Received January 8, 2020, accepted January 19, 2020, date of publication January 24, 2020, date of current version

February 4, 2020.
Digital Object Identifier 10.1109/ACCESS.2020.2969357

Energy-Efficient UAV-to-User Scheduling to

Maximize Throughput in Wireless Networks
SHAKIL AHMED 1 , (Student Member, IEEE),
MOSTAFA ZAMAN CHOWDHURY 2 , (Senior Member, IEEE),
AND YEONG MIN JANG 3 , (Member, IEEE)
1 Department of Electrical and Computer Engineering, The University of Arizona, Tucson, AZ 85721, USA
2 Department of Electrical and Electronic Engineering, Khulna University of Engineering and Technology, Khulna 9203, Bangladesh
3 Department of Electronics Engineering, Kookmin University, Seoul 02707, South Korea
Corresponding author: Yeong Min Jang ([email protected])
This research was supported by the MSIT (Ministry of Science and ICT), Korea, under the ITRC 547 (Information Technology Research
Center) support program (IITP-2018-0-01396) supervised by the IITP 548 (Institute for Information and Communications Technology
Promotion).

ABSTRACT The unmanned aerial vehicle (UAV) communication is a potential technology to meet the
excessive next-generation cellular users’ demand due to its reliable connectivity and cost-effective deploy-
ment. However, UAV communications have to be energy efficient so that it can save energy. Thus, the UAV
flies sufficiently long enough time to serve the ground users with limited on-board energy. In this paper,
we investigate an energy-efficient UAV communication via designing the UAV trajectory path. We consider
throughput and the UAV propulsion energy consumption jointly. We assume that the UAV flies at a fixed
altitude such that it can avoid tall obstacles. A binary decision variable is assigned to schedule UAV-to-user
communication. First, we derive the UAV-to-user channel model based on the line of sight and non-line of
sight communication links and jointly optimize the trajectory, transmit power, and the speed of UAV; and
UAV-to-user scheduling to maximize throughput. Then, we apply the UAV propulsion energy consumption,
which is a function of the UAV trajectory and speed. Finally, we formulate the UAV energy-efficiency
maximization problem, which is defined as the total bits of information sent to the ground users by consuming
the UAV energy for a given UAV flight duration. The formulated energy-efficiency maximization problem
is non-convex, fractional, and mixed-integer non-linear programming in nature. We propose an efficient
algorithm based on successive convex approximation and classical Dinkelbach method to achieve the optimal
solution of energy-efficient UAV. We present simulation results to validate the efficacy of our proposed
algorithms. The results show a significant performance improvement compared to the benchmark methods.

INDEX TERMS UAV, throughput, UAV propulsion energy, energy-efficiency, UAV-user scheduling.

I. INTRODUCTION The specific system reconfiguration provides flexible

Recently, the unmanned aerial vehicles (UAVs) communi- and reliable connections between the UAV and its users
cation has attracted substantial attention by fifth-generation with reduced power consumption. The BS can be exces-
(5G) and beyond wireless networks researchers due to sively crowded nowadays while serving the exploding traffic
its salient features to support convenient connectivity demands. Moreover, the BS may malfunction anytime. Thus,
with enhanced spectral efficiency [1]. The UAV provides there arises a burning issue to serve the users in the event
on-demand, cost-effective deployment, on-board communi- of too congested traffic, the BS hardware limitation, etc.
cation, and the flexible system reconfiguration compared to Fortunately, the UAV potentially overcomes this issue owing
the base stations (BSs) on the ground [2]–[4]. In particular, to its cost-effective and energy-efficient features. Moreover,
it can support better communication links between air and the UAVs offer a better line of sight (LOS) communication
ground terminals due to less signal blockage and shadowing links by significantly shortening the UAV to user distance.
effects. UAV communication has many potential applications, which
can be categorized as follows:
The associate editor coordinating the review of this manuscript and 1) The UAV works as the aerial BS o support the ground
approving it for publication was Javed Iqbal . users [5]–[7]. In any case, the UAV also provides

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see http://creativecommons.org/licenses/by/4.0/
VOLUME 8, 2020 21215
 S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

