am+D2
m→u
α0+D2
m→u−log2β0+Bm+D2
m→u
β0+D2
m→u≥ϕ,
(31)
with
α0=P
i6=m,i∈M
piTr (Gi,uWi),
β0=P
i6=m,i∈M
piTr (Ge,iWi).
(32)
After simplification and analysis, (31) is further reformu-
lated as shown in (33), at the bottom of this page.
We take a replacement of Lm,u =D2
m→uand and refor-
mulate P3as
P5 : Max
{Lm,u}ϕ(34)
s.t.: aL2
m,u +bLm,u +c≤0,(34a)
Lm,u ≥0,(34b)
b2−4ac > 0,(34c)
where a, b, and care constants. Based on (33), since 2ϕ−1≥
0, the parabola with variable Lm,u should be at least two points
of intersection with the horizontal axis of Lm,u and thus (34b)
should be satisfied.
Remark 1. By carefully analyzing the simplifed Problem P5,
we can draw an interesting finding, which are discussed as
follows. Based on (34a), we have
Lm,u ≤−b+√b2−ac
2a.(35)
Particularly, by taking the derivation of Lm,u, we have
∂Rm
∂Lm,u
= log2eBm
(β0+Bm+Lm,u) (β0+Lm,u )
−Am
(α0+Am+Lm,u) (α0+Lm,u )
≤0,(36)
which indicates that the secrecy rate of IUmdecreases
monotonously as the distance between the UAV and IUm.
To address problem P5in the second stage, we introduce
an iterative optimization strategy that synergizes the bisection
and coordinate descent methods for optimizing the UAV’s
positioning, as outlined in Algorithm 1.
Algorithm 1 commences with defining the search parame-
ters for ϕwith a lower (ϕmin) and upper (ϕmax) boundary.
The algorithm repeats until it converges, assessed via a pre-
established tolerance .
Each iteration incorporates two crucial steps:
Step 1: The bisection approach is utilized to revise ϕ. At
each iteration, the current search range’s midpoint is calculated
as:
ϕ=ϕmin +ϕmax
2.(37)
Subsequently, coefficients a,b, and care updated based on
the new ϕvalue. The problem’s feasibility is then validated
through the discriminant condition:
∆ = b2−4ac. (38)
(2ϕ−1) D4
m→u+D2
m→u[2ϕ(α0+β0+Bm)−α0−β0−Am]+2ϕ(α0β0+α0Bm)≤α0β0+Amβ0(33)
⇔aL2
m,u +bLm,u +c≤0; a∆
= 2ϕ−1, b ∆
= 2ϕ(α0+β0+Bm)−α0−β0−Am, c ∆
= 2ϕ(α0β0+α0Bm)−α0β0+Amβ0.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
8
Algorithm 2: Synergized Bisection and Coordinate
Descent algorithm
1Initialization: ϕmin = 0,ϕmax represents the upper
limit, and > 0is the tolerance;
2while non-convergence do
3Step 1: ϕis updated via the bisection method:
4ϕ←ϕmin+ϕmax
2;a,b, and ccoefficients are
revised based on the new ϕ;
5if b2−4ac > 0then
6Feasible problem: Adjust ϕmin or ϕmax;
7else
8Infeasible problem: Modify ϕmin or ϕmax;
9Step 2: Establish the optimal Lm,u by minimizing
constraint (25a) through coordinate descent;
Refresh Lm,u and evaluate convergence based on
the tolerance ;
10 Output: Optimal values of ϕand Lm,u with (xu, yu).
If ∆>0, the problem is deemed feasible; otherwise, it’s
considered infeasible. Dependent on the feasibility, the search
range for ϕis updated accordingly.
Step 2: With the revised ϕ, the optimal Lm,u is ascertained
by minimizing the constraint in (25a) via coordinate descent.
The update rule for Lm,u is enforced, and the algorithm’s
convergence is evaluated based on the tolerance .
Finally, the convergence can be attained by alternating
iterations between these two stages optimizations in Algorithm
1 and Algorithm 2.
