This paper has been accepted for publication in IEEE Network Magazine, ©2020 IEEE.

Please cite it as: M. Giordani and M. Zorzi, ”Non-Terrestrial Networks in the 6G Era: Challenges and Opportunities,”
in IEEE Network, vol. 35, no. 2, pp. 244-251, Mar. 2021.

Non-Terrestrial Networks in the 6G Era:

Challenges and Opportunities
Marco Giordani, Member, IEEE, Michele Zorzi, Fellow, IEEE

Abstract—Many organizations recognize non-terrestrial net- dimensional space, 6G envisions a three-dimensional (3D) het-
works (NTNs) as a key component to provide cost-effective erogeneous architecture in which terrestrial infrastructures are
and high-capacity connectivity in future 6th generation (6G) complemented by non-terrestrial stations including Unmanned
wireless networks. Despite this premise, there are still many
questions to be answered for proper network design, including Aerial Vehicles (UAVs), High Altitude Platforms (HAPs), and
those associated to latency and coverage constraints. In this satellites [5]. Not only can these elements provide on-demand
arXiv:1912.10226v3 [cs.NI] 26 Apr 2021

paper, after reviewing research activities on NTNs, we present cost-effective coverage in crowded and unserved areas, but
the characteristics and enabling technologies of NTNs in the 6G they can also guarantee trunking, backhauling, support for
landscape and shed light on the challenges in the field that are high-speed mobility, and high-throughput hybrid multiplay
still open for future research. As a case study, we evaluate the
performance of an NTN scenario in which aerial/space vehicles services. Notably, the potential of NTNs has been acknowl-
use millimeter wave (mmWave) frequencies to provide access edged in the standard activities. A work item for 3GPP Rel-
connectivity to on-the-ground mobile terminals as a function of 17 has indeed been approved in December 2019 to define and
different networking configurations. evaluate solutions in the field of NTNs for NR, with a priority
Index Terms—6G; non-terrestrial network (NTN); satellites; on satellite access. Study items have also been identified for
unmanned aerial vehicles (UAVs), millimeter waves (mmWaves). Rel-18 and Rel-19, thus acknowledging long-term research
within the timeframe of 6G.
Research studies on NTN are not only limited to 3GPP
reports. For instance, Babich et al. presented a novel network
I. I NTRODUCTION architecture for an integrated nanosatellite-5G system oper-
ating in the millimeter wave (mmWave) domain [6], while
While network operators have already started deploying in our previous work [5] we identified the most promising
commercial 5th generation (5G) cellular networks, the research configuration(s) for satellite networking and discussed some
community is discussing use cases, requirements, and enabling design trade-offs in this domain. UAVs were also considered
technologies towards 6th generation (6G) systems [1]. Among as a tool to complement terrestrial connectivity in critical
other challenges, current networks fall short of providing scenarios [7]. Additionally, there currently exist several case
adequate broadband coverage to rural regions [2]. Moreover, studies of NTN deployments in different countries, in addition
even in the most technologically advanced countries, exist- to efforts by international foundations and initiatives [2].
ing cellular infrastructures may lack the level of reliability, Nevertheless, despite such earlier investigations, there are
availability, and responsiveness requested by future wireless still several questions to be answered for proper network
applications, and show vulnerability to natural disasters. Con- design. In particular, while some prior work typically focuses
nectivity outages during natural disasters, in particular, may on standalone aerial/space architectures, a formalization of
slow down or impede appropriate reaction, create significant the challenges and opportunities pertaining to a multi-layered
damage to business and property, and even loss of lives. network, in which heterogeneous non-terrestrial stations co-
One solution to increase network resiliency would be to operate at different altitudes in an integrated fashion, has
densify cellular sites, which however involves prohibitive not yet been provided. Some other articles, e.g., [8], review
deployment and operational expenditures for network oper- how to improve the protocol stack design in a space-air-
ators and requires high-capacity backhaul connections [2], ground integrated network, but do not thoroughly explore
[3]. Moreover, network deployment in rural areas (i.e., the the most recent technological advancements to achieve per-
most under-connected areas) is further complicated by the formance optimization. Moreover, a complete description of
varying degree of terrain that may be encountered when NTN enabling technological solutions and future research
installing cables or fibers between cellular stations. Network directions is currently scattered in several technical reports,
densification will also inevitably lead to an energy crunch with which makes them confusing and tiresome to follow without
serious economic and environmental concerns. the proper background.
To address these issues, 6G research is currently focusing This paper addresses these challenges by formalizing how
on the development of non-terrestrial networks (NTNs) to NTNs can be practically deployed to satisfy emerging 6G
promote ubiquitous and high-capacity global connectivity [4]. application requirements. We focus on (i) new architecture
While previous wireless generation networks have been tra- advancements in the aerial/space industry, (ii) novel spectrum
ditionally designed to provide connectivity for a quasi bi- technologies, e.g., operating in the mmWave and optical bands,
(iii) antenna design advancements, and (iv) transport layer
Marco Giordani and Michele Zorzi are with the Department of Infor-
mation Engineering, University of Padova, Padova, Italy (email: {giordani, developments. Moreover, we shed light on the research chal-
zorzi}@dei.unipd.it). lenges associated to NTNs, providing a full-stack perspective
 GEO ~ 36000 km Relaying
5G Mobile
Edge Cloud
LEO ~ 300 km
Backhauling