reliable connectivity with low latency. In this scenario, channel model. An energy-efficient UAV communication by
the UAV flies while it has a quasi-stationary position optimizing the trajectory is investigated in [18]. Further-
on the air. This kind of UAV communication can apply more, the authors in [18] considered the fixed UAV altitude
in many areas, such as recovering the natural disaster, to optimize energy-efficiency for various trajectory designs.
remote areas, etc. The authors proposed an efficient algorithm to design the
2) The UAV can also work as a relay to support the energy-efficient UAV communication considering the single
distance/remote users [8], [9]. The UAV relay can be user and simple channel model. Though they proved the
mobile/static and is a great choice to support smart improved performance for the single user, they did not con-
cities in the next-generation wireless networks. sider the decision binary variable to schedule the UAV-to-user
3) UAV communication can be used to send/receive in the event of BS malfunctioning.
real-time information [10]. This also suitable for the The authors in [19] proposed the UAV-enabled mobile edge
periodic sensing applications such that the UAV can computing, where the offloading is performed via trajectory
fly over the sensors. This leads to potential network design, and throughput maximization for UAV fixed alti-
lifetime enhancement. tude is studied. Authors in [20] proposed the resource allo-
When the UAV serves like the terrestrial wireless com- cation algorithm to design multicarrier solar-powered UAV
munication infrastructure, it can enlarge the next generation networks, which serves ground users. On the other hand,
wireless capacity due to its reliable uplink and downlink an algorithm is proposed to design the UAV trajectory, con-
communication, mobility, swift deployment, and on-demand sidering energy-efficient UAV communication in [21]. They
service, etc. Moreover, the UAV deployment as an aerial BS also aimed to secure the wireless network via UAV. In [22],
can also compensate signal blockage due to its LOS channel the authors studied the underlaid D2D communications. They
advantages. Furthermore, the UAV deployment can limit the also investigate the spectrum sharing with UAV-assisted wire-
higher transmit power compared to the BS on the ground less networks. The UAV implementation can substitute the
because the UAV can easily adjust its mobility and altitude traditional BS of the cellular networks due to its innovative
based on the user’s location. These features eventually pro- method with flexible, robust, and low latency wireless com-
vide a solution of the energy-efficient UAV deployment to munication [23].
serve the users on the ground in the event of the malfunction- The authors in [24] studied the energy-efficient machine-
ing of the BS on the ground [5], [11]. to-machine communication system, where they assumed the
Though the UAV has many potential features, it still has fixed UAV flying speed. A novel scheme is proposed in [25],
the challenge of energy-efficient UAV deployment due to its where the users and the UAV share the same frequency,
limited on-board energy. Thus, our primary motivation is to and there is no user-to-user communication. In their system,
design the energy-efficient UAV communication when the the total number of orthogonal channels is higher than the
BS has the hardware limitation. It is essential to maximize total number of users, and the received signal to interference
the amount of information per unit UAV energy consump- and noise ratio (SINR) is higher than the minimum SINR
tion during the UAV flight time to achieve optimal energy level. The author in [26] investigated the average worst-case
efficiency. This is because of the fixed weight and size of secrecy rate considering a simple channel model in the pres-
UAV, which may limit the overall system performance. Thus, ence of the unknown adversaries. However, energy-efficient
designing energy-efficient UAV communication is more chal- UAV communication is not studied in the proposed sys-
lenging than energy-efficient terrestrial BS communication tem model in [26]. None of [15]–[18], [24]–[26] considers
infrastructure. Unfortunately, the UAV has limited power the decision variable for the UAV-to-users scheduling and
resources on-board due to its wight, flight time, etc. How- LOS/NLOS based channel model in their investigations.
ever, the UAV is required sending the maximum amount of In this paper, we maximize the system throughput and
information with these limited resources to attain the quality energy-efficiency of the UAV via optimizing the UAV tra-
of service (QoS). jectory optimization by considering the air-to-ground chan-
Energy-efficient UAV wireless communication has been nel based on LOS and non-line of sight (NLOS), and the
an active research topic lately. There are many examples UAV propulsion energy consumption. We consider orthog-
of simplified models for UAV-assisted networks in the lit- onal frequency-division multiple access (OFDMA) in the
erature [12]–[26]. For example, the UAV is not designed proposed system. Table 1 describes the mathematical symbol
to support the distant users on the ground because of its used in the paper. The main contributions of our paper are
limited energy supplies [14] unless the UAV works as the described as follows:
relay. However, the UAV moves on-demand basis, which is • We formulate the channel model based on both LOS
a good fit for nearby users. Throughput optimization using and NLOS communications links. We use LOS and
multiple antennae [15] and system throughput maximization NLOS based channel model to formulate throughput
using the UAV [16] were studied based on trajectory opti- maximization problem via designing the UAV trajec-
mization. However, the energy-efficient UAV deployment is tory while multiple users are present on the ground.
not considered in these works. The authors in [17] investi- The binary decision variable indicates the connec-
gated the energy-efficiency approach by considering a simple tivity of the UAV-to-user. However, the formulated

21216 VOLUME 8, 2020

S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

TABLE 1. List of mathematical symbols used in the paper. optimizing the UAV trajectory radius, UAV transmit
power, UAV-users scheduling, and UAV mobility. How-
ever, the formulated problem is non-convex, fractional,
and MINLP.
• To reduce the complexity of the formulated energy-
efficiency optimization problem, we perform the
approximation of the formulated problem. The opti-
mization problem is solved effectively by our proposed
efficient algorithm based on SCA, which tackles the con-
convexity. Moreover, the Dinkelbach method deals with
the fractional problem. Moreover, cvx solver, mosek, can
tackle MINLP.
• Finally, we present the improved performance of the pro-
posed algorithm via the simulation results by using the
optimal parameter configuration for the UAV trajectory,
the UAV height, decision variable, and UAV mobility.
The rest of the sections of the paper are organized
as follows: We present the system model in Section II.
In Section III, the throughput maximization problem is for-
mulated and solved. The UAV optimal energy-efficiency
maximization problem is analyzed in Section IV. We
propose two efficient algorithms, which solve throughput
and energy-efficiency maximization in Section V. Finally,
the proposed schemes are validated via simulation results in
Section VI. We conclude in Section VII.
Notations: Lower case boldface letter, italic letters, log2 (.),
k . k, [.], (.0 ), (.)i , (.)i+1 ,and tan(.)−1 denote vectors, scalars,
logarithm with base 2, norm, function of time, transpose
of vector, i iteration, (i + 1) iteration, and inverse tangent
function, respectively.

II. SYSTEM MODEL AND KEY DEFINITIONS

Nowadays, the base stations (BSs) are too congested with
next-generation users, which may prevent users from attain-
ing the required quality of service (QoS). As an alternative,
the unmanned aerial vehicles (UAVs) can support excessive
users, especially when the BS has the hardware limitation or
the malfunctioning. In that case, the UAV can also serve as a
terrestrial wireless network infrastructure [28].

A. SYSTEM MODEL
In Fig. 1, we consider a wireless communication system in
a geographical area, containing a UAV, and a set of multiple
users U on the ground, where U = {1, 2, 3, . . . , U } and U is
the total number of users. The UAV is dedicated to supporting
a U set of users. In our investigation, the UAV dynamically
moves to serve the users. The UAV flight duration to serve
problem is non-convex and mixed-integer non-linear the users on the ground is 0 ≤ t ≤ T .
problems (MINLP) problems. We tackle to throughput The locations of the users are entirely known to the UAV,
maximization problem by using the successive con- which is used for designing the trajectory. The location of the
vex approximation (SCA) [27]. cvx solver, mosek can static user u on the ground is denoted as (xu , yu ). We investi-
address MINLP. gate the LOS and NLOS communication links based channel
• We investigate the UAV propulsion energy consumption, model, which has negligible shadowing and multipath effect.
which is a function of UAV trajectory and speed. The Thus, we leave these issues as our future work. We con-
UAV energy-efficiency problem is formulated via jointly sider the time varying location of the UAV is (x(t), y(t)).

VOLUME 8, 2020 21217

 S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

such as urban, suburban, etc. Moreover, fc is the carrier

frequency, α is the path loss exponent, and cl is the speed
of light. The path loss model for the NLOS links between the
UAV and the user u is:
4π fc h α u
 
u
PLNLoS (t) = δ2 pNLoS (t), (6)
cl
where δ2 is the excessive path loss coefficient for NLOS links,
which is also the environment constant depending on the area
type. However, it is impossible to determine the path loss type
experienced by the UAV and the user u by only knowing ru if
it is LOS or NLOS paths. Thus, to find the path loss model,
we determine an average path loss model for these two types.