Complexity analysis: The proposed algorithm provides an
efficient and effective way to find the optimal values for ϕ
and Lm,u, which can be utilized in the optimization of UAV
placement in the given communication system. The complexity
of our proposed algorithm mainly includes two aspects: the
bisection method used for updating ϕand the optimization
technique employed for finding the optimal value of Lm,u.
Bisection method: The bisection method is known for its
logarithmic convergence rate. At each iteration, the search
interval is reduced by half. Let Nϕdenote the maximum
iteration number for the bisection method to converge. Given a
predefined tolerance , the complexity of the bisection method
is O(log2(ϕmax−ϕmin
)).
Optimize Lm,u:The complexity of this step depends on the
optimization technique used for minimizing the constraint in
(25a). Coordinate descent is one possible method for solving
this problem. Let NLm,u denote the maximum number of
iterations for the optimization of Lm,u to converge. The
complexity of this step is O(NLm,u ), where the actual value of
NLm,u depends on the specific convergence rate of the chosen
optimization technique.
Considering both aspects, the overall complexity of the
proposed algorithm is O(NϕNLm,u ).
Based on the above subsection A and B, the two-stage
optimization approach is completed. In the first stage, we
solve optimization problem P3using Algorithm 1, obtaining
the uplink power allocation and the UAV’s beamforming
vector. Based on these results, in the second stage, we solve
optimization problem P5using Algorithm 2 to determine the
position of the UAV.
The problem defined in this paper is non-convex and multi-
variable, thus obtaining a global optimum solution is challeng-
ing. The problem is simplified to a bi-convex form and solved
using a two-tier approach. Utilizing an alternating iterative
framework, fast convergence is achieved. Although each tier
achieves optimal solutions, the overall outcome is considered
near-optimal.
V. PERFORMANCE EVAL UATIONS
In this section, various simulations are conducted to assess
the secrecy performance of uplink transmissions for IoT users.
The simulation parameters are configured as follows: IoT users
are distributed according to a 2D PPP within an area with
a radius of 100 meters. The initial position of the UAV is
set in 3D coordinates as (0m, 0m, 100m). The imperfect
channel state information (CSI) has an estimation error of
kυk= 0.2, such that ˆg†
m,e =g†
m,e +υ. The channel power
gains from IoT users to the UAV in the uplink and to Eve at a
reference distance of 1m are -40dB and -38.6dB, respectively.
The Ricean factor for the channel from IoT users to the UAV
is set at 10dB. The Nakagami-mparameters for the wiretap
channel from IoT users to the ground-based Eve are set to
(m=2, Ω=1). Additionally, specific parameters are introduced
for each individual simulation, e.g., the maximum transmission
power PS, the number of UAV receive antennas N, the number
of IoT users within coverage of UAV is set to M, the height of
UAV hu. The number of receive antennas of Eve is same as N.
For the benchmarks, we adopt two variants of the proposed,
which are respectively labeled as ”Fixed PA and ZF-BF &
placement opt.”, representing the approach that optimizes
only the UAV’s position with a fixed power allocation using
equal division and the UAV beamforming using Zero-Forcing
beamforming associated to the eavesdropping channel, and
”Fixed location &joint PA and BF opt.”, representing the
approach that fixes the UAV’s location while jointly optimizing
the uplink power allocation and the UAV’s beamforming.
Besides, we draw from established optimization strategies
in relevant research, formulating an ”Alternating SDP and
Dinkelbach Optimization”. Semidefinite Programming (SDP)
is frequently employed for power allocation and beamforming,
whilst Dinkelbach’s method is regularly utilized for planning
drone trajectories [47], [48]. As a benchmark, we amalgamate
these techniques to address the secure fairness problem.
Fig. 2 shows the influence of the maximum transmission
power on the minimum secrecy rate. As observed, the secrecy
rate performance increases with the maximum transmission
power, which is consistent with Theorem 1. According to
Theorem 1, allocating more power to IoT users as the maxi-
mum transmission power increases results in improved secrecy
rate performance. Besides, the proposed joint optimization
approach involving UAV placement, IoT user power allocation,
and UAV beamforming demonstrates superior performance
compared to the evaluated benchmarks. Comparing the curves
in Fig. 2, the following observations can be made: (1) The
presence of channel estimation error can degrade the max-
min secrecy rate performance; (2) Optimizing only the UAV
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
9
5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Proposed, perfect CSI
Proposed, imperfect CSI
Alternating SDP and Dinkelbach opt.
