IEEE INTERNET OF THINGS JOURNAL, VOL. 6, NO.
4, AUGUST 2019
Distributed and Multilayer UAV Networks for
Next-Generation Wireless Communication and
Power Transfer: A Feasibility Study
Yiming Huo , Member, IEEE, Xiaodai Dong , Senior Member, IEEE, Tao Lu, Member, IEEE,
Wei Xu , Senior Member, IEEE, and Marvin Yuen, Member, IEEE
Abstract—Unmanned aerial vehicles (UAVs) for wireless com-
munications have rapidly grown into a research hotspot as the
mass production of high-performance, low-cost, and intelligent
UAVs becomes practical. In the meantime, the fifth genera-
tion (5G) wireless communication and Internet-of-Things (IoT)
technologies are being standardized and planned for global
deployment. During this process, UAVs are becoming an impor-
tant part of 5G and IoT, and expected to play a crucial role
in enabling more functional diversity for wireless communica-
tions. In this paper, we first present a summary of mainstream
UAVs and their use in wireless communications. Then, we pro-
pose a hierarchical architecture of UAVs with multilayer and
distributed features to facilitate the integration of different
UAVs into the next-generation wireless communication networks.
Finally, we unveil the design tradeoffs with the consideration
of power transfer, wireless communication, and aerodynamic
principles. In particular, empirical models and published mea-
surement data are used to analyze power transfer efficiency, and
meteorological impacts on UAVs enabled next-generation wireless
communications.
Index Terms—Aerial base station (ABS), fifth generation (5G),
Internet-of-Things (IoT), laser power transfer, photovoltaic cell
(PV-cell), unmanned aerial vehicle (UAV).
I. I NTRODUCTION
N RECENT years, unmanned aerial vehicles (UAVs) have
I experienced a rapid transition from the initial military
exploitation, to fast-growing civilian applications, such as
industrial inspection, scientific research, agricultural practice,
security surveillance, emergency rescue, entertainment [2]–[5],
etc. In the meantime, the fifth generation (5G) and Internet-
of-Things (IoT) wireless networks and systems are being
planned for rapid deployment, and diverse paradigms to accel-
erate this progress and enrich application scenarios have been
Manuscript received March 9, 2019; revised April 12, 2019; accepted
April 29, 2019. Date of publication May 1, 2019; date of current version
July 31, 2019. This work was supported in part by the NSERC of Canada
under Grant 522620, and in part by the NSFC under Grant 61871109. This
work was presented in part in a preprint version [1]. (Corresponding author:
Xiaodai Dong.)
Y. Huo, X. Dong, and T. Lu are with the Department of Electrical and
Computer Engineering, University of Victoria, Victoria, BC V8P 5C2, Canada
(e-mail: ymhuo@uvic.ca; xdong@ece.uvic.ca; taolu@uvic.ca).
W. Xu is with the National Mobile Communications Research Laboratory,
Southeast University, Nanjing 210096, China (e-mail: wxu@seu.edu.cn).
M. Yuen was with the Viterbi School of Engineering, University of
Southern California, Los Angeles, CA 90089 USA. He is now with Kingston
Technology, Fountain Valley, CA 92708 USA. (e-mail: marvinyu@usc.edu).
Digital Object Identifier 10.1109/JIOT.2019.2914414
2327-4662 c 2019 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
7103
investigated. UAV-aided 5G wireless has sparked a large
interest as it can facilitate various use cases such as those
speculated in the three key principle application scenarios of
the International Telecommunication Union (ITU) [6]. They
are namely enhanced mobile broadband (eMBB), ultra reli-
able low latency communications (uRLLCs), and massive
machine type communications (mMTCs). For example, UAV
can play a critical role in providing network service recovery
in a disaster-stricken region, enhancing public safety networks,
or handling other emergency situations when uRLLC is
required. In particular, UAV-assisted eMBB could be an impor-
tant complement to the 5G cellular network where a 1000
times comprehensive performance improvement over 4G is
expected [6].
Since July 2016, when the Federal Communication
Committee (FCC) adopted a new Upper Microwave Flexible
Use Service [7], the use of millimeter wave (mmWave)
bands for cellular services has become an immediate real-
ity. However, alongside promising opportunities, e.g., larger
bandwidths and faster speed, mmWave cellular communi-
cations face significant challenges, particularly in terrestrial
environments where large propagation loss and shadowing
effects are ubiquitous [8]. The propagation loss challenge
can be overcome by adopting beamforming techniques at the
cost of more hardware resources and higher power consump-
tion [9]. Shadowing effects are more difficult to mitigate as
mmWave has an increased shadowing variance in nonline-
of-sight (NLoS) compared to sub-6 GHz [10]. Moreover,
mmWave channels hold sparse nature with limited channel
elements. Deploying UAV-assisted wireless networks can be
an effective solution to this issue as it enables more line-of-
sight (LoS) communications. A proof-of-concept (PoC) UAV
deign enabling multibeam multistream for multiuser applica-
tion scenarios is demonstrated in [11]. In the near future,
UAV-satellite communications [12] can enable more diverse
Earth-space communication and hereby make 5G global access
robust and reliable.
In general, UAV-aided wireless communications can fall
into three representative categories of use cases [13], namely,
UAV-aided ubiquitous coverage, relaying, and information dis-
semination and data collection. The former two categories are
suitable for 5G base station (BS) offloading and wireless con-
nectivity relaying. However, there are several major hurdles
preventing UAVs from 5G networks. First, UAVs could cause
7104
safety concerns [16]. For this reason, the Federal Aviation
Administration (FAA) and many other countries’ civil avia-
tion authorities have implemented laws and rules on operating
commercial UAVs [17] to regulate the weight, the maximum
altitude and speed, the minimum distance from airports, con-
structions, vehicles, people, etc. Second, the UAV design is
confronted with various technical bottlenecks. One of the
most significant limitations is on-board power. For example,
a 3-pound mainstream miniature UAV (mini-UAV) usually
carries single 1-pound lithium polymer (LiPo) battery. Such
battery typically has a total energy of more than 80 Wh and can
barely support a maximum flight time of more than 30 min.
Except for control and nonpayload communication (CNPC)
system that should have outstanding performance in latency
and security [16], adding and enabling 5G wireless communi-
cation functions into UAVs will necessitate additional payload
and power-hungry wireless hardware that further limits the
operation duration. Moreover, adverse weather conditions can
pose serious challenges to UAV operating time and wireless
communication quality.
Generally, aircraft engine consumes much higher energy
than the communication systems of a microcell or picocell.
In the literature, several solutions have been proposed to
enhance robustness, flight time, and energy efficiency of a
UAV communications system.
One solution is derived from the wireless system design
perspective. On the one hand, Alzenad et al. [18] presented
algorithms to maximize the number of users covered by find-
ing energy-efficient 3-D placement of a UAV-BS within the
region of interest. On the other hand, UAV path planning
plays a crucial role in attaining satisfactory energy efficiency
and quality of service (QoS), and the planned paths depend
on the specific application scenarios. As an example, [19]
optimized UAV’s flying direction for uplink communications
by assuming a constant speed that the UAV maintains. In
addition, the UAV trajectory optimization in [20] took into
account the propulsion energy consumption of fixed-wing
UAVs (FW-UAVs).
Another approach is to directly advance the energy resource
technology and improve the energy management. In general,
both gasoline and jet fuel are at least 20 times higher in
energy density or specific energy (MJ/kg) than the LiPo battery
widely used for mini-UAVs. In spite of the facts that the elec-
tric motor generally demonstrates higher efficiency and speed
adjustment capability than the petrol engines, the energy den-
sity gap cannot be filled easily. As predicted, petrol engines
or hybrid-electric engines should play a critical role in future
UAVs used in 5G communications to achieve longer flight
duration.
New energy systems and energy harvesting techniques could
further accelerate the pace of implementing 5G-oriented UAV
networks. For example, [21] presents a solar power man-
agement system for aircraft and UAV applications, with a
maximum power tracking system to increase the operating
efficiency of solar cells. Moreover, some other techniques
for addressing energy transfer and storage challenges, such
as laser power beam techniques can also be integrated to
enable 24-h flight working without landing or refueling [22].