D. THROUGHPUT
Now, we achieve the average path loss using (3) - (6). Thus,
FIGURE 1. The UAV flies over the users on the ground. A binary decision
variable schedules UAV-to-user communication in the event of the BS
the average path loss between the UAV and the user u is:
malfunctioning or hardware limitation, etc. Thus, the UAV works as
4π fc h α
 
terrestrial wireless network infrastructure. u
PLavg (t) = (δ1 puLoS (t) + δ2 puNLoS (t)). (7)
cl
The distance between the UAV and user u is:
q The average channel gain for the user u is the inverse of (7)
ru (t) = (x(t) − xu )2 + (y(t) − yu )2 + h. (1) [4, Eq. (3)]. The channel gain is expressed as follows:
where h is the UAV fixed altitude. In particular, ru can bal- 1
hcu (t) = u (t)
, (8)
ance throughput maximization and eventfully results in the PLag
energy-efficient UAV.
where hcu (t) represents the channel gain based on LOS and
B. THE UAV-TO-USER SCHEDULING NLOS communication links. We apply the Shannon capacity
Binary decision variable, bu (t), is used as an indication of the to define the throughput for user, u during the UAV flight time
connectivity between the user u and the UAV. If bu (t) is 1, 0 ≤ t ≤ T as follows:
pu (t)hcu (t)
 
the user u is supported by the UAV. If bu (t) equals 0, then
cu (t) = bu (t) log2 1 + , (9)
otherwise. We can express as follows: σ2
(
1, the user u supported by the UAV, where pu is the UAV to user u transmit power and σ 2 defines
bu (t) = (2)
0, otherwise. the additive white Gaussian noise (AWGN) power at the
receiver. (9) also considers the UAV-to-user u scheduling
C. CHANNEL MODELING while calculating the channel gain.
The channel between the UAV and the users consists of both
LOS and NLOS paths. Firstly, the probability of existing a E. UAV PROPULSION ENERGY CONSUMPTION
LOS link [29] between UAV and user u is: The amount of energy consumed by UAV is propulsion
1 energy consumption, which has a significant impact on sys-
puLoS (t) = , (3) tem performance. Energy consumption due to the signal pro-
1 + ψ exp 8 π tan−1 ruh(t) − ψ
 180

cessing energy, radiation, and circuitry energy consumption
where 8 and ψ are constant values depending on the envi- of the UAV has a negligible impact on the overall system
ronment such as urban, and suburban. The probability of performance. If UAV is flying with fixed wings with no
NLOS between the UAV and user u is: abnormalities, such as no engine abnormality to generate a
backward thrust against forwarding speed, then the required
puNLoS (t) = 1 − puLoS (t). (4)
total propulsion energy consumption is a function of ru during
The increment of ru (t) results in the decrements of puLoS (t) 0 ≤ t ≤ T period. Moreover, for energy-efficient trajectory
in (3) and the increment of puNLoS (t) in (4). However, the path design, UAV velocity and energy consumption should have
loss model for the LOS link between the UAV and the user u an optimal trade-off, which can be obtained by designing
is [4]: the UAV trajectory. Thus, UAV energy consumption [18] due
4π fc h α u to user u for a circular trajectory path can be expressed as
 
PLLoS (t) = δ1
u
pLoS (t), (5) follows:
cl
b k vu (t) k3
 
where δ1 is the excessive path loss coefficient for LOS links, b
3
eu (t) = a kvu (t)k + + , (10)
which is an environment constant depending on the area type, k vu (t) k ru (t)2 g2

21218 VOLUME 8, 2020

S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

where a and b are both UAV weight dependent constants. transparent and understandable, we summarize these essen-
Furthermore, vu is the velocity, while flying over the user u. g tial steps as follows. Firstly, the transformation of P1−→P2
is gravitational constant, while ru is the distance between the is performed, where we employ the discrete space represen-
UAV and the user u. tation, i.e. replacing t by n. Secondly, the transformation of
P2−→P3 is developed. We add the auxiliary variable to the
F. ENERGY-EFFICIENCY objective function. However, the transformed problem P3 is
We maximize the UAV energy-efficiency by jointly opti- still a non-convex problem due to its one of the newly added
mizing throughput and energy consumption. First we for- constraint. Therefore, we approximate (12) to make them
mulate the UAV energy-efficiency problem, δuav ee . Thus, convex. We then propose an algorithm based on SCA to solve
energy-efficiency for the total number of users during the the problem P1 optimally.
UAV flight time is formulated by combining (9) and (10) as:
PU PT 1) TRANSFORMATION OF P1 to P2
t=1 cu (t)
δuav (t) = Pu=1
ee
U PT
. (11) The problem P1 can be rewritten as follows:
u=1 t=1 eu (t)
U T
1 XX
max cu [n], (13a)
x[n],y[n],bu [n],pu [n],vu [n] UN
III. THROUGHPUT MAXIMIZATION u=1 t=1
A. PROBLEM FORMULATION s.t. 0 ≤ pu [n] ≤ u ,
pmax ∀n, (13b)
We now formulate throughput maximization problem under ru [0] = ru [N ], (13c)
the constraints of the UAV trajectory, binary variable, trans- XU
mit power, and velocity. This maximization problem is: bu [n] ≤ 1, ∀n, (13d)
U X
T u=1
1 X
P1 : max cu (t) (12a) (x[n + 1]−x[n])2 +(y[n + 1]−y[n])2
x(t),y(t),bu (t),pu (t) UT
u=1 t=1
≤ (vu ρ)2 , ∀n, (13e)
u ,
s.t. 0 ≤ pu (t) ≤ pmax ∀t, (12b)
bu [n] ∈ {1, 0}. (13f)
ru (0) = ru (T ), (12c)
XU where
bu (t) ≤ 1, ∀t, (12d) pu [n]hcu [n]
 