Fixed PA and ZF-BF & placement opt.
Fixed location & joint PA and BF opt.
Fig. 2. Maximum transmission power PsVs. the minimum secrecy rate.
(N= 4, M = 4, hu= 100m)
4 5 6 7 8 9 10 11 12
0
1
2
3
4
5
6
Proposed, perfect CSI
Proposed, imperfect CSI
Alternating SDP and Dinkelbach opt.
Fixed PA and ZF-BF & placement opt.
Fixed location & joint PA and BF opt.
Fig. 3. Number of UAV receive antennas NVs. the minimum secrecy rate.
(M= 4, Ps=10dB, hu= 100m)
placement, while using fixed power allocation and Zero-
Forcing (ZF)-based UAV beamforming, could lead to leakage
of legitimate signals due to ZF-based beamforming; (3) As
the maximum transmission power increases, the optimal UAV
placement contributes more significantly to security enhance-
ment than the optimal power allocation and beamforming
strategies.
Fig. 3 shows the effect of UAV receive antenna number
on the minimum secrecy rate. As Nincreases, the secrecy
rate performance exhibits an improvement. This enhancement
can be attributed to the increased capacity of the UAV to
effectively shape the uplink signal beams from IoT users with
a higher number of antennas, enabling better discrimination
between inter-user interference and minimizing energy leakage
of valuable signals. The proposed approach presented in this
paper demonstrates a notable superiority over the examined
benchmarks. A detailed comparisons of the curves in Fig. 3
reveals that the presence of channel estimation error results
2345678910
0
0.5
1
1.5
2
2.5
3
3.5
Fig. 4. Number IoT users MVs. the minimum secrecy rate. (N= 4, Ps=
10dB, hu= 100m)
in a decline in secrecy rate performance. Furthermore, when
an adequate number of antennas are utilized, optimizing the
uplink power allocation and UAV beamforming contributes
more significantly to the improvement of the secrecy rate
compared to the optimization of the UAV deployment location.
Fig. 4 shows the impact of the increasing number of uplink
IoT users on the minimum secrecy rate performance, which
demonstrates a decline as the user count grows. This outcome
can be ascribed to the escalation of inter-user interference
stemming from the expanding number of IoT users, leading
to a consequent reduction in the SINR at the UAV receiver.
Furthermore, the proposed approach consistently outperforms
the other benchmark approaches, and channel errors contribute
to a diminished secrecy rate performance. In large-scale IoT
applications with a substantial user population, the joint op-
timization of UAV placement, uplink power allocation, and
UAV beamforming can more proficiently safeguard the overall
secrecy rate performance for IoT users.
Fig. 5 shows the impact of UAV deployment height on the
minimum secrecy rate, highlighting a decrease in the secrecy
rate performance as the UAV deployment height. This decline
can be attributed to the exacerbation of air-to-ground channel
fading due to the heightened UAV position, which undermines
the quality of legitimate signal reception. Consequently, the
main channel capacity of the uplink from IoT users is reduced
while the eavesdropping channel remains unaltered, leading
to diminished secrecy rate performance. Once again, Fig. 5
demonstrates that the proposed joint optimization of UAV
placement, uplink power allocation, and UAV beamforming
outperforms other benchmark approaches. Channel estimation
errors can also contribute to the degradation of the max-
min secrecy rate performance. Furthermore, the alteration
in UAV height does not impact the significance of UAV
deployment optimization over uplink power allocation and
UAV beamforming optimization in terms of enhancing secrecy
rate performance.