IEEE INTERNET OF THINGS JOURNAL, VOL. 6, NO. 4, AUGUST 2019
Ouyang et al. [23] have proposed a throughput maximization
scheme for balancing tradeoffs between laser energy harvest-
ing and wireless communication performance.
In order to overcome aforementioned challenges, some
UAVs networking architectures are proposed. For exam-
ple, [14] illustrates a multitier drone-cell network based on
which a novel drone management framework for efficient oper-
ation is proposed. In [15], the feasibility of a multitier drone
architecture over single-tier drones is studied in terms of spec-
tral efficiency, with the impact of different urban environments.
Single drone in [14] and multiple drones in [15] are used to
support cells of different sizes, creating multitier cells.
The main contributions of this paper are summarized as the
following.
1) Investigation of various UAVs genres from both aero-
dynamics and wireless communication aspects. Based
on the unique features of each UAV type, we pro-
pose a multilayer and distributed networking architec-
ture facilitating next-generation UAV enabled wireless
communication.
2) Feasibility analysis and comparison of radio
frequency/microwave-based wireless power transfer and
laser power transfer for the proposed architecture.
3) Study of UAV communications channels under various
weather conditions.
The rest of this paper is organized as follows. Section II
presents several main types of UAVs, their aerodynamic char-
acteristics, and suitability to 5G applications. Section III
proposes a novel distributed and multilayer UAVs (DAMU)
network, where the aerial network is composed of multiple
layers that interact with each other. Section IV studies
the critical power supply issue, comparing laser-based and
microwave-based power transfer within the DAMU network.
Section V characterizes the impact of meteorological condi-
tions on the UAV communication channel attenuation property.
Finally, Section VI concludes this paper with suggested future
works.
II. W HEN A ERODYNAMICS M EETS
5G C OMMUNICATIONS
In this section, we investigate different types of UAVs from
the aerodynamics perspective and their suitability for 5G com-
munications. As depicted in Fig. 1, there are three categories
of UAVs: 1) balloon; 2) fixed-wing; and 3) rotary-wing. A
detailed comparison of these UAVs is summarized in Table I.
Among them, balloons have been widely used for greater
than 10 km in altitude, high altitude platforms (HAPs), and
ultrahigh altitude (UHA) applications. For example, NASA’s
helium balloons can lift instruments upto hundreds of kilo-
grams at an altitude over 30 km for more than 100 days. On
the other hand, Google’s Project Loon [24] has successfully
enabled a balloon network at an altitude of 20 km, extending
the Internet connectivity in rural and remote areas worldwide.
In October 2017, Project Loon provided emergency long-term
evolution (LTE) service recovery to Puerto Rico in the after-
math of Hurricane Maria. With solar panels and advanced
predictive models of the winds and other meteorological
information from National Oceanographic and Atmospheric
HUO et al.: DAMU NETWORKS FOR NEXT-GENERATION WIRELESS COMMUNICATION AND POWER TRANSFER
TABLE I
C OMPARISON OF T HREE C ATEGORIES OF M AINSTREAM UAV S
(b)
(a) (c)
Fig. 1. Three main categories of UAVs. (a) Balloon. (b) Fixed-wing.
(c) Rotary-wing (quadcopter).
Administration (NOAA), the balloons can be navigated and
deployed as requested. Furthermore, a balloon-UAV can facili-
tate quasi-stationary communications with propellers adjusting
balance and position.
The second type of UAV is the FW-UAV. In 1995, MQ-1
Predator was first introduced by General Atomics. An FW-
UAV can achieve a wide range of altitude with the fastest
horizontal speed due to powerful turbine engines. The max-
imum payload weight is determined by the lift force (L)
according to [25]
CL × ρ × V 2 × A
L= (1)
2 the sky with the longest duration. Considering the significant
where L is the lift force that equals to the airplane’s weight in
pounds; CL is the coefficient of lift determined by the airfoil
type and angle of attack (AOA); ρ is the air density and its
specific value can be obtained from the atmospheric model of
the International Standard Atmosphere (ISA); V stands for the
velocity of the airfoil; and A is the surface area of the wing. For
a medium-sized FW-UAV with 11 m2 , effective wing area fly-
ing with an AOA of 15◦ , at 5-km high and 50 m/s (180 km/h),
7105
it can generate a lift force of 11 800 Newtons. On the one hand,
in 5G applications, an FW-UAV needs to maintain a minimum
speed to carry the weight of both equipment and UAV itself.
On the other hand, it shall fly as slow as possible to minimize
the Doppler effect and avoid complicating channel modeling
and system design. Moreover, hybrid-electric engines and solar
panels can further improve the energy efficiency and flight
time for cost-effective UAV enabled 5G services. The solar-
powered aircraft, Solar Impulse, is built with electric motors,
lithium-ion (Li-ion) batteries, and solar panels, and can 118 h
of flights without landing.
The third type of UAV is known as the rotary-wing UAV
(RW-UAV) and has been popularly deployed as an LiPo bat-
tery powered quadcopter in Fig. 1(c). Such an RW-UAV can
achieve high aerodynamic flexibility and mobility with reli-
able hovering capability. In the UAV-based delivery system
developed by Amazon Prime Air [26], RW-UAV with multiple
propellers is demonstrated. However, the battery signifi-
cantly limits its flight time and payload. Consequently, the
petrol engine-based RW-UAV equipped with 6 or 8 pro-
pellers for industrial applications are developed for longer
flight times (generally 3–5 times), better balance, and pay-
load. Nevertheless, the overall energy efficiency of RW-UAV
is about 80% lower than FW-UAV. Its application for 5G
is limited due to unsatisfactory flight time and weight
capacity.
III. D ISTRIBUTED AND M ULTILAYER UAV N ETWORKS
As aforementioned, a balloon-UAV may be the most suit-
able aircraft for carrying heavy 5G equipment and hovering in
coverage for more than 20 km in radius, an energy-efficient
balloon-UAV can serve as a 5G macrocell BS that could weigh
up to hundreds of kilograms. A high altitude in the sky can
enable higher probability of LoS links [8]. Furthermore, an
FW-UAV may carry a 5G light macrocell or microcell and
fly within a flexible altitude below 10 km. Finally, an RW-
UAV is ideal for installing lightweight 5G small cell/picocell
equipment and executing short duration tasks that require fast
7106
Fig. 2. DAMUs network architecture.
deployment. Based on the characteristics of different UAVs,
we propose a DAMU network architecture as depicted in
Fig. 2.
1) The balloon-UAVs function as quasi-stationary cellu-
lar towers and stay in the stratosphere, 20 km above
sea level. Typically, a balloon can be recycled and
relaunched before or after a regular maintenance over
a period of several months.
2) The FW-UAVs are deployed below 10 km and above
1 km. In order to minimize the Doppler shift and the
associated system design challenges, FW-UAVs need to
cruise at the slowest speed possible.
3) The RW-UAVs are dispatched below 1 km, serving
as microcell/picocell/small cell BSs. Low cruise alti-
tudes can enable RW-UAVs to be frequently and quickly
recharged or replenished.
Fig. 3 illustrates an example of DAMU network that realizes
5G communications, energy harvesting, and power transfer
among its constituent UAVs. DAMU enables close interactions
among the three layers and dynamic configurations depending
on the application requirements.
1) 5G macrocell communications and CNPC systems are
integrated and enabled in the balloon-UAVs. The bal-
loons provide both 5G new radio (NR) facility and back-
wards compatibility to legacy 3GPP standards. In addi-
tion, the balloons communicate with other aerial BSs
(ABSs) or ground BSs (GBSs), as well as ground termi-
nals. In a typical LoS communication scenario in Fig. 3,
Balloon 1 operates its phased arrays to form multiple
beams to align with the beams from FW-UAV-1 and FW-
UAV-2, respectively. On the other hand, Balloon 1 is also
connected with other Balloons such as Balloon 2/3, and
IEEE INTERNET OF THINGS JOURNAL, VOL. 6, NO. 4, AUGUST 2019
other unmanned spacecraft/satellites (not drawn in the
figure) to form wider coverage networks and additional
hierarchical links. These balloons can act as backhauls
for UAVs of lower layers.