cu [n] = bu [n] log2 1 + . (14)
u=1 σ2
(x(t + 1) − x(t))2 + (y(t + 1)−y(t))2 ≤ v2u ,
We explain the detailed approximation in (13), as follows:
(12e) we divide the UAV flight time period T into N equal and
bu (t) ∈ {1, 0}. (12f) static number of time slots with a slot size ρ = NT and
n = 1, 2, 3, ...N . The time slots are represented by N vector
where (12b) defines the UAV transmit power control limit
sequences for designing the trajectory of the UAV. Further-
during 0 ≤ t ≤ T period. pmax u is the maximum UAV
more, (x, y), vu , and bu can be approximated for the each
transmit power to user u. In (12c), the UAV returns to its initial
time interval, ρ. Thus, (13b) - (13f) can be expressed as the
location, where ru (0) is initial and ru (T ) is final locations of
equivalent expressions to (12b) - (12f). The binary variable
the UAV. Moreover, bu (t) allows user u to be served by the
in (13f) is described as the continuous variables [30]. Thus,
UAV. The UAV and user u association, scheduled by bu (t),
it can be expressed as follows:
is described in (12f). We define (12d) as follows. Recall that
a binary variable bu (t) ∈ {1, 0} schedules the UAV and user 0 ≤ bu [n] ≤ 1, ∀n, ∀U . (15)
u communication, which is shown in (2). Here, if bu (t) is 1,
then the UAV is connected to support the user u. Thus, (12d) The relaxation of binary variable bu [n] results that the
defines the UAV and U set of users connectivity. The mobility objective function serves as a upper bound for (16), Moreover,
constraint of the UAV is defined in (12e). (13e) is the UAV mobility constraint. The new optimization
However, it is readily observed that P1 is not a convex problem is reformulated as P2 :
problem with fractional objection function. Thus, P1 can- U N
1 XX
not be solved directly due to 1) continuous function, ru (t), P2 : max cu [n],
x[n],y[n],bu [n],pu [n],vu [n] UT
2) continuous UAV flight time, and 3) MINLP nature. Hence, u=1 n=1
we approximate P1 to the new problem to solve it optimally. s.t. (13b) − (13e), (15). (16a)
The transformation of problem P1 is presented in the next
subsection. 2) TRANSFORMATION OF P2 to P3
P2 is not convex due to its objective function. We approx-
B. PROBLEM TRANSFORMATION imate the objective function of (16a) to a convex function
The transformation of the P1 is performed in several steps, by introducing a new variable. Throughput based on Shan-
resulting in the final approach as convex. To make this process non capacity is defined in (14). Thus, we approximate the

VOLUME 8, 2020 21219

 S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

channel gain, hcu [n], into fu [n] for user u.The throughput with IV. UAV OPTIMAL ENERGY EFFICIENCY
the new variable is appeared as follows: A. PROBLEM FORMULATION
  We now formulate the energy-efficiency maximization prob-
cn [n] = bu [n] log2 1+fu [n] − log2 (σ 2 ).
u
(17) lem. We use the optimal throughput maximization problem
solution from Section III. However, we still need the UAV
propulsion energy consumption from (10) and is continuous
where (17) is equivalent to (14).
in nature. We replace the continuous-time series into the
Proof: The proof is given in Appendix VII.
discrete state-space representation. The UAV energy efficient
Now, fu [n] can be expressed as follows:
maximization problem is:
 
1 1 1 U X
N  n 
fu [n] = , (18) X cu [n]
,
Wru [n] δ1 puLoS [n] + δ2 puNLoS [n] P4 : max
opt eu [n]
(24a)
x[n],y[n],fu [n] u=1 n=1
c vu [n],pu [n],bu [n]
where W = 4πf c
. New variable, fu [n] can be replaced with s.t. vmin ≤ vu [n] ≤ vmax , ∀n,
ru [n] in (13b), which is the UAV transmit power control.
opt
Therefore, we have the new upper limit fu,max [n] can be (13b) − (13e), (15), (23b). (24b)
defined as follows: where
b k vu [n] k3
   
1 1 1 b
fu [n] = . (19) eu [n] = a k vu [n] k3 + + . (25)
Wrmax δ1 puLoS [n] + δ2 puNLoS [n] k vu [n] k ru [n]2 g2
where (24b) defines the limit of vu in 0 ≤ t ≤ T period, where
It is seen that (19) is a concave-surrogate function. We
vmin and vmax are the minimum and maximum UAV velocity,
make a further explanation of the concave surrogate fum [n]
opt respectively. P4 is not a convex problem due to its fractional
for fu [n] in Theorem 1.
objection function. Thus, P4 cannot be solved directly due
Theorem 1: At ith iteration, the concave surrogate fum [n]
opt to 1) intractable fractional energy-efficiency problem, and
for fu [n] is:
2) MINLP nature. Hence, we approximate P4 to the new

puLoSi [n]
 problem to solve it optimally. The transformation of problem
fum [n] = XA [n]+XB [n]+ u ≤ fuopt [n], (20) P4 is presented in the next subsection.
pNLoSi 2 [n]
B. PROBLEM TRANSFORMATION
where
The transformation of the P4 is also performed in several
2δ22 − δ15 puLoSi 2 [n]puLoSi+1 3 [n] steps, resulting in the final approach as convex. The trans-
XA [n] = , (21) formation of P4 −→ P5 is developed by adding the UAV
δ22 δ14 puNLoSi+1 2 [n]puLoSi 2 [n]puLoSi+1 2 [n]
trajectory variable to the denominator. By doing so, these can
kni couple with other trajectory variables. Hence, it transforms
XB [n] =  .
the denominator as a convex function. Finally, the transfor-
δ14 puLoSi+1 2 [n]puNLoSi+1 2 [n] puLoSi [n] + puNLoSi [n]2 mation of P5−→P6 is derived, where the Dinkelbach method
(22) is applied with a numerical constant, which is iteratively
updated. We propose an efficient algorithm based on SCA
kni is defined as follows: and Dinkelbach to solve the problem P4 optimally.

−2−δ14 puLoSi+1 2[n]puNLoSi+1 2 [n](3puLoSi [n]+puNLoSi [n]). 1) TRANSFORMATION OF P4 to P5

Due to the non-convexity of the objective function, we add
Finally, the problem P2 is transformed to P3 with newly a new variable, vnv , to the denominator [11] in (25). Hence,
added optimizing variables and constraints as follows: optimizing the new variable, vnv , meaning that we jointly
optimize the trajectory variable, vu , and ru . We rewrite the
U N
1 XX n UAV propulsion energy consumption in (25) as follows:
P3 : max cu [n], (23a)
x[n],y[n],fu [n] UT  3 
3 b kvnv [n]k b
vu [n],pu [n],bu [n] u=1 n=1 n
eu [n] = a kvnv [n]k + 2 + . (26)
opt
s.t. 0 ≤ fu [n] ≤ fu,max [n], n = 1, 2, 3, . . . N , g ru [n]2 vu [n]
(13b) − (13e), (15). (23b) The problem P4 is transformed to P5 with newly added
optimizing variables and constraints as follows:
U X
N
The constraint in (23b) is proved to be convex due to the X cn [n]
newly added variable. Thus, (23) is a convex problem and P5 : max u
, (27a)
opt
x[n],y[n],fu [n] enu [n]
ready to solve using the standard optimization toolbox [31]. u=1 n=1
vu [n],pu [n],bu [n],vnv [n]