Multi-connectivity for
Service boosting for users service continuity
Distributed computation
HAPS ~ 20 km in crowded areas
and content broadcasting eMBB in unserved
and disaster areas
UAV ~ 100 m

Fig. 1: Non-terrestrial stations (left) and use cases enabled by the integration of terrestrial and non-terrestrial networks (right).

with considerations related to spectrum usage, medium access services, without the prohibitive costs of terrestrial infrastruc-
and higher layers, coverage and mobility management con- tures. However, HAPs may suffer from the need for refueling
straints, thus stimulating further research on this topic. Finally, and challenges related to stabilization in the air.
as a case study, we validate the feasibility of establishing Satellites. Satellite stations can be classified according to
non-terrestrial communication at mmWaves to provide access their orbit characteristics. Geostationary Earth Orbit (GEO)
connectivity to terrestrial nodes. satellites orbit on the Earth’s equatorial plane at an altitude of
about 35,800 km and, despite the significant signal propagation
II. N ON -T ERRESTRIAL N ETWORKS IN 6G delay and attenuation experienced at such long distance, can
cover very large geographical areas and are continuously
NTNs refer to (segments of) networks operating through
visible from terrestrial terminals. Low Earth Orbit (LEO) and
an air/spaceborne vehicle for communication. While the pos-
Medium Earth Orbit (MEO) satellites, instead, orbit at an
sibility to integrate satellite technologies to provide access
altitude between 200 and 2,000 km and 2,000 and 35,000 km,
connectivity on the ground was first introduced by 3GPP in
respectively, and guarantee better signal strength and lower
Rel-15, more recent activities have been promoted for Rel-16
propagation delay compared to GEO systems. However, these
and Rel-17 to define deployment scenarios and parameters,
satellites are non-stationary relative to the Earth’s surface and
and identify key potential impacts in NR [9]. Specifications
must operate in a constellation to maintain service continuity.
may also continue with enhancements in Rel-18 and Rel-19.
The 3GPP is promoting different NTN architectures depending
Based on this introduction, in Sec. II-A we describe a typical
on the degree of integration among the different air/spaceborne
non-terrestrial architecture, while in Sec. II-B we present
elements [9]. Specifically, the 3GPP envisions:
potential use cases and related deployment scenarios.
(1) a transparent satellite-based Radio Access Network
(RAN) architecture in which the satellite repeats the
A. General Architecture user’s signal from the feeder link to the service link and
Non-terrestrial systems feature (i) a terrestrial terminal, vice versa;
(ii) an aerial/space station, which may operate similarly to (2) a regenerative satellite-based RAN architecture in which
a terrestrial base station, (iii) a service link between the the satellite payload implements regeneration of the sig-
terrestrial terminal and the aerial/space station, and (iv) a nals received from the Earth, while also providing inter-
gateway that connects the non-terrestrial access network to the satellite connectivity;
core network through a feeder link. Different types of stations (3) a multi-connectivity architecture involving two transpar-
can be considered, as depicted in Fig. 1 (left). ent RANs (either GEO or LEO or a combination thereof),
where integration of terrestrial and non-terrestrial access
Unmanned Aerial Vehicle (UAV). UAVs fly at low altitudes
is also supported.
(e.g., a few hundred meters) and, thanks to their flexibility,
have recently gained increasing attention to provide broadband
wide-scale wireless connectivity during disasters or temporary B. Use Cases
events, and relay services for terrestrial mobile nodes. On the For many years, non-terrestrial devices have been consid-
one hand, UAVs can be deployed on-demand, thereby promot- ered to support services like home delivery, meteorology,
ing energy efficiency compared to always-on fixed terrestrial video surveillance, television broadcasting, remote sensing,
infrastructures. On the other hand, UAVs incur high propulsion and navigation. However, recent technological developments
energy consumption to maintain and support their movement, in the aerial/space industry have opened up the way towards
thereby posing severe power management constraints. integration between terrestrial and non-terrestrial technologies
High Altitude Platform (HAP). HAPs operate in the strato- to enable more advanced use cases, as illustrated in Fig. 1
sphere at an altitude of around 20 km. Thanks to their (right) and summarized below.
quick deployment and geographical coverage of hundreds of Communication resilience and service continuity. Non-
kilometers, these elements are indeed being considered to terrestrial stations can be deployed to assist existing base sta-
support ultra-flexible deployment and cost-effective wireless tions in providing high-capacity wireless coverage, e.g., in hot-
 TABLE I: Enabling technologies for non-terrestrial networks.