Fig. 6 shows the convergence performance of the joint
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
10
40 50 60 70 80 90 100 110 120 130 140
0
0.5
1
1.5
2
2.5
3
3.5
4
Fig. 5. UAV deployment height huVs. the minimum secrecy rate. (N=
4, M = 4, Ps=10dB)
1 2 3 4 5 6 7 8 9 10
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.55
Fig. 6. Convergence analysis. (N= 4, M = 4, PS= 5dB, hu= 100m)
optimization algorithm for UAV placement, uplink power allo-
cation, and UAV beamforming. The convergence performance
of the algorithm is evaluated by simulating the number of
iterations of the alternating optimization and the corresponding
minimum secrecy rate performance. Comparing the curves in
the figure, it can be seen that our proposed method, despite
jointly optimizing multiple dimensions, does not significantly
increase the number of iterations, and the alternating optimiza-
tion using Algorithm 1 and Algorithm 2 can quickly converge.
VI. CONCLUSION
This paper has explored the use of UAV for enhancing the
uplink secrecy rate performance in satellite-supported IoT, and
the secrecy fairness among IoT users is realized. A framework
is proposed that optimizes UAV placement, uplink power
allocation, and UAV beamforming to ensure secure uplink
transmissions while considering the energy constraints of IoT
users. The non-convex max-min uplink secrecy rate problem
is addressed using a two-stage optimization approach, which
includes SCA-based algorithm for jointly optimizing uplink
power allocation of IoT users and UAV beamforming, and
synergized bisection and coordinate descent algorithm for op-
timizing UAV placement. In addition, numerical results verify
the effectiveness of our proposed approach. Future research
directions include investigating the impact of mobility on the
proposed framework and optimizing the trade-off between
security and energy efficiency.
REFERENCES
[1] X. Shen and H. St¨
uttgen, “Growing our technical portfolio through
cooperation with other ieee entities,” IEEE Commun. Mag., vol. 61,
no. 3, pp. 4–6, 2023.
[2] T. Liu, J. Xia, Z. Ling, X. Fu, S. Yu, and M. Chen, “Efficient
federated learning for aiot applications using knowledge distillation,”
IEEE Internet Things J., vol. 10, no. 8, pp. 7229–7243, 2023.
[3] N. Cheng, J. He, Z. Yin, C. Zhou, H. Wu, F. Lyu, H. Zhou, and X. Shen,
“6g service-oriented space-air-ground integrated network: A survey,”
Chin. J. Aeronaut., vol. 35, no. 9, pp. 1–18, 2022.
[4] Z. Yin, N. Cheng, Y. Hui, W. Wang, L. Zhao, K. Aldubaikhy, and
A. Alqasir, “Multi-domain resource multiplexing based secure transmis-
sion for satellite-assisted IoT: AO-SCA approach,” IEEE Trans. Wireless
Commun., 2023, early access.
[5] T. K. Rodrigues and N. Kato, “Hybrid centralized and distributed
learning for mec-equipped satellite 6g networks,” IEEE J. Sel. Areas
Commun., 2023.
[6] D. Han, Q. Ye, H. Peng, W. Wu, H. Wu, W. Liao, and X. Shen, “Two-
timescale learning-based task offloading for remote iot in integrated
satellite-terrestrial networks,” IEEE Internet Things J., 2023, early
access.
[7] B. Al Homssi, A. Al-Hourani, K. Wang, P. Conder, S. Kandeepan,
J. Choi, B. Allen, and B. Moores, “Next generation mega satellite net-
works for access equality: Opportunities, challenges, and performance,”
IEEE Commun. Mag., vol. 60, no. 4, pp. 18–24, 2022.
[8] F. Tang, X. Chen, M. Zhao, and N. Kato, “The roadmap of communica-
tion and networking in 6g for the metaverse,” IEEE Wireless Commun.,
pp. 1–15, 2022.
[9] M. L ´
opez, S. B. Damsgaard, I. Rodr´
ıguez, and P. Mogensen, “An
empirical analysis of multi-connectivity between 5g terrestrial and leo
satellite networks,” in 2022 IEEE Globecom Workshops (GC Wkshps).
IEEE, 2022, pp. 1115–1120.
[10] I. Ashraf, Y. Park, S. Hur, S. W. Kim, R. Alroobaea, Y. B. Zikria, and
S. Nosheen, “A survey on cyber security threats in iot-enabled maritime
industry,” IEEE Trans. Intell. Transp. Syst., 2022.