21220 VOLUME 8, 2020

S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

s.t. k vu [n] k≤ vmax , n = 1, 2, 3, . . . N , (27b) Algorithm 1 Solution of Throughput i.e., P1

vnv [n] ≤k v[n] k ,
2 2
n = 1, 2, 3, . . . N , 1: Inputting : σ 2 , 8, ψ, and fc .
(13b) − (13e), (15), (23b). (27c) 2: Initializing : iterative number i = 1, ru [n], pu [n], bu [n],
and vn [n]
3: Optimization :
The constraints in (27b) is proved to be convex due to 4: repeat 1
opt
the added variable. However, we still have the non-convex 5: Calculate fu [n], fu,max [n], puLoS [n], and puNLoS [n]
constraint (27c) due to its non-linearity. Therefore, we make 6: Update i ←− i + 1
an approximation to reformulate (27c) to make it convex. Our 7: until convergence
approximation is expressed as: 8: repeat 2
  9: Solve P3 using cvx and cvx solver mosek
k vnv [n] k≤ − k vi+1 [n] k +4vi vi+1 [n] ,
2
(28) 10: Update i ←− i + 1
11: until convergence
where (28) is defined as follows. Firstly, kvnv [n]k is convex.
Moreover, kvi+1 [n]k is differential function w.r.t. kvi [n]k, for
any local point vi [n] obtained at i-th iteration [11], which is Algorithm 1 has complexity, which is polynomial in the
approximated from (27c). Thus, newly added variable vnv [n] worst case [34]. This is because Algorithm 1 solves the con-
in (26) makes the problem P5 optimally solvable by using the vex optimization problem at each iteration. Moreover, achiev-
SCA method. However, we also need to tackle the fractional ing the UAV optimal trajectory is found offline before the
problem in order to get the optimal solution of the optimiza- UAV dispatch at the ground control station. This also has a
tion problem, which is discussed in the following subsection. high computational capability.

2) TRANSFORMATION OF P5 to P6 Algorithm 2 Solution of Energy-Efficiency i.e., P4

To solve the problem P5 , we would apply the Dinkelbach 1: Inputting : σ 2 , a, b, g, 8, ψ, and fc .
method to a fractional problem and then formulate the P6 with 2: Initializing : iterative number i = 1, ru [n], pu [n], bu [n],
the new objective function. We briefly explain the Dinkelbach and vn [n]
method in the following (interested readers can find more 3: Optimization :
P(r) repeat 1
information in [32], [33]). Here, f (r) = Q(r) is described as 4:
opt
f (r)=P(r)−λQ(r) under all convex constraints [12], where 5: Calculate fu [n], fu,max [n], vnv [n], puLoS [n], and
u
λ is a constant. This value is iteratively updated by λj+1 = pNLoS [n]
prj
6: Update i ←− i + 1
Qrj , where j is the iterative index. This process guarantees
7: until convergence
the convergence, and hence, the locally optimal solution is
8: repeat 2
achieved. Hence, P5 can be approximated as P6 , where the
9: Solve P6 for given cnu [n], enu [n], and iteratively
objective function and all constraints are convex as follows:
updated λm , using cvx and cvx solver mosek
X N 
U X  10: Update i ←− i + 1
P6 : max cnu [n] − λi enu [n] , 11: until convergence
xi+1 [n],yi+1 [n],vi+1 [n],
opt u=1 n=1
fu,i+1 [n],vnv [n],bu [n],pu [n]
s.t. (13b) − (13e), (15), (23b), (27b), (28).
B. ENERGY-EFFICIENCY MAXIMIZATION ALGORITHM
(29a)
The efficient algorithm, based on SCA and Dinkelbach
where λi is a new numerical
 value that can be iteratively methods, can solve the UAV energy-efficiency maximization
PU PN cnu [n] problem optimally, which is summarized in Algorithm 2.
updated as u=1 n=1 en [n] . P6 is a convex problem and
u The classical Dinkelbach method tackles the fractional prob-
is ready to solve using the standard optimization toolbox. lem. The cvx solver mosek in the algorithm can help to
solve the problem efficiently. The proposed algorithm to
V. PROPOSED ALGORITHMS maximize the UAV energy efficiency is summarized in
A. THROUGHPUT MAXIMIZATION ALGORITHM Algorithm 2.
The efficient algorithm based on the SCA method can Algorithm 2 also has complexity polynomial in the worst
solve the problem P3 optimally, which is summarized in case as Algorithm 2 solves the convex optimization problem
Algorithm 1. Moreover, there is a binary variable of bu [n], at each iteration.
which makes the problem MINLP. The cvx solver mosek in
the algorithm can help to solve the problem efficiently. The VI. SIMULATION RESULTS
proposed algorithm to maximize throughput is summarized We present simulation results in this section to show the
in Algorithm 1. improved performance of the proposed scheme. We compare

VOLUME 8, 2020 21221

 S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

TABLE 2. List of parameters used in the simulation results.

FIGURE 2. The UAV trajectory design. ee .

FIGURE 3. Only one environment dependent maximum point of δuav

the proposed scheme with a static system and an uncon- shorter UAV flight time, many of the ground users also reside
strained approach. The static scheme is defined as the objec- under the UAV hovering path. Due to the closeness of the
tive function is maximized after achieving the optimal UAV users to the UAV hovering path, we achieve the improved
fixed trajectory radius. We evaluate the system model per- performance of both throughput and energy-efficiency max-
formance for a suburban environment. The list of parameters imization problem. The UAV trajectory path both for shorter
used in the simulation is presented in Table 2. and higher UAV flight time proves the supremacy of the
Fig. 2 shows the UAV trajectory design for the given UAV proposed algorithms.
fight time. The UAV has a different hovering path for dif- We study the impact of the UAV-to-user, u distance,
ferent UAV flight time. When the UAV flight time is higher, i.e., ru on the proposed δuav
ee performance in Fig. 3. To do

the hovering path is larger. Thus, for higher UAV flight time, u u
so, we determine pLoS , pNLoS , and UAV energy consumption.
it can support more users on the ground. Fig. 2 also shows In various environments, it can be readily found that proposed
the location of the users while the UAV is hovering. We δuav
ee has a maximum global value in each area. Fig. 3 clearly

consider the random distribution of the users in our proposed shows the maximum point of δuav ee in various areas such as the

model. From Fig. 2, it is seen that the ground users reside suburban, urban, dense urban, and ultra-dense metropolitan
under/nearby the UAV hovering trajectory path. Even for area. We obtain the best performance in the suburban area,

21222 VOLUME 8, 2020

S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

TABLE 3. Performance comparison for energy-efficient UAV algorithms.