2) FW-UAVs serve as either a macrocell or a microcell.
In a representative 5G usage scenario, an FW-UAV
may communicate with both a balloon-UAV macrocell
and multiple RW-UAVs-based small cells/picocells. It
also can communicate with GBSs and other ground
terminals. As depicted in Fig. 3, RW-UAV-5 is live-
streaming a sport event over the stadium and uploading
the ultrahigh resolution (UHD) video to RW-UAV-
3 that further communicates with FW-UAV-1 using
beamforming.
3) RW-UAVs are dispatched and deployed mainly for
microcell/picocell applications below an altitude of
1 km. They can enable fast communication access and
services whenever or wherever there is such a need.
An RW-UAV can be connected to either ABSs/GBSs
or aerial/ground user terminals. Moreover, RW-UAVs
can also execute multiple tasks such as searching
and surveillance under emergency situations. Either an
RW-UAV or FW-UAV can generate multiple mmWave
beams to increase spatial multiplexing gain and chan-
nel capacity, and mitigate the interference. For exam-
ple, by adopting a distributed phased array MIMO
(DPA-MIMO) reconfigurable architecture in [9], [11],
and [27] where multiple phased arrays are placed in
a distributed manner on an RW-UAV, multigiga-bits
speed communication for multiuser scenarios can be
enabled and has been preliminarily verified from field
tests.
4) Instead of carrying full BS equipment, some UAVs of
each layer can also act as distributed antennae or remote
radio heads for a GBS or ABS cell. This is particu-
larly applicable to RW-UAVs whose payload and power
are limited. Moreover, in certain application scenarios,
a UAV may function as a relay.
5) Under specific emergency conditions, each vertical layer
of UAVs should be able to work independently to sustain
a full-function network access and service when other
layers of UAVs are not available.
6) When an FW-UAV-based ABS enables 5G communi-
cations, its cruising velocity should be well maintained
at the lowest possible speed, and its flight path needs
to be well planned. However, as a tradeoff according
to (1), the velocity will need to increase when carry-
ing more payload (5G and beyond equipment) or/and
flying at a higher altitude. Furthermore, energy effi-
ciency should be taken into account when determining
the cruising speed. Moreover, assume that an FW-UAV
is needed to provide 5G service coverage for area of
interest with a radius of 2 km, we need to dispatch it
to fly at least 2-km high [28]. As an FW-UAV cannot
hover at a fixed location, it should be programmed to
fly with a predefined pattern that is known and easy
to track by other ABSs/GBSs and terminals, in order
to build reliable communication links. For example, the
HUO et al.: DAMU NETWORKS FOR NEXT-GENERATION WIRELESS COMMUNICATION AND POWER TRANSFER
Fig. 3. Wireless communications and power transfer in a DAMU enabled 5G wireless communication and power transfer network.
cruising path can be a simple circle with a small radius
of turn given by [25]
V2
R=
11.26 × tan(θ )
where R is in the unit of feet, V is the velocity in
the unit of knot, and θ is the bank angle. Accordingly,
a medium-sized FW-UAV (>100 kg of extra payload)
should maintain a cruising speed at 10 m/s (19.4 knot).
In Fig. 4, we assume an FW-UAV flies at an altitude of
2 km and a speed of 20 m/s. For simplicity, we exclude
the wind effect in this stage. If the bank angle is 30◦ ,
the radius of turn, R, should be 70.8 m. Furthermore,
assume a balloon-UAV that hovers 20-km high and 5 km
horizontally away from the center of circle, while a
quasi-stationary RW-UAV is 1.9 km vertically beneath
the center of circle, both communicating with the FW-
UAV. Angles α and β that are formed by the LoS
communication and indicated in Fig. 4, are calculated
to be 0.42◦ and 2.13◦ , respectively. They affect beam
7107
alignment and beam tracking when mmWave beamform-
ing is used in balloon-FW and FW-RW links. Smaller
α and β ease beam tracking, particularly when they are
(2) smaller than or close to half of the half power beamwidth
(HPBW) of the radiation patterns, as illustrated in
Fig. 4.
7) The trajectory design or path planning is another major
concern for the proposed DAMU networks. When UAVs
serve as ABSs, the trajectory design is focused on
navigation with practical constraints, such as collision
avoidance [29] and dedication to the optimization of
the communication performance. For example, succes-
sive convex optimization technique has been used for
the energy efficiency maximization of UAV-enabled
communication [20] and throughput maximization for
UAV-enabled multiuser downlink communication [30].
For DAMU networks in particular, trajectory design
becomes more complicated than prior works and needs
to consider cooperating different types of UAVs and
energy harvesting/power transfer altogether.
7108
Fig. 4. FW-UAV circular cruising mode in a DAMU system.
8) Secure communication of UAVs in DAMU networks
is a critical factor to the success of the architec-
ture. Due to the broadcast nature of the medium,
wireless communications are vulnerable to malicious
attacks or eavesdropping [31]. Compared to conven-
tional key-based cryptographic techniques that require
secret keys and complex algorithms, physical layer
security (PLS) is more suitable for large-scale decen-
tralized wireless networks [32] such as DAMU. In light
of the complexity of DAMU architecture, hierarchi-
cal network management is needed. Moreover, multiple
techniques may be adopted to co-enable secure UAVs
communication and maximize the secrecy rate. For
example, Li et al. [33] focused on maximizing the
secrecy rate by jointly designing the transmit beam-
former and artificial noise (AN) subject to the transmit
power constraint of UAV; [31] pointed out that aerial
blockchain technique can be considered as a decentral-
ized solution to prevent privacy leakage; [34] enhanced
security performance through controlling channel gains
via adjusting the UAV trajectory and employing the
conventional power/rate adaptation.
IV. P OWER T RANSFER IN DAMU N ETWORKS
The power transfer strategy based on the DAMU architec-
ture is critical and may encompass multiple energy harvesting
and power transfer technologies to realize cost-effective energy
charging to the UAVs. Take DJI’s Matrice 100 for example,
IEEE INTERNET OF THINGS JOURNAL, VOL. 6, NO. 4, AUGUST 2019
Fig. 5. Generic block diagram of wireless power transfer via RF/microwave.
with a maximum takeoff weight of 3600 g and a TB48D bat-
tery, it can hover with 500-g payload at an operation power of
380 W under this mode [35]. Therefore, the minimum transfer
and/or harvest power of the same amount is required in order
to keep the UAV hovering without landing.
Natural energy harvesting is currently the most suitable
technique on UAVs. Fig. 3 illustrates that UAVs can har-
vest both solar and wind energy. The balloon-UAVs can be
equipped with solar panels, batteries, and wind generators
to achieve a self-sustaining energy system. Staying in strato-
sphere layer facilitates efficient solar energy harvesting as no
weather occurs at this layer.
Near-field wireless power transfer, such as inductive cou-
pling and magnetic resonance coupling have seen commercial
successes in recent years, but cannot be satisfactorily applied
to power transfer and energy charging in DAMU networks
due to their limited distance typically ranging from a few cen-
timeters to a few meters [36]. For long range far-field wireless
power transfer, WPT via RF/microwave is one of the candidate
technologies suitable for DAMU networks.