[11] Y. Liu, L. Ni, and M. Peng, “A secure and efficient authentication
protocol for satellite-terrestrial networks,” IEEE Internet Things J.,
vol. 10, no. 7, pp. 5810–5822, 2023.
[12] B. Wang, J. Jiao, S. Wu, R. Lu, and Q. Zhang, “Age-critical and secure
blockchain sharding scheme for satellite-based internet of things,” IEEE
Trans. Wireless Commun., vol. 21, no. 11, pp. 9432–9446, 2022.
[13] B. Kuang, A. Fu, W. Susilo, S. Yu, and Y. Gao, “A survey of remote at-
testation in internet of things: Attacks, countermeasures, and prospects,”
Comput Secur, vol. 112, p. 102498, 2022. [Online]. Available:
https://www.sciencedirect.com/science/article/pii/S0167404821003229
[14] K. Yu, J. Yu, and C. Luo, “The impact of mobility on physical layer
security of 5g iot networks,” IEEE/ACM Trans. Netw., pp. 1–14, 2022.
[15] D. Chen, H. Wang, N. Zhang, X. Nie, H.-N. Dai, K. Zhang, and K.-K. R.
Choo, “Privacy-preserving encrypted traffic inspection with symmetric
cryptographic techniques in iot,” IEEE Internet Things J., vol. 9, no. 18,
pp. 17 265–17 279, 2022.
[16] M. Shen, K. Ye, X. Liu, L. Zhu, J. Kang, S. Yu, Q. Li, and K. Xu,
“Machine learning-powered encrypted network traffic analysis: A com-
prehensive survey,” IEEE Communi. Surve. & Tutori., vol. 25, no. 1, pp.
791–824, 2023.
[17] Y. Liu, Z. Su, C. Zhang, and H.-H. Chen, “Minimization of secrecy
outage probability in reconfigurable intelligent surface-assisted mimome
system,” IEEE Trans. Wireless Commun., vol. 22, no. 2, pp. 1374–1387,
2023.
[18] Z. Yin, N. Cheng, T. H. Luan, and P. Wang, “Physical layer security in
cybertwin-enabled integrated satellite-terrestrial vehicle networks,” IEEE
Trans. Veh. Technol., vol. 71, no. 5, pp. 4561–4572, 2022.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
11
[19] P. Angueira, I. Val, J. Montalban, ´
O. Seijo, E. Iradier, P. S. Fontaneda,
L. Fanari, and A. Arriola, “A survey of physical layer techniques for
secure wireless communications in industry,” IEEE Communi. Surv. &
Tutori., vol. 24, no. 2, pp. 810–838, 2022.
[20] S. Han, J. Li, W. Meng, M. Guizani, and S. Sun, “Challenges of physical
layer security in a satellite-terrestrial network,” IEEE Netw., vol. 36,
no. 3, pp. 98–104, 2022.
[21] K. Cao, B. Wang, H. Ding, L. Lv, R. Dong, T. Cheng, and F. Gong,
“Improving physical layer security of uplink noma via energy harvesting
jammers,” IEEE Trans. Inf. Forensics Security, vol. 16, pp. 786–799,
2021.
[22] V. Bankey, S. Sharma, S. R, and A. Madhukumar, “Physical layer
security of haps-based space-air-ground integrated network with hybrid
fso/rf communication,” IEEE Trans. Aerosp. Electron. Syst., pp. 1–8,
2022.
[23] S. Aggarwal, N. Kumar, and S. Tanwar, “Blockchain-envisioned uav
communication using 6g networks: Open issues, use cases, and future
directions,” IEEE Internet Things J., vol. 8, no. 7, pp. 5416–5441, 2021.
[24] W. U. Khan, E. Lagunas, Z. Ali, M. A. Javed, M. Ahmed, S. Chatzinotas,
B. Ottersten, and P. Popovski, “Opportunities for physical layer security
in uav communication enhanced with intelligent reflective surfaces,”
IEEE Wireless Commun., vol. 29, no. 6, pp. 22–28, 2022.