FIGURE 4. Convergence of algorithm 1.

FIGURE 6. Throughput versus time.

FIGURE 5. Convergence of algorithm 2.

FIGURE 7. Energy-efficiency maximization versus the UAV-user distance.

as shown in Fig. 3. So, from now, we only consider this area

for the remaining results. We show the optimizing variables the initial drop up to 10% of our proposed throughput from
and compared it with [18] in Table 3. α = 4 to α = 2. That observation proves the superiority
We also present the convergence of Algorithm 1 and of our proposed throughput and compares it with the uncon-
Algorithm 2 in Fig. 4 and Fig. 5, respectively. Both straint and static throughput.
Algorithm 1 and Algorithm 2 are proved to be efficient due We compare the performance between energy-efficiency
to their fast convergence. maximization, static energy-efficiency, and unconstraint
Throughput vs. time is shown in Fig. 6. As LOS and energy-efficiency problem in Fig. 7. For α = 4, there is a sig-
NLOS communication links are both considered to design nificant improvement in energy-efficiency for our proposed
the channel model, various values of path loss exponent α = energy-efficient Algorithm 2 compared to the benchmark
2, 3, 4 have been studied to analyze the best performance. methods. When ru = 150 m or more substantial, both our
We investigate the improved performance of throughput max- proposed δuavee and unconstraint energy-efficiency problem

imization based on Algorithm 1 compared to other schemes. are investigated as impractical due to their low performance.
Our proposed throughput maximization illustrates the signif- On the other hand, the static scheme has low performance
icant improvement for α = 2, 3, 4. Our proposed throughput due to the optimal static location of the UAV. The UAV is
almost maintains a constant and higher value though there is not expected to cover larger areas when the UAV altitude

VOLUME 8, 2020 21223

 S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

is too high. though Thus, proposed energy-efficient UAV is and

proved useful for the 5G and beyond wireless networks.
2(ri − r) 2ri
VII. CONCLUSION
>B (r | ri ) = − 2. (37)
y2 x
The UAV provides a better communication link based on LOS Now, (36) and (37) become the same at r = ri . The second
and NLOS, which guarantees a better QoS. In this paper, surrogate condition is satisfied by (31).
an energy-efficient UAV serves the users via jointly optimiz- Thus, the proof is complete.
ing the UAV trajectory path, the UAV to users scheduling,
the UAV transmit power, and the UAV mobility. We design REFERENCES
the UAV-user channel model based on LOS and NLOS com- [1] M. Z. Chowdhury, M. Shahjalal, S. Ahmed, and Y. M. Jang, ‘‘6G wireless
munication links. A binary variable is introduced to schedule communication systems: Applications, requirements, technologies, chal-
the connectivity between UAV to the user. Before formu- lenges, and research directions,’’ Sep. 2019, arXiv:1909.11315. [Online].
Available: https://arxiv.org/abs/1909.11315
lating the energy-efficient problem formulation, we develop [2] H. Shakhatreh, A. H. Sawalmeh, A. Al-Fuqaha, Z. Dou, E. Almaita,
a throughput maximization problem and propose an effi- I. Khalil, N. S. Othman, A. Khreishah, and M. Guizani, ‘‘Unmanned
cient algorithm to solve it optimally. We also formulate aerial vehicles (UAVs): A survey on civil applications and key research
challenges,’’ IEEE Access, vol. 7, pp. 48572–48634, 2019.
an energy-efficiency maximization problem via optimiz- [3] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, ‘‘Unmanned aerial
ing throughput and UAV propulsion energy. We also for- vehicle with underlaid device-to-device communications: Performance and
mulate the UAV propulsion energy expression. We derive tradeoffs,’’ IEEE Trans. Wireless Commun., vol. 15, no. 6, pp. 3949–3963,
Jun. 2016.
the energy-efficient maximization problem. The problem is [4] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, ‘‘Mobile unmanned
investigated as a non-convex fractional, and MINLP problem. aerial vehicles (UAVs) for energy-efficient Internet of Things communi-