A. Wireless Power Transfer via RF/Microwave
A block diagram of WPT via RF system typically oper-
ating in industrial, scientific, and medical (ISM) bands, such
as 915 MHz, 2.45 GHz, and 5.8 GHz, is given in [37] and
redrawn in Fig. 5. The end-to-end power transfer efficiency
can be expressed as [37]
Prdc Ptrf Prrf Prdc
ηrf = = (3)
Ptdc Ptdc Ptrf Prrf
  
η1 η2 η3
where η1 , η2 , and η3 stand for the DC-to-RF, RF-to-RF, and
RF-to-DC power conversion/transmission efficiency. Ptrf and
Prrf are the RF power at the transmitter and receiver. Ptdc
and Prdc are the dc power at the transmitter and receiver,
respectively. In particular, η1 is mainly determined by, the
RF up-converter efficiency, the power amplifier (PA) effi-
ciency indexed by power added efficiency (PAE), and the
antenna radiation efficiency (RE), respectively. η1 is further
formulated as
out − Pin Pant
Prf rf rad
η1 = ηup-conv (4)
tot
Pdc Pin
   
ant
PAE RE
where Prf rf tot
out , Pin , and Pdc stand for the PA’s output, input,
and total power consumption. Parameters Pin rad
ant and Pant rep-
resent the input and radiated power of the antenna. Under
the condition that the conduction losses and dielectric losses
are minimized, and impedance matching is achieved between
the PA and antenna(s)’s feed port(s), Pin ant can be deemed
HUO et al.: DAMU NETWORKS FOR NEXT-GENERATION WIRELESS COMMUNICATION AND POWER TRANSFER
out . Moreover, ηup-conv depends on the specific cir-
equal to Prf
cuits and systems design of up-converting the dc signal to an
RF signal.
Furthermore, η2 is mainly determined by wireless channels
and can be quantitatively analyzed through link budget calcu-
lations. At the RF-to-DC conversion end, a rectenna is usually
used to receive an RF signal and transforms it into dc power. It
consists of an antenna and a rectifying circuit, which decides
η3 with a suitable dynamic range at the frequency of interest.
Matsunaga et al. [38] presented a highly efficient (44.1% with
0.041 W/m2 input) stacked differential rectenna working at
5.8 GHz; [39] demonstrates a maximum RF-to-DC conver-
sion efficiency of 16.3%–45.3%, with wide angle coverage
at 2.45 GHz. Usually, it is more challenging to rectify and
convert lower power than high power in the harvester.
Although WPT via RF/microwave is a promising technol-
ogy that can be applied to DAMU networks under some appli-
cation scenarios, there are several known technical limitations
summarized below.
1) High free space path loss (FSPL). FSPL can be approx-
imated as [40]
(4π d)2
FSPL = (5)
Gt Gr λ2
where d is the distance between the transmit and receive
antennas, λ is the wavelength of the carrier, and, Gt and
Gr are the gains of transmit and receive antennas. The
FSPL is quadratically proportional to the distance, and
it becomes so significant that over 1-km distance omni-
directional antennae experience about 100-dB FSPL at
carrier frequencies. Such large loss can substantially
degrade the overall efficiency ηrf whereas the channel
models of specific application scenarios are sometimes
more complicated.
2) Using phased arrays can increase the gain and effective
isotropic radiated power (EIRP) of the propagation path
and mitigate FSPL. However, such approach is imprac-
tically expensive, e.g., an array of 1000 antennas can
boost the gain by 30 dB. This compensation is still
incomparable to the large FSPL, overly large in physical
dimension, expensive in hardware, and extremely high
in power consumption. Moreover, the RF amplifiers with
output more than 50 dBm (100 W) are usually designed
and implemented using high-performance and expen-
sive monolithic microwave integrated circuit (MMIC)
process, such as gallium nitride (GaN). High-power
PAs satisfying the massive power transfer requirements
(hundreds of watts and above, plus good linearity) of
DAMU networks are still rare and challenging to design
according to state-of-the-art designs [41], [42].
3) Since the sub-6 GHz spectrum is overly crowded with
unlicensed and licensed commercial bands for cellular
and WiFi, high-power transfer that takes advantage of
the ISM bands can introduce considerable interference
to wireless data transmission that is normally in much
lower signal strength. As a result, the isolation and
linearity of RF circuits are of critical importance.
7109
Fig. 6. Generic block diagram of OWPT via laser.
B. Laser Power Transfer
To overcome limitations, optical wireless power transfer
(OWPT) via laser may provide a complementary solution due
to the following advanced features.
1) Highly concentrated energy in the laser beam with well
controlled divergence angle as small as 0.1 microradians.
2) Enabling long-range point-to-point power transfer with
much less attenuation.
3) Photovoltaic (PV) receivers are efficient for laser charg-
ing [43], e.g., gallium arsenide (GaAs)-based photo-
voltaic cells (PV-cells) can achieve 50% efficiency [44].
A generic block diagram of the OWPT via laser system
is illustrated in Fig. 6. Ps and Pl are the input and output
power of the optical transmitter which transforms the electric-
ity to laser, Pr is the power after being transmitted through
the optical channel, and Pout is the output power of the opti-
cal receiver (PV-cell) that converts the laser to electricity. The
overall system efficiency from input at transmitter to output at
receiver is written as
Pl Pr Pout
ηl = (6)
Ps Pl Pr
  
ηs ηt ηout
where ηs is the conversion efficiency of electricity to laser,
ηt is the efficiency of laser transmission over the open space,
ηout is the conversion efficiency of laser to electricity, and
Pout = Vout Iout . Modeling laser transmission in the space and
atmosphere is significantly different from its RF/microwave
counterpart which can be partly characterized with Friis (5).
Based on modeling and formulation in [45], laser transmission
efficiency is defined as
ηt = e−σ d (7)
 −q
3.91 λ
σ = (8)
V 550 nm
where d is the transmission distance between the optical trans-
mitter and receiver, V stands for visibility in km, λ is laser
wavelength, σ is atmospheric attenuation coefficient in the unit
of dB/km, and q characterizes the size distribution of the scat-
tering particles as they can be composed of both Rayleigh and
Mie type ones. Typically, q = 1.3 for 6 km < V < 50 km,
q = 0.16 V + 0.34 for haze visibility (1 km < V < 6 km),
q = V − 0.5 for mist visibility (0.5 km < V < 1 km), and
q = 0 for fog visibility (V < 0.5 km).
For laser transmission, spot diameter is another important
parameter since it directly impacts the optical (laser) receiver
design and laser beam tracking and alignment. According
to [44], the minimum spot diameter is diffraction limited by
Dspot = 2.44dλ/Dlens (9)
where d is the distance between the transmitter and receiver,
λ is the laser wavelength, and Dlens is the lens diameter. For
7110 IEEE INTERNET OF THINGS JOURNAL, VOL. 6, NO. 4, AUGUST 2019

Fig. 7. Block diagram of PV-cell model.

example, with a 1-m diameter lens, after propagating 10 km,

the minimum spot size is only 3.8 cm.
Regarding ηs for electricity-to-laser conversion, we can fur-
ther expand and formulate it using the relationship in [46] Fig. 8. Maximum laser transmission efficiency of 1550 and 810 nm, under
given by temperature of 0◦ , 25◦ , and 50◦ , respectively.

Pl q (It − Ith )
ξ hν
ηs = = e (10)
Ps It Vt
where It and Vt are the driving current and voltage, respec-
tively, Ith is the laser threshold current, h denotes the Plank
constant which equals to 6.62606957×10−34 J·s, ν is the laser
frequency, qe is the electronic charge constant and equals to
1.6 × 10−19 C, and ξ is an empirical coefficient with value
between 0 and 1 that is determined based on measurements
and given in [47].
For the laser-to-electricity conversion efficiency, ηout , it is
directly determined by the converting capability of a PV-cell
which can be modeled [49] as follows. As illustrated in Fig. 7,
the equivalent electrical circuit of PV cell can be represented
by using the lumped model that includes a light-induced cur-
rent source, a parallel diode, a shunt resistor, and a series
resistor. Thus, Iout is derived as
Fig. 9. Maximum laser transmission efficiency for haze with 2 and 4 km
Iout = Ig − Id − Ish (11) visibility, respectively.
 v +I R  
out out sr vout + I out Rsr
Iout = Ig − Is e nkB Tc /qe − 1 − (12)
Rsh 5% after 20 m propagation. In addition, according to [45], very
where Ig is the photograph current related to the optical radi- heavy snow and rain can result in 60–100 dB/km attenuation.
ation and PV-cell temperature Tc , Is is the dark saturation Subsequently, using laser power transfer technology is, to a
current which is also a function of cell temperature, kB is large extent, weather and distance dependent.