[25] Z. Yin, M. Jia, N. Cheng, W. Wang, F. Lyu, Q. Guo, and X. Shen, “Uav-
assisted physical layer security in multi-beam satellite-enabled vehicle
communications,” IEEE Trans. Intell. Transp. Syst., vol. 23, no. 3, pp.
2739–2751, 2022.
[26] X. Wang, Z. Fei, J. A. Zhang, and J. Huang, “Sensing-assisted secure
uplink communications with full-duplex base station,” IEEE Commun.
Lett., vol. 26, no. 2, pp. 249–253, 2022.
[27] K. Cao, B. Wang, H. Ding, L. Lv, J. Tian, and F. Gong, “On the security
enhancement of uplink noma systems with jammer selection,” IEEE
Trans. Commun., vol. 68, no. 9, pp. 5747–5763, 2020.
[28] M. Centenaro, C. E. Costa, F. Granelli, C. Sacchi, and L. Vangelista, “A
survey on technologies, standards and open challenges in satellite iot,”
IEEE Communi. Surve. & Tutori., vol. 23, no. 3, pp. 1693–1720, 2021.
[29] X. Zhu and C. Jiang, “Integrated satellite-terrestrial networks toward
6g: Architectures, applications, and challenges,” IEEE Internet Things
J., vol. 9, no. 1, pp. 437–461, 2021.
[30] K.-Y. Lam, S. Mitra, F. Gondesen, and X. Yi, “Ant-centric iot security
reference architecture—security-by-design for satellite-enabled smart
cities,” IEEE Internet Things J., vol. 9, no. 8, pp. 5895–5908, 2021.
[31] H. Guo, J. Li, J. Liu, N. Tian, and N. Kato, “A survey on space-air-
ground-sea integrated network security in 6g,” IEEE Communi. Surv. &
Tutori., vol. 24, no. 1, pp. 53–87, 2022.
[32] Y. Tian, G. Pan, M. A. Kishk, and M.-S. Alouini, “Stochastic analysis
of cooperative satellite-uav communications,” IEEE Trans. Wireless
Commun., vol. 21, no. 6, pp. 3570–3586, 2022.
[33] C. Liao, K. Xu, H. Zhu, X. Xia, Q. Su, and N. Sha, “Secure transmission
in satellite-uav integrated system against eavesdropping and jamming:
A two-level stackelberg game model,” China Commun., vol. 19, no. 7,
pp. 53–66, 2022.
[34] D. He, Z. Sha, H. Liu, T. Mao, and Z. Wang, “Uav-assisted satellite-
terrestrial secure communication using large-scale antenna array with
one-bit adcs/dacs,” IEEE Trans. Commun., vol. 71, no. 1, pp. 580–594,
2023.
[35] D. Wang, T. He, Y. Lou, L. Pang, Y. He, and H.-H. Chen, “Double-edge
computation offloading for secure integrated space-air-aqua networks,”
IEEE Internet Things J., pp. 1–1, 2023.
[36] R. Han, L. Bai, J. Liu, J. Choi, and Y.-C. Liang, “A secure structure
for uav-aided iot networks: Space-time key,” IEEE Wireless Commun.,
vol. 28, no. 5, pp. 96–101, 2021.
[37] W. Wang, X. Li, R. Wang, K. Cumanan, W. Feng, Z. Ding, and O. A.
Dobre, “Robust 3d-trajectory and time switching optimization for dual-
uav-enabled secure communications,” IEEE J. Sel. Areas Commun.,
vol. 39, no. 11, pp. 3334–3347, 2021.
[38] G. Zhang, Q. Wu, M. Cui, and R. Zhang, “Securing uav communications
via joint trajectory and power control,” IEEE Trans. Wireless Commun.,
vol. 18, no. 2, pp. 1376–1389, 2019.
[39] S. Li, B. Duo, M. D. Renzo, M. Tao, and X. Yuan, “Robust secure
uav communications with the aid of reconfigurable intelligent surfaces,”
IEEE Trans. Wireless Commun., vol. 20, no. 10, pp. 6402–6417, 2021.