We propose an efficient algorithm based on SCA and Dinkel- cations,’’ IEEE Trans. Wireless Commun., vol. 16, no. 11, pp. 7574–7589,
Nov. 2017.
bach method, non-convexity, fractional, and MINLP prob- [5] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, ‘‘Efficient deployment
lems are tackled by SCA, Dinkelbach, and cvx solver mosek, of multiple unmanned aerial vehicles for optimal wireless coverage,’’ IEEE
respectively. Simulation results prove the supremacy of the Commun. Lett., vol. 20, no. 8, pp. 1647–1650, Aug. 2016.
[6] C. Qiu, Z. Wei, Z. Feng, and P. Zhang, ‘‘Joint resource allocation, place-
algorithm in terms of the energy-efficiency and throughput ment and user association of multiple UAV-mounted base stations with
maximization. We claim our proposed model is an excel- in-band wireless backhaul,’’ IEEE Wireless Commun. Lett., vol. 8, no. 6,
lent fit to design the energy-efficient UAV communication pp. 1575–1578, Dec. 2019.
[7] S. Yin, Y. Zhao, and L. Li, ‘‘Resource allocation and basestation placement
in 5G and beyond wireless technology due to the consider- in cellular networks with wireless powered UAVs,’’ IEEE Trans. Veh.
ation of the feasible channel model and the UAV trajectory Technol., vol. 68, no. 1, pp. 1050–1055, Jan. 2019.
design. [8] P. Zhan, K. Yu, and A. L. Swindlehurst, ‘‘Wireless relay communications
with unmanned aerial vehicles: Performance and optimization,’’ IEEE
Trans. Aerosp. Electron. Syst., vol. 47, no. 3, pp. 2068–2085, Jul. 2011.
APPENDIX [9] W. Zhao, M. Ammar, and E. Zegura, ‘‘A message ferrying approach for
PROOF data delivery in sparse mobile ad hoc networks,’’ in Proc. 5th ACM Int.
Symp. Mobile Ad Hoc Netw. Comput., 2004, pp. 187–198.
Consider a function of r [25], [33] as [10] C. Zhan, Y. Zeng, and R. Zhang, ‘‘Energy-efficient data collection in UAV
enabled wireless sensor network,’’ IEEE Wireless Commun. Lett., vol. 7,
1 1
>A (r) = , (30) no. 3, pp. 328–331, Jun. 2018.
x [11] V. Chetlur and H. Dhillon, ‘‘Downlink coverage analysis for a finite 3D
wireless network of unmanned aerial vehicles,’’ IEEE Trans. Commun.,
where x =k r k2 +y. x > 0 and y > 0 are constant. vol. 65, no. 10, pp. 4543–4558, Jul. 2017.
We can write the following expression [12] F. Cheng, S. Zhang, Z. Li, Y. Chen, N. Zhao, F. R. Yu, and V. C. M. Leung,
‘‘UAV trajectory optimization for data offloading at the edge of multiple
1 cells,’’ IEEE Trans. Veh. Technol., vol. 67, no. 7, pp. 6732–6736, Jul. 2018.
4>B (r | ri ) = g1 + g2 , (31) [13] F. Cheng, G. Gui, N. Zhao, Y. Chen, J. Tang, and H. Sari, ‘‘UAV-relaying-
assisted secure transmission with caching,’’ IEEE Trans. Commun., vol. 67,
where no. 5, pp. 3140–3153, May 2019.
[14] S.-Y. Lien, K.-C. Chen, and Y. Lin, ‘‘Toward ubiquitous massive accesses
k r k2−krik2+2r 0ri in 3GPP machine-to-machine communications,’’ IEEE Commun. Mag.,
g1 = (32)
y2 vol. 49, no. 4, pp. 66–74, Apr. 2011.
[15] J. Cui, H. Shakhatreh, B. Hu, S. Chen, and C. Wang, ‘‘Power-efficient
x +2 k r k2 −2r 0ri
g2 = . (33) deployment of a UAV for emergency indoor wireless coverage,’’ IEEE
x2 Access, vol. 6, pp. 73200–73209, 2018.
[16] Y. Zeng, R. Zhang, and T. J. Lim, ‘‘Throughput maximization for UAV-
For ri , >B (r) needs to satisfy the following conditions enabled mobile relaying systems,’’ IEEE Trans. Commun., vol. 64, no. 12,
pp. 4983–4996, Dec. 2016.
4>A (ri ) = 4>B (r | ri ), (34) [17] D. H. Choi, S. H. Kim, and D. K. Sung, ‘‘Energy-efficient maneuvering
and communication of a single UAV-based relay,’’ IEEE Trans. Aerosp.
>A (ri ) ≥ >B (r | ri ). (35) Electron. Syst., vol. 50, no. 3, pp. 2320–2327, Jul. 2014.
[18] Y. Zeng and R. Zhang, ‘‘Energy-efficient UAV communication with tra-
Also, it can be written as a gradient of 4>A (r) and 4>B (r | jectory optimization,’’ IEEE Trans. Wireless Commun., vol. 16, no. 6,
ri ) w.r.t. r pp. 3747–3760, Jun. 2017.
[19] F. Zhou, Y. Wu, H. Sun, and Z. Chu, ‘‘UAV-enabled mobile edge
2r computing: Offloading optimization and trajectory design,’’ Feb. 2018,
>A (r) = , (36) arXiv:1802.03906. [Online]. Available: https://arxiv.org/abs/1802.03906
x2
21224 VOLUME 8, 2020
S. Ahmed et al.: Energy-Efficient UAV-to-User Scheduling to Maximize Throughput in Wireless Networks