Boltzmann constant, and n is the ideality factor of PV-cell Nevertheless, compared to its RF/microwave counterpart,
with the value 1 for an ideal diode. advantages of laser power transfer are very obvious and is
Fig. 8 depicts the maximum transmission efficiency, ηt , of quantitatively analyzed in Fig. 12. The transmission efficiency
1550 and 810 nm laser versus distance for clear air with 10 km of WPT-RF at 915 MHz/2.45 GHz/5.80 GHz and laser power
visibility, under different temperatures of 0◦ , 25◦ , and 50◦ . transfer at 1550 and 810 nm are plotted with respect to distance
As observed, lower temperatures help achieve a better overall ranging from 10 m to 20 km. In particular, η1 and η3 are
transmission efficiency while 1550 nm outperforms 810 nm assumed to be 0.5 and 0.9 for WPT, based on the state-of-the-
for longer distance transmission. Laser transmission in haze art designs. Both omni-directional antenna (antenna gain = 1)
with 2 and 4 km visibility are further plotted in Fig. 9. For all and phased array are considered for comparison. Both a 100-
cases in this figure, a transmission efficiency of at least 10% element phased array (20 dB additional gain for the link) and
is maintained for a distance of no further than 500 m. a 1000-element phased-array (30 dB additional gain) are taken
Likewise, the mist/fog scenarios with visibility of 1 km, into account, to enhance the transmission efficiency.
200 m, and 50 m are analyzed in Figs. 10 and 11. It is apparent As observed, in general, WPT-RF is largely outperformed
that even the medium fog with 200 m visibility can degrade the by the LPT, even with a very large phased array (high in
laser power transfer efficiency quite substantially while a thick dimension, cost, and power consumption) at the transmit-
fog with 50 m visibility reduces the efficiency quickly below ter to compensate the quick efficiency drop. Consequently,
HUO et al.: DAMU NETWORKS FOR NEXT-GENERATION WIRELESS COMMUNICATION AND POWER TRANSFER
Fig. 10. Maximum laser transmission efficiency for mist with visibility
of 1 km.
Fig. 11. Maximum power transmission efficiency over distance for thick and
medium fog with visibility of 50 and 200 m, respectively.
it is hardly possible to rely only on WPT-RF to realize
hovering-without-landing functions of RW-UAVs while LPT
can conditionally beam highly concentrated energy (hundreds
of Whs) over long distance without significant attenuation.
WPT-RF can be used in scenarios where transmission dis-
tance is not very large. When good visibility is available for
LPT, 5% transmission efficiency can be maintained even after
10 km propagation. By applying LPT technology to DAMU
networks, RW UAV-1, RW-UAV-8, and RW UAV-3 are power-
supplied by the laser beam generators installed on the roof of
high-rise building and stadium, for example. Balloon 1 may
also transfers laser beams to fixed-wing UAV-1/UAV-2 and
RW UAV-6. FW-UAVs may also provide laser power transfer
to other RW UAVs.
Furthermore, delivering high-power laser beams over high-
performance optical fibers directly to the RW UAVs is
currently a more practical solution to the atmospheric and
weather-introduced attenuation of the wireless LPT, and has
been demonstrated by [22].
7111
Fig. 12. Transmission efficiency comparison of wireless power transfer via
RF and laser power transfer.
One major concern of LPT in the DAMU network is the
safe laser power transmission since the transferred power can
be up to hundreds of watts. Laser safety limits are based on
maximum permissible exposure (MPE) values which define
the exposure time below which a laser power density has neg-
ligible effect on the human eye [48]. On the one hand, laser
power transfer over fiber (LPT-OF) as aforementioned is cur-
rently an available and affordable option to implement safe
laser power transfer. On the other hand, the wireless laser
power transfer channels should strictly depend on the LoS.
Consequently, laser tracking and LoS detection is extremely
critical. For the existing conceptual designs and prototypes, the
distributed laser charging technology [47], [50] can ensure the
immediate stop of laser power transfer when LoS is blocked by
any object. In addition, [51] demonstrates a laser beam charg-
ing prototype that enables shutting down the high–power laser
before a human moving with a maximum speed of 44 m/s.
This design implements low power lasers to create a guard
ring around the high-power laser and direct them at the retro-
reflectors on the receiver, which enables a motion detection
system that operates on the order of nanoseconds for 10-m
distance. Moreover, other living creatures such as animals and
plants should be well protected, and environmental damages
are supposed to be strictly controlled. Note that the high trans-
mission efficiency of LPT means the power transfer duration
can be relatively short which is desirable.
Some other alternative power supply methods are also crit-
ical complements for power transfer in DAMU networks. For
example, RW-UAV-3 is wired through an electrical power
cable to a power source, and the cable length can be more
than 100 m according to experimental practice.
V. UAV C OMMUNICATION C HANNELS U NDER
M ETEOROLOGICAL C ONDITIONS
For UAV communications, air-to-ground (A2G) and air-to-
air (A2A) channel characteristics are significantly different
7112
from terrestrial ones. The currently proposed channel models
have been thoroughly reviewed in [8] and [52]. According to
the level of blockage, there are mainly three classes of links:
1) LoS; 2) obstructed LoS (OLoS); and 3) NLoS. The proba-
bility of LoS occurrence (LoS, OLoS, or NLoS) is a function
of environments. Following the UAV channel model [53],
the LoS probability under four selected environments, includ-
ing suburban, urban, dense urban, and high-rise urban, can
be approximated to a simple modified Sigmoid function
given by
1 γc = Kl M
PLoS = (13)
1 + a exp(−b[θ − a])
where a and b are S-curve parameters directly linked to the
environment variables specified in [54], θ is the elevation angle
which equals to arctan(h/r), and h and r are the height and
coverage radius of UAV, respectively. Moreover, the path loss
under LoS and NLoS scenarios can be formulated as
(4π d)2
PLLoS = 10 log + ηLoS (14)
Gt Gr λ2
(4π d)2
PLNLoS = 10 log + ηNLoS (15)
Gt Gr λ2
where the first term of both equations is the FSPL, ηLoS and
ηLoS are
√ mean values of excessive path loss given in [55], and
d = h2 + r2 . Therefore, the probabilistic mean path loss
values can be written as
PL(h, r) = PLoS × PLLoS + PNLoS × PLNLoS . (16)
Particularly, A2G and A2A channels have smaller path loss
exponent and lighter small-scale fading than terrestrial ones.
Moreover, the A2A channel model plays a crucial role in
the DAMU networks as UAV-to-UAV communication exists.
However, either conventional ground-to-ground (G2G) chan-
nels or aforementioned A2G channels are difficult to extend
to the A2A channel [57] in 3-D space. One of the very
few works that characterizes the A2A channels [57] proposes
an empirical propagation channel prediction model, based on
the ray-tracing. In order to extract more realistic A2A chan-
nel model parameters, UAV channel measurements should be
conducted.
As previously discussed, weather conditions play a critical
role in DAMU enabled 5G networks with apparent impacts
on wireless communications, power transfer, and UAV work-
ing status. It should be noted that global weather and climate
patterns are dramatically diverse. Therefore, in this section,
we focus on A2A attenuation modeling for frequency below
100 GHz with weather factors taken into account.
First, atmospheric humidity largely affects gaseous attenu-
ation, particularly at mmWave bands and above. The quan-
titative analysis of gaseous attenuation over frequency as a
variable of water vapor density, is given in prediction mod-
els recommended by ITU [58]. Generally, gaseous attenuation
due to water vapor increases over frequency. At sea level and
under standard atmosphere (7.5 g/m3 water vapor density), the
total gaseous attenuation (dry air plus water vapor) grows to
more than 1 dB/km from 53 to 67 GHz and sharply peaks at
IEEE INTERNET OF THINGS JOURNAL, VOL. 6, NO. 4, AUGUST 2019
15 dB/km for 61 GHz. This is attributed, to a large extent, by
oxygen absorption.