[40] A. S. Abdalla and V. Marojevic, “Securing mobile multiuser transmis-
sions with uavs in the presence of multiple eavesdroppers,” IEEE Trans.
Veh. Technol., vol. 70, no. 10, pp. 11 011–11 016, 2021.
[41] T.-X. Zheng, X. Chen, C. Wang, K.-K. Wong, and J. Yuan, “Physical
layer security in large-scale random multiple access wireless sensor
networks: A stochastic geometry approach,” IEEE Trans. Commun.,
vol. 70, no. 6, pp. 4038–4051, 2022.
[42] B. A. Homssi and A. Al-Hourani, “Modeling uplink coverage perfor-
mance in hybrid satellite-terrestrial networks,” IEEE Commun. Lett.,
vol. 25, no. 10, pp. 3239–3243, 2021.
[43] P.-. Series, “Propagation model for IF77,” ITU-R Report, pp. 1–72,
2020. [Online]. Available: http://www.itu.int/publ/R-REP/en
[44] B. Hu, L. Wang, S. Chen, J. Cui, and L. Chen, “An uplink throughput op-
timization scheme for uav-enabled urban emergency communications,”
IEEE Internet Things J., vol. 9, no. 6, pp. 4291–4302, 2022.
[45] T.-H. Vu, T.-V. Nguyen, T.-T. Nguyen, D. B. da Costa, and S. Kim,
“Performance analysis of short-packet noma-based cdrt networks over
nakagami-mfading channels,” IEEE Trans. Veh. Technol., vol. 71,
no. 12, pp. 12 928–12 942, 2022.
[46] S. Zargari, A. Khalili, Q. Wu, M. R. Mili, and D. W. K. Ng, “Max-
min fair energy-efficient beamforming design for intelligent reflecting
surface-aided swipt systems with non-linear energy harvesting model,”
IEEE Trans. Veh. Technol., vol. 70, no. 6, pp. 5848–5864, 2021.
[47] J. Chu, R. Liu, M. Li, Y. Liu, and Q. Liu, “Joint secure transmit beam-
forming designs for integrated sensing and communication systems,”
IEEE Trans. Veh. Technol., vol. 72, no. 4, pp. 4778–4791, 2023.
[48] X. Zhou, Q. Wu, S. Yan, F. Shu, and J. Li, “Uav-enabled secure
communications: Joint trajectory and transmit power optimization,”
IEEE Trans. Veh. Technol., vol. 68, no. 4, pp. 4069–4073, 2019.
Zhisheng Yin (Member, IEEE) received his Ph.D.
degree from the School of Electronics and Infor-
mation Engineering, Harbin Institute of Technology,
Harbin, China, in 2020, and the B.E. degree from the
Wuhan Institute of Technology, the B.B.A. degree
from the Zhongnan University of Economics and
Law, Wuhan, China, in 2012, and the M.Sc. degree
from the Civil Aviation University of China, Tianjin,
China, in 2016. From Sept. 2018 to Sept. 2019,
Dr. Yin visited in BBCR Group, Department of
Electrical and Computer Engineering, University of
Waterloo, Canada. He is currently an Associate Professor in Xidian University,
Xi’an, China. He is also an Associate Editor of IEEE Internet of Things
Journal. His research interests include space-air-ground integrated networks,
wireless communications, cybertwin, and physical layer security.
Nan Cheng (Senior Member, IEEE) received the
Ph.D. degree from the Department of Electrical and
Computer Engineering, University of Waterloo in
2016, and B.E. degree and the M.S. degree from
the Department of Electronics and Information En-
gineering, Tongji University, Shanghai, China, in
2009 and 2012, respectively. He worked as a Post-
doctoral fellow with the Department of Electrical
and Computer Engineering, University of Toronto,
from 2017 to 2019. He is currently a professor
with State Key Lab. of ISN and with School of
Telecommunication Engineering, Xidian University, Shaanxi, China. His
current research focuses on B5G/6G, space-air-ground integrated network,
big data in vehicular networks, and self-driving system. His research interests
also include performance analysis, MAC, opportunistic communication, and
application of AI for vehicular networks.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
12
Yunchao Song (Member, IEEE) received the B.E.
degree in electronic science and technology and
the Ph.D. degree in circuits and systems from the
Nanjing University of Posts and Telecommunica-
tions (NJUPT), Nanjing, China, in 2010 and 2016,
respectively. He is currently an associate professor
with the NJUPT. His research interests includes
wireless communications, signal processing for com-
munications and reinforcement learning.