[20] Y. Sun, D. Xu, D. W. K. Ng, L. Dai, and R. Schober, ‘‘Optimal 3D- MOSTAFA ZAMAN CHOWDHURY (Senior
trajectory design and resource allocation for solar-powered UAV commu- Member, IEEE) received the B.Sc. degree in
nication systems,’’ IEEE Trans. Commun., vol. 67, no. 6, pp. 4281–4298, electrical and electronic engineering from the
Jun. 2019. Khulna University of Engineering and Technol-
[21] Y. T. B. Cai, Z. Wei, R. Li, D. W. K. Ng, and J. Yuan, ‘‘Energy- ogy (KUET), Bangladesh, in 2002, and the M.Sc.
efficient resource allocation for secure UAV communication sys- and Ph.D. degrees in electronics engineering from
tems,’’ Jan. 2019, arXiv:1901.09308. [Online]. Available: https://arxiv. Kookmin University, South Korea, in 2008 and
org/abs/1901.09308
2012, respectively. In 2003, he joined the Elec-
[22] H. Wang, J. Wang, G. Ding, L. Wang, T. A. Tsiftsis, and P. K. Sharma,
trical and Electronic Engineering Department,
‘‘Resource allocation for energy harvesting-powered D2D communication
underlaying UAV-assisted networks,’’ IEEE Trans. Green Commun. Netw., KUET, as a Lecturer, where he is currently a Pro-
vol. 2, no. 1, pp. 14–24, Mar. 2018. fessor. He was a Postdoctoral Researcher with Kookmin University, from
[23] H. Wang, G. Ding, F. Gao, J. Chen, J. Wang, and L. Wang, ‘‘Power control 2017 to 2019. He has published around 125 research articles in national
in UAV-supported ultra dense networks: Communications, caching, and and international conferences and journals. His research interests include
energy transfer,’’ IEEE Commun. Mag., vol. 56, no. 6, pp. 28–34, Jun. 2018. small-cell networks, the Internet of Things, eHealth, 5G, and beyond (5GB)
[24] C.-Y. Tu, C.-Y. Ho, and C.-Y. Huang, ‘‘Energy-efficient algorithms and communications, artificial intelligence (AI), machine learning, and optical
evaluations for massive access management in cellular based machine to wireless communication. He has served as a TPC Member for several
machine communications,’’ in Proc. IEEE Veh. Technol. Conf. (VTC Fall), IEEE conferences. In 2008, he received the Excellent Student Award from
San Francisco, CA, USA, Sep. 2011, pp. 1–5. Kookmin University. His three articles received the Best Paper Award at
[25] S. Eom, H. Lee, J. Park, and I. Lee, ‘‘UAV-aided wireless communication several international conferences around the world. He received the Best
designs with propulsion energy limitations,’’ Jan. 2018, arXiv:1801.02782. Reviewer Award 2018 by ICT Expressed journal. Moreover, he received
[Online]. Available: https://arxiv.org/abs/1801.02782 the Education and Research Award 2018 given by Bangladesh Community
[26] S. Ahmed, ‘‘Robust resource allocation to secure physical layer using UAV- in South Korea. He served as a Reviewer for many international journals,
assisted mobile relay communications in 5G technology,’’ M.S. thesis,
including the IEEE, Elsevier, Springer, ScienceDirect, MDPI, and Hindawi
College of Eng, Utah State Univ., Logan, UT, USA, Aug. 2019. [Online].
published journals, and the IEEE conferences. He was he TPC Chair of the
Available: https://digitalcommons.usu.edu/etd/7575/
[27] Y. Yang and M. Pesavento, ‘‘A unified successive pseudoconvex approx- International Workshop on 5G/6G Mobile Communications, in 2017 and
imation framework,’’ IEEE Trans. Signal Process., vol. 65, no. 13, 2018. He was the Publicity Chair of the International Conference on Arti-
pp. 3313–3328, Jul. 2017. ficial Intelligence in Information and Communication, in 2019. He has been
[28] M. Chen, W. Saad, and C. Yin, ‘‘Liquid state machine learning for resource working as an Editor of ICT Express, an Associate Editor of IEEE ACCESS,
and cache management in LTE-U unmanned aerial vehicle (UAV) net- a Lead Guest Editor of Wireless Communications and Mobile Computing,
works,’’ IEEE Trans. Wireless Commun., vol. 18, no. 3, pp. 1504–1517, and a Guest Editor of Applied Sciences. He has been involved in several
Mar. 2019. Korean Government projects.
[29] R. I. Bor-Yaliniz, A. El-Keyi, and H. Yanikomeroglu, ‘‘Efficient 3-D
placement of an aerial base station in next generation cellular networks,’’ YEONG MIN JANG (Member, IEEE) received
in Proc. IEEE Int. Conf. Commun. (ICC), Kuala Lumpur, Malaysia, the B.E. and M.E. degrees in electronics engineer-
May 2016, pp. 1–5. ing from Kyungpook National University, South
[30] Q. Wu, Y. Zeng, and R. Zhang, ‘‘Joint trajectory and communication Korea, in 1985 and 1987, respectively, and the
design for UAV-enabled multiple access,’’ IEEE Trans. Wireless Commun., Ph.D. degree in computer science from the Univer-
vol. 17, no. 3, pp. 2109–2121, Mar. 2018. sity of Massachusetts, USA, in 1999. He was with
[31] M. Grant, S. Boyd, and Y. Ye. CVX: MATLAB Software for Disciplined the Electronics and Telecommunications Research
Convex Programming. Accessed: Nov. 31, 2019. [Online]. Available: Institute (ETRI), from 1987 to 2000. Since 2002,
http://www.stanford.edu/âĹijboyd/cvx/ he has been with the School of Electrical Engi-
[32] W. Dinkelbach, ‘‘On nonlinear fractional programming,’’ Manage. Sci., neering, Kookmin University, Seoul, South Korea,
vol. 13, no. 7, pp. 492–498, Mar. 1967. where he has also been the Director of the LED Convergence Research
[33] B. R. Marks and G. P. Wright, ‘‘A general inner approximation algo- Center, since 2010, and the Director of the Energy Internet Research Center,
rithm for non-convex mathematical programs,’’ Oper. Res., vol. 26, no. 2,
since 2018. His research interests include 5G/6G mobile communications,
pp. 681–683, Aug. 1978.
[34] S. T. L. Boyd and L. Vandenberghe, Convex Optimization Cambridge,
the Internet of Energy, AI platform, eHealth, public safety, optical wireless
U.K.: Cambridge Univ. Press, 2004. communications, optical camera communication, and the Internet of Things.
He is currently a Life Member of the Korean Institute of Communications
and Information Sciences (KICS). He received the Young Science Award
from the Korean Government, from 2003 to 2006. He was a recipient of the
KICS Dr. Irwin Jacobs Award, in 2017. He has organized several conferences
SHAKIL AHMED (Student Member, IEEE) and workshops, such as the International Conference on Ubiquitous and
received the B.S. degree in electrical and electronic Future Networks (ICUFN), the International Conference on ICT Conver-
engineering from the Khulna University of Engi- gence (ICTC), the International Conference on Information Networking
neering and Technology (KUET), Bangladesh, (ICOIN), the International Conference on Artificial Intelligence in Infor-
in 2014, and the M.S. degree in electrical engineer- mation and Communication (ICAIIC), and the International Workshop on
ing from Utah State University, Logan, UT, USA, Optical Wireless Communication Networks (OWCN). He had served as
in 2019. He is currently pursuing the Ph.D. degree the Founding Chair of the KICS Technical Committee on Communication
in electrical engineering with The University of Networks, in 2007 and 2008. He was a Visiting Professor with the Cornell
Arizona, Tucson, AZ, USA. He received the pres- University, Ithaca, NY, USA, in 2008 and 2009. He had served as the Execu-
tigious Presidential Doctoral Research Fellowship tive Director of KICS, from 2006 to 2014, the Vice President of KICS, from
Award by the school of graduate studies, Utah State University. He has 2014 to 2016, the Executive Vice President of KICS, in 2018, and the Pres-
published multiple research articles in international conferences and jour- ident of KICS, in 2019. He has been the Steering Chair of the MultiScreen
nals. One of his articles received the Best Paper Award at the international Service Forum, since 2011, and the Society Safety System Forum, since
conference. He served as a Reviewer for international journals, such as the 2015. He had served as the Chairman of the IEEE 802.15 Optical Camera
IEEE TRANSACTIONS ON COGNITIVE COMMUNICATIONS AND NETWORKING, IEEE Communications Study Group, in 2014, and also served as the Chairman of
ACCESS, the IEEE SYSTEMS JOURNAL, Wireless Communications and Mobile the IEEE 802.15.7m Optical Wireless Communications Task Group. He is
Computing, and the Applied Electromagnetics Society. His current research also the Chairman of the IEEE 802.15 Vehicular Assistive Technology (VAT)
interests include next-generation wireless communications, wireless network Interest Group. He is also serving as the Co-Editor-in-Chief of ICT Express,
design and optimization, unmanned aerial vehicle (UAV), physical layer which is published by Elsevier.
security, and covert/low probability of detection (LPD).
VOLUME 8, 2020 21225