From the meteorological perspective, precipitation can be
in the forms of drizzle, rain, sleet, snow, and hail. According
to ITU rain attenuation models, heavy rainfall can cause sig-
nificant attenuation at 5-G mmWave bands [56]. When the
weather gets cloudy and foggy, there are two methodologies
to calculate attenuation [58]. The first one is to obtain the spe-
cific attenuation (dB/km) within a cloud or fog, which can be
written as
(17)
where Kl is the specific attenuation coefficient
((dB/km)/(g/m3 )), and it is a function of frequency and
permittivity of water; M is the liquid water density (LWD)
in cloud or fog (g/m3 ). For medium and thick fog, the LWD
is around 0.05 and 0.5 g/m3 , respectively [59]. Additionally,
advection fog can be more than 2-km thick above ground
level.
The second methodology is to calculate attenuation due to
clouds for a given probability. This attenuation is correlated
with the statistics of the total columnar content of cloud liquid
water L (kg/m2 ) for a given geographic location, and it is
expressed as
LKl
A= (18)
sin(θ )
where Kl is the specific attenuation coefficient and θ is the
elevation angle within a range from 5◦ to 90◦ . The value L
can be obtained from map-based data files [58]. For exam-
ple in some regions of Southeast Asia, L can be as high
as 2 for a yearly exceedance probability of 1%. However,
the second method is not able to directly estimate the worst-
case A2A communications scenarios, especially before heavy
rainfall.
Out of many forms and types of clouds in the Earth’s
atmosphere, cumulonimbus clouds are a dense, towering ver-
tical clouds that can be very tall and thick with the highest
LWD. According to cloud thickness estimation from GOES-
8 satellite data [59], precipitating cumulus clouds (precursor
of cumulonimbus cloud) have a mean thickness of 9.32 km.
Using these variables, we can plot the atmospheric attenua-
tion over frequency, with various meteorological conditions
introduced.
Assuming the LoS exists between a balloon-UAV (at 20 km)
and FW-UAV (hovering at 1 km) and is normal to the ground
plane, and such an LoS channel normally exists during a clear
day with good visibility. As depicted in Fig. 13, if a very thick
cumulonimbus cloud with high LWD (12 km, 3 g/m3 ) exists,
it causes the highest attenuation for frequencies over 40 GHz.
Moreover, if a 2-km vertically thick advection fog emerges,
the resulting attenuation is 0.68 and 1.28 dB at 28 GHz and
40 GHz, respectively. If the precipitation happens, the atten-
uation caused by medium rain, heavy rain, and violent rain
dramatically varies. For example, a violent rain (100 mm/h)
can result in a 38.3-dB attenuation at 40 GHz, compared to a
3.4-dB attenuation caused by medium rain. Therefore, a dense
cumulonimbus cloud before or during precipitation will lead to
HUO et al.: DAMU NETWORKS FOR NEXT-GENERATION WIRELESS COMMUNICATION AND POWER TRANSFER
Fig. 13. Attenuation due to typical meteorological conditions for 5G high
bands.
the most significant attenuation. The effects of weather should
be thoroughly considered with other types of propagation loss
when conducting the DAMU networks design and link budget
calculation for 5G and beyond applications.
VI. C ONCLUSION
UAV enabled 5G communication is promising to become an
immediate reality. The design challenges of UAV enabled 5G
communications are more pressing than terrestrial 5G com-
munications because of more comprehensive system designs
by considering several critical factors, namely power transfer-
ring, wireless communication, aerodynamic constraints, and
meteorological variables. Our proposed DAMU networking
architecture is hierarchical, flexible, and can be reconfigured
according to specific deployment and application scenarios.
Our numerical analysis of the efficiency of laser transmis-
sion and the propagation attenuation introduced by typical
meteorological conditions provides the guidelines for an over-
all link budget calculation and system robustness design
analysis.
Further research on DAMU networks may include but not
limited to:
1) quantitative analysis and optimization of UAVs’ 3-D
deployment and trajectory design in the DAMU
networks which can involve various urban and suburban
environments;
2) interference cancellation and sum rate maximization for
wireless communication and power transfer co-design;
3) precise modeling of A2A channel propagation and the
nonstationary features of UAVs; analysis of their impacts
on the mmWave communications and power transfer;
4) development of computer vision enabled and deep
learning-aided safe laser power transfer technologies,
which can be used not only for DAMU networks but
also for other IoT applications;
7113
5) in order to overcome the challenging issues of UAV
communications during the adverse weather conditions,
a paradigm of judiciously using both sub-6 GHz and
mmWave bands should be developed;
6) secure UAV communication designs with hierarchical
network management to cooperate the UAVs in the
DAMU networks.
R EFERENCES
[1] Y. Huo, X. Dong, T. Lu, W. Xu, and M. Yuen, “Distributed and
multilayer UAV network for the next-generation wireless communica-
tion,” arXiv preprint arXiv:1805.01534, 2018.
[2] T. Tomic et al., “Toward a fully autonomous UAV: Research platform
for indoor and outdoor urban search and rescue,” IEEE Robot. Autom.
Mag., vol. 19, no. 3, pp. 46–56, Sep. 2012.
[3] D. Erdos, A. Erdos, and S. E. Watkins, “An experimental UAV system
for search and rescue challenge,” IEEE Aerosp. Electron. Syst. Mag.,
vol. 28, no. 5, pp. 32–37, May 2013.
[4] M. Erdelj, E. Natalizio, K. R. Chowdhury, and I. F. Akyildiz, “Help from
the sky: Leveraging UAVs for disaster management,” IEEE Pervasive
Comput., vol. 16, no. 1, pp. 24–32, Jan./Mar. 2017.
[5] S. Ortiz et al., “An UAV-based content delivery architecture for rural
areas and future smart cities,” IEEE Internet Comput., vol. 23, no. 1,
pp. 29–36, Jan./Feb. 2019.
[6] “IMT vision framework and overall objectives of the future develop-
ment of IMT for 2020 and beyond,” Int. Telecommun. Union, Geneva,
Switzerland, ITU-Recommendation M 2083, 2015.
[7] FCC. (2016). Report and Order and Future Notice of Proposed
Rulemaking. [Online]. Available: https://docs.fcc.gov/public/
attachments/FCC-16-89A1.pdf
[8] B. Li, Z. Fei, and Y. Zhang, “UAV communications for 5G and beyond:
Recent advances and future trends,” IEEE Internet Things J., to be
published.
[9] Y. Huo, X. Dong, and W. Xu, “5G cellular user equipment: From theory
to practical hardware design,” IEEE Access, vol. 5, pp. 13992–14010,
2017.
[10] T. S. Rappaport et al., “Overview of millimeter wave communica-
tions for fifth-generation (5G) wireless networks—With a focus on
propagation models,” IEEE Trans. Antennas Propag., vol. 65, no. 12,
pp. 6213–6230, Dec. 2017.
[11] Y. Huo and X. Dong, “Millimeter-wave for unmanned aerial vehicles
networks: Enabling multi-beam multi-stream communications,” arXiv
preprint arXiv:1810.06923, 2018.
[12] J. Zhao et al., “Beam tracking for UAV mounted SatCom on-the-move
with massive antenna array,” IEEE J. Sel. Areas Commun., vol. 36, no. 2,
pp. 363–375, Feb. 2018.
[13] Y. Zeng, R. Zhang, and T. J. Lim, “Wireless communications
with unmanned aerial vehicles: Opportunities and challenges,” IEEE
Commun. Mag., vol. 52, no. 2, pp. 122–130, May 2016.
[14] I. Bor-Yaliniz and H. Yanikomeroglu, “The new frontier in RAN hetero-
geneity: Multi-tier drone-cells,” IEEE Commun. Mag., vol. 54, no. 11,
pp. 48–55, Nov. 2016.
[15] S. Sekander, H. Tabassum, and E. Hossain, “Multi-tier drone architecture
for 5G/B5G cellular networks: Challenges, trends, and prospects,” IEEE
Commun. Mag., vol. 56, no. 3, pp. 96–103, Mar. 2018.
[16] S. F. Yunas, M. Valkama, and J. Niemela, “Spectral and energy efficiency
of ultra-dense networks under different deployment strategies,” IEEE
Commun. Mag., vol. 53, no. 1, pp. 90–100, Jan. 2015.