Yilong Hui (Member, IEEE) received the Ph.D.
degree in control theory and control engineering
from Shanghai University, Shanghai, China, in 2018.
He is currently an Associate Professor with the State
Key Laboratory of Integrated Services Networks,
and with the School of Telecommunications Engi-
neering, Xidian University, China. He has published
over 50 scientific articles in leading journals and
international conferences. His research interests in-
clude wireless communication, mobile edge com-
puting, vehicular networks, intelligent transportation
systems and autonomous driving. He was the recipient of the Best Paper
Award of International Conference WiCon2016 and IEEE Cyber-SciTech2017.
Yunhan Li received her B. Eng. degrees in Urban
Planning from both Xi’an Jiaotong-liverpool Uni-
versity, P. R. China, and University of Liverpool,
UK. in 2016, and M. Eng. Degree in Urbanism
(Urban Design) from University of Sydney, Aus-
tralia. Currently, she is a Highway Engineer, Opera-
tion Management Branch of Shaanxi Transportation
Holding Group Co., Ltd., Xi’an, Shaanxi, China.
Her research interests include smart transportation,
vehicle information security and vehicle networking.
Tom H. Luan (Senior Member, IEEE) received
the B.Eng. degree from Xi’an Jiaotong University,
Xi’an, China, in 2004, the M.Phil. degree from The
Hong Kong University of Science and Technology,
Hong Kong, in 2007, and the Ph.D. degree from the
University of Waterloo, Waterloo, ON, Canada, in
2012. He is currently a Professor with the School of
Cyber Engineering, Xidian University, Xi’an. He has
authored or coauthored more than 40 journal articles
and 30 technical articles in conference proceedings.
His research interests include content distribution
and media streaming in vehicular ad hoc networks, peer-to-peer networking,
and the protocol design and performance evaluation of wireless cloud comput-
ing and edge computing. Dr. Luan was the recipient of one U.S. patent. He was
a TPC Member of the IEEE Global Communications Conference, the IEEE
International Conference on Communications, and the IEEE International
Symposium on Personal, Indoor and Mobile Radio Communications, and
the Technical Reviewer for multiple IEEE Transactions, including the IEEE
Transactions on Mobile Computing, the IEEE Transactions on Parallel and
Distributed Systems, the IEEE Transactions on Vehicular Technology, the
IEEE Transactions on Wireless Communications, and the IEEE Transactions
on Intelligent Transportation Systems.
Shui Yu (Fellow, IEEE) obtained his PhD from
Deakin University, Australia, in 2004. He is a Pro-
fessor of School of Computer Science, Deputy Chair
of University Research Committee, University of
Technology Sydney, Australia. His research interest
includes Cybersecurity, Network Science, Big Data,
and Mathematical Modelling. He has published five
monographs and edited two books, more than 500
technical papers at different venues, such as IEEE
TDSC, TPDS, TC, TIFS, TMC, TKDE, TETC, ToN,
and INFOCOM. His current h-index is 72. Professor
Yu promoted the research field of networking for big data since 2013, and
his research outputs have been widely adopted by industrial systems, such
as Amazon cloud security. He is currently serving the editorial boards of
IEEE Communications Surveys and Tutorials (Area Editor) and IEEE Internet
of Things Journal (Editor). He served as a Distinguished Lecturer of IEEE
Communications Society (2018-2021). He is a Distinguished Visitor of IEEE
Computer Society, and an elected member of Board of Governors of IEEE
VTS and ComSoc, respectively. He is a member of ACM and AAAS, and a
Fellow of IEEE.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3313197
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on October 15,2023 at 01:54:22 UTC from IEEE Xplore. Restrictions apply.