[17] Federal Aviation Administration. Unmanned Aircraft Systems. Accessed:
Mar. 1, 2019. [Online]. Available: https://www.faa.gov/uas/
[18] M. Alzenad, A. El-keyi, F. Lagum, and H. Yanikomeroglu, “3D place-
ment of an unmanned aerial vehicle base station (UAV-BS) for energy-
efficient maximal coverage,” IEEE Wireless Commun. Lett., vol. 6, no. 4,
pp. 434–437, Aug. 2017.
[19] F. Jiang and A. L. Swindlehurst, “Optimization of UAV heading for
the ground-to-air uplink,” IEEE J. Sel. Areas Commun., vol. 30, no. 5,
pp. 993–1005, Jun. 2012.
[20] Y. Zeng and R. Zhang, “Energy-efficient UAV communication with tra-
jectory optimization,” IEEE Trans. Wireless Commun., vol. 16, no. 6,
pp. 3747–3760, Jun. 2017.
[21] J.-K. Shiau, D.-M. Ma, P.-Y. Yang, G.-F. Wang, and J. H. Gong, “Design
of a solar power management system for an experimental UAV,” IEEE
Trans. Aerosp. Electron. Syst., vol. 45, no. 4, pp. 1350–1360, Oct. 2009.
7114
[22] Powerlight. Accessed: Mar. 1, 2019. [Online]. Available:
https://powerlighttech.com/the-technology
[23] J. Ouyang, Y. Che, J. Xu, and K. Wu, “Throughput maximization for
laser-powered UAV wireless communication systems,” in Proc. IEEE
ICC Workshops, Kansas City, MO, USA, 2018, pp. 1–6.
[24] Project Loon. Accessed: Mar. 1, 2019. [Online]. Available:
https://x.company/loon
[25] Federal Aviation Administration, Pilot’s Handbook of Aeronautical
Knowledge: FAA-H-8083-25B, Newcastle, WA, USA: Aviation Supplies
Acad., 2016.
[26] Amazon. Accessed: Mar. 1, 2019. [Online]. Available: https://
www.amazon.com/Amazon-Prime-Air/b?ie=UTF8&node=8037720011
[27] Y. Huo, X. Dong, W. Xu, and M. Yuen, “Cellular and WiFi co-design
for 5G user equipment,” in Proc. IEEE 5G World Forum (5GWF),
Silicon Valley, CA, USA, Jul. 2018, pp. 256–261.
[28] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, “Efficient
deployment of multiple unmanned aerial vehicles for optimal wire-
less coverage,” IEEE Commun. Lett., vol. 20, no. 8, pp. 1647–1650,
Aug. 2016.
[29] Y. Zeng, X. Xu, and R. Zhang, “Trajectory design for completion
time minimization in UAV-enabled multicasting,” IEEE Trans. Wireless
Commun., vol. 17, no. 4, pp. 2233–2246, Apr. 2018.
[30] Q. Wu, Y. Zeng, and R. Zhang, “Joint trajectory and communication
design for multi-UAV enabled wireless networks,” IEEE Trans. Wireless
Commun., vol. 17, no. 3, pp. 2109–2121, Mar. 2018.
[31] B. Li, Z. Fei, Y. Zhang, and M. Guizani, “Secure UAV communication
networks over 5G,” IEEE Wireless Commun., vol. 33, no. 1, pp. 78–83,
Jan./Feb. 2019.
[32] C. Liu, J. Lee, and T. Q. S. Quek, “Safeguarding UAV communica-
tions against full-duplex active eavesdropper,” IEEE Trans. Wireless
Commun., to be published.
[33] B. Li, Z. Fei, Y. Dai, and Y. Zhang, “Secrecy-optimized resource alloca-
tion for UAV-assisted relaying networks,” in Proc. IEEE Glob. Commun.
Conf. (GLOBECOM), Abu Dhabi, UAE, Dec. 2018, pp. 1–6.
[34] G. Zhang, Q. Wu, M. Cui, and R. Zhang, “Securing UAV communi-
cations via joint trajectory and power control,” IEEE Trans. Wireless
Commun., vol. 18, no. 2, pp. 1376–1389, Feb. 2019.
[35] DJI. [Online]. Available: https://www.dji.com/ca/matrice100/info#specs
[36] C. R. Valenta and G. D. Durgin, “Harvesting wireless power: Survey
of energy-harvester conversion efficiency in far-field, wireless power
transfer systems,” IEEE Microw. Mag., vol. 15, no. 4, pp. 108–120,
Jun. 2014.
[37] B. Clerckx et al., “Fundamentals of wireless information and power
transfer: From RF energy harvester models to signal and system
designs,” IEEE J. Sel. Areas Commun., vol. 37, no. 1, pp. 4–33,
Jan. 2019.
[38] T. Matsunaga, E. Nishiyama, and I. Toyoda, “5.8-GHz stacked differ-
ential rectenna suitable for large-scale rectenna arrays with DC connec-
tion,” IEEE Trans. Antennas Propag., vol. 63, no. 12, pp. 5944–5949,
Dec. 2015.
[39] Y.-Y. Hu, S. Sun, H. Xu, and H. Sun, “Grid-array rectenna with wide
angle coverage for effectively harvesting RF energy of low power den-
sity,” IEEE Trans. Microw. Theory Techn., vol. 67, no. 1, pp. 402–413,
Jan. 2019.
[40] A. Goldsmith, Wireless Communications. Cambridge, U.K.: Cambridge
Univ. Press, 2005.
[41] G. Lv, W. Chen, X. Liu, and Z. Feng, “A dual-band GaN MMIC
power amplifier with hybrid operating modes for 5G application,” IEEE
Microw. Wireless Compon. Lett., vol. 29, no. 3, pp. 228–230, Mar. 2019.
[42] J. Wong, N. Watanabe, and A. Grebennikov, “High-power high-
efficiency broadband GaN HEMT Doherty amplifiers for base station
applications,” in Proc. IEEE Topical Conf. RF/Microw. Power Amplifiers
Radio Wireless Appl. (PAWR), Anaheim, CA, USA, 2018, pp. 16–19.
[43] A. V. Blank et al., “Conversion characteristics of silicon photo-
voltaic cells for optical beaming,” Proc. SPIE, vol. 10688, May 2018,
Art. no. 106881F.
[44] G. A. Landis, “Photovoltaic receivers for laser beamed power in space,”
J. Propulsion Power, vol. 9, no. 1, pp. 105–112, 1993.
[45] I. I. Kim, B. McArthur, and E. J. Korevaar, “Comparison of laser beam
propagation at 785 nm and 1550 nm in fog and haze for optical wire-
less communications,” in Proc. SPIE, vol. 4214. Boston, MA, USA,
Nov. 2001, pp. 26–37.
[46] B. V. Zeghbroeck, Principles of Semiconductor Devices, Univ.
Colorado, Boulder, CO, USA, 2011. [Online]. Available:
http://ecee.colorado.edu/ bart/book/book/index.html
IEEE INTERNET OF THINGS JOURNAL, VOL. 6, NO. 4, AUGUST 2019
[47] Q. Zhang et al., “Distributed laser charging: A wireless power trans-
fer approach,” IEEE Internet Things J., vol. 5, no. 5, pp. 3853–3864,
Oct. 2018.
[48] Safety of Laser Products—Part 1: Equipment Classification and
Requirements, IEC Standard 60825-1, 2007.
[49] S. Liu and R. A. Dougal, “Dynamic multiphysics model for solar
array,” IEEE Trans. Energy Convers., vol. 17, no. 2, pp. 285–294,
Jun. 2002.
[50] Q. Liu et al., “Charging unplugged: Will distributed laser charging for
mobile wireless power transfer work?” IEEE Veh. Technol. Mag., vol. 11,
no. 4, pp. 36–45, Dec. 2016.
[51] V. Iyer, E. Bayati, R. Nandakumar, A. Majumdar, and S. Gollakota,
“Charging a smartphone across a room using lasers,” Proc. ACM
Interact. Mobile Wearable Ubiquitous Technol., vol. 1, no. 4, p. 143,
2018.
[52] M. Mozaffari, W. Saad, M. Bennis, Y.-H. Nam, and M. Debbah, “A
tutorial on UAVs for wireless networks: Applications, challenges, and
open problems,” arXiv preprint arXiv:1803.00680, 2018.
[53] A. Al-Hourani, S. Kandeepan, and S. Lardner, “Optimal LAP altitude
for maximum coverage,” IEEE Wireless Commun. Lett., vol. 3, no. 6,
pp. 569–572, Dec. 2014.
[54] J. Holis and P. Pechac, “Elevation dependent shadowing model
for mobile communications via high altitude platforms in built-up
areas,” IEEE Trans. Antennas Propag., vol. 56, no. 4, pp. 1078–1084,
Apr. 2008.
[55] A. Al-Hourani, S. Kandeepan, and A. Jamalipour, “Modeling air-to-
ground path loss for low altitude platforms in urban environments,” in
Proc. IEEE Glob. Commun. Conf. (GLOBECOM), Austin, TX, USA,
Dec. 2014, pp. 2898–2904.
[56] T. S. Rappaport et al., “Millimeter wave mobile communications
for 5G cellular: It will work!”, IEEE Access, vol. 1, pp. 335–349,
2013.
[57] L. Zhou, Z. Yang, G. Zhao, S. Zhou, and C. Wang, “Propagation char-
acteristics of air-to-air channels in urban environments,” in Proc. IEEE
Glob. Commun. Conf. (GLOBECOM), Abu Dhabi, UAE, Dec. 2018,
pp. 1–6.
[58] “Attenuation due to clouds and fog,” Int. Telecommun. Union, Geneva,
Switzerland, ITU-Recommendation P.840-7, Dec. 2017. [Online].
Available: https://www.itu.int/rec/R-REC-P.840-7-201712-I/en
[59] V. Chakrapani, D. R. Doelling, A. D. Rapp, and P. Minnis, “Cloud
thickness estimation from GOES-8 satellite data over the ARM-SGP
site,” in Proc. 12th ARM Sci. Team Meeting, St. Petersburg, FL, USA,
Apr. 2002, pp. 1–7.
Yiming Huo (S’08–M’18) received the B.Eng.
degree in information engineering from Southeast
University, Nanjing, China, in 2006, the M.Sc.
degree in system-on-chip from Lund University,
Lund, Sweden, in 2010, and the Ph.D. degree in elec-
trical engineering from the University of Victoria,
Victoria, BC, Canada, in 2017.
He is currently a Post-Doctoral Research Fellow
with the University of Victoria. His current research
interests include 5G and beyond wireless systems,
Internet of Things, and machine learning. He has
worked with several companies and institutes, including Ericsson, Lund,
Sweden, ST-Ericsson, Lund, the Chinese Academy of Sciences, Shanghai,
China, STMicroelectronics, Shenzhen, China, and Apple Inc., Cupertino, CA,
USA.
Dr. Huo was a recipient of the Best Student Paper Award of 2016 IEEE
ICUWB, the Excellent Student Paper Award of 2014 IEEE ICSICT, the
Bronze Leaf Certificate of 2010 IEEE PrimeAsia, and the ISSCC-STGA
Award of the IEEE Solid-State Circuits Society in 2017. He is a member
of several IEEE societies. He has served as a Program Committee member
of IEEE ICUWB 2017, IEEE VTC 2018/2019, and IEEE ICC 2019, the
Session Chair of IEEE 5G World Forum 2018, the Publication Chair of IEEE
PACRIM 2019, and as a Technical Reviewer for multiple premier IEEE con-
ferences and journals. He is also a member of the IEEE Future Networks
Technology Roadmap Working Group.
HUO et al.: DAMU NETWORKS FOR NEXT-GENERATION WIRELESS COMMUNICATION AND POWER TRANSFER
Xiaodai Dong (S’97–M’00–SM’09) received the
B.Sc. degree in information and control engineer-
ing from Xi’an Jiaotong University, Xi’an, China,
in 1992, the M.Sc. degree in electrical engineer-
ing from the National University of Singapore,
Singapore, in 1995, and the Ph.D. degree in
electrical and computer engineering from Queen’s
University, Kingston, ON, Canada, in 2000.
From 1999 to 2002, she was with Nortel
Networks, Ottawa, ON, Canada, where she per-
formed research on the base transceiver design of
third-generation mobile communication systems. From 2002 to 2004, she
was an Assistant Professor with the Department of Electrical and Computer
Engineering, University of Alberta, Edmonton, AB, Canada. Since 2005, she
has been with the University of Victoria, Victoria, BC, Canada, where she
is currently a Professor with the Department of Electrical and Computer
Engineering. Her current research interests include 5G, mmWave communi-
cations, radio propagation, Internet of Things, machine learning, localization,
wireless security, e-health, smart grid, and nano-communications.
Dr. Dong served as an Editor for the IEEE T RANSACTIONS ON
C OMMUNICATIONS from 2001 to 2007 and the IEEE T RANSACTIONS ON
W IRELESS C OMMUNICATIONS from 2009 to 2014. She is an Editor of the
IEEE T RANSACTIONS ON V EHICULAR T ECHNOLOGY. She was the Canada
Research Chair (Tier II) from 2005 to 2015.
Tao Lu (M’17) received the B.Sc. degree from
the Department of Physics, University of Manitoba,
Winnipeg, MB, Canada, in 1995, the M.Sc. degree
from the Department of Electrical and Computer
Engineering, Queen’s University, Kingston, ON,
Canada, in 1998, and the Ph.D. degree from the
Department of Physics, University of Waterloo,
Waterloo, ON, Canada, in 2005.
He has worked in industry with various companies
including Nortel Networks, Ottawa, ON, Canada;
Kymata Canada; and Peleton, during which time he
was involved with optical communications. He was a Post-Doctoral Fellow
with the Department of Applied Physics, California Institute of Technology,
Pasadena, CA, USA, from 2006 to 2008. He is currently with the University of
Victoria, Victoria, BC, Canada. His current research interests include optical
microcavities and their applications to ultra-narrow linewidth laser sources,
bio- and nano-photonics, machine learning algorithms with applications to
spectral analysis, Internet of Things, and indoor localization.
7115
Wei Xu (S’07–M’09–SM’15) received the B.Sc.
degree in electrical engineering and the M.S. and
Ph.D. degrees in communication and information
engineering from Southeast University, Nanjing,
China, in 2003, 2006, and 2009, respectively.
From 2009 to 2010, he was a Post-Doctoral
Research Fellow with the Department of Electrical
and Computer Engineering, University of Victoria,
Victoria, BC, Canada. He is currently a Professor
with the National Mobile Communications Research
Laboratory, Southeast University. He has coauthored
over 80 refereed journal papers in addition to holding 28 granted domestic
patents and 3 U.S. patents. His current research interests include cooperative
communications, information theory, signal processing, and machine learning
for wireless communications.
Dr. Xu was a recipient of the Best Paper Awards of IEEE MAPE in 2013,
IEEE/CIC ICCC in 2014, IEEE GLOBECOM in 2014, IEEE ICUWB in 2016,
and WCSP in 2017. He was a co-recipient of the First Prize of the Science
and Technology Award of Jiangsu Province, China, in 2014. He has served on
Technical Program Committees for international conferences, including IEEE
GLOBECOM, IEEE ICC, IEEE WCNC, IEEE VTC, and IEEE PIMRC. He
was an Editor of IEEE C OMMUNICATIONS L ETTERS, from 2012 to 2017 and
he is currently an Editor of the IEEE T RANSACTIONS ON C OMMUNICATIONS
and IEEE ACCESS.
Marvin Yuen (S’09–M’13) received the B.Sc.
degree in electrical and computer engineering from
The Johns Hopkins University, Baltimore, MD,
USA, and the M.S. degree in electrical engineering
from the Viterbi School of Engineering, University
of Southern California, Los Angeles, CA, USA.
He was with USPTO, Alexandria, VA, USA,
as a Patent Examiner. He is currently a System
Application Engineer with Kingston Technology,
Fountain Valley, CA, USA.