Technology Advantage
Nano/pico satellites Small component costs, low latency, low energy consumption
Gallium Nitride (GaN) Feasible to install, small form-factor and more efficient components
Multi-layered networks Better spatial and temporal coverage by deploying satellites in different orbits
Architecture Solid-state lithium batteries Safe and efficient source of power
Software Defined Networking (SDN) Improved flexibility, automation, agility through Virtualization Network Functions (VNFs)
Flexible payloads Dynamic adaptation of beam patterns, frequency, and power allocation
Hybrid payloads Better trade-off between performance and payload complexity
Millimeter waves Feasibility of ultra-fast connections, antenna gain, spatial isolation and security
UWB modulation Reduced non-linear signal distortion by encoding the transmitted pulse
Spectrum
Cognitive spectrum Reduced interference through dynamic spectrum utilization in different frequency bands
Optical communications Feasibility of terabits-per-second connections through extreme bandwidth and directivity
Reconfigurable phased antennas Reduced power consumption, size and weight
Metasurface antennas Component miniaturization, high directivity, low sidelobes, fine beamwidth control
Antenna Inflatable/fractal antennas High-directivity in dynamic scenarios
Coherent antenna arrays Maintainability, scalability, flexibility, robustness to single points of failure
Multi-beam architectures High spectrum efficiency through spatial diversity
TCP spoofing Fast TCP full-buffer capacity through TCP acknowledgements
Higher layers
TCP multiplexing High performance by splitting TCP session into multiple data flows

spot areas or when terrestrial infrastructures are overloaded. saving terrestrial resources for the access traffic and avoiding
Non-terrestrial elements can also provide a secondary backup the costs of traditional fiber-like deployments. Satellites and
route to preserve the connection when the primary path is other aerial platforms can also complement the terrestrial
unavailable, e.g., in rural areas or oceans, or when terrestrial backhaul in dense regions with high peak traffic demands,
towers are out of service, e.g., after natural disasters. Addi- thus achieving load balancing.
tionally, these elements can provide on-demand extra capacity Energy-efficient hybrid multiplay. Air/spaceborne platforms
to cell-edge users, the most resource-constrained network enti- have the ability to provide high-speed connectivity while
ties, thereby promoting fairness in the network. Finally, aerial promoting energy efficiency. On one side, aerial platforms like
platforms can host Mobile Edge Cloud (MEC) functionalities UAVs, while consuming significant energy for hovering, can
to offer on-the-ground terminals additional computing and be deployed on demand implementing smart duty cycle control
storage capabilities, thereby evolving coverage towards 3D. mechanisms, thereby reducing management costs of always-
Even though the limited energy support from battery may on fixed terrestrial infrastructures. On the other side, space
render the MEC environment challenging, machine-learning- platforms like satellites can be operated by solar panels which
assisted migration and technologies for renewable energy provide efficient, clean, and renewable energy compared to
production/harvesting and storage are studied to minimize traditional energy sources powering terrestrial devices.
power consumption [10].
Global satellite overlay. When the distance between two ter- III. N ON -T ERRESTRIAL N ETWORKS :
restrial infrastructures increases, inter-site connectivity through E NABLING T ECHNOLOGIES
optical fiber may become too expensive. A constellation The evolution of NTNs will be favored by recent technolog-
of satellites, where each spacecraft is interconnected with ical advancements in the aerial/space industry, as summarized
other neighboring spacecrafts via inter-satellite links, can then in Table I and described in the following subsections. We focus
provide high-capacity access connectivity to on-the-ground on the innovations that do not currently fall within the scope
devices by relaying the user’s signals through an overlay space of early 5G standard activities but could flourish in 6G.
mesh network.
Ubiquitous Internet of Things (IoT) broadcasting. The wide A. Architecture advancements
geographical coverage and the inherent broadcast nature of Space manufacturers are improving satellite technologies
aerial/space platforms make it possible to convey multimedia while further reducing the operational costs for satellite launch,
and entertainment contents to a very large number of user deployment, and maintenance. Nano- and pico-satellites in
equipments, including in-motion terminals that cannot benefit the LEO orbits, in particular, are emerging as game-changing
from terrestrial coverage like planes or vessels. UAVs and innovations thanks to their reduced component costs, and low
satellites can also play the role of moving aggregators for communication latency and energy consumption. Moreover,
IoT traffic, thereby offering global continuity of service for the adoption of the Gallium Nitride (GaN) technologies on
applications that rely on sensors. satellites allows the use of smaller form factors and more
Advanced backhauling. Non-terrestrial terminals can serve efficient components compared to their silicon counterparts,
on-the-ground backhaul requests wirelessly, e.g., for locations thereby saving fuel and area on the payload and improv-
where no wired backhaul solutions are available, thereby ing operational efficiency [11]. Today, the commercialization
of GaN products is restricted to military applications, with bandwidth and directivity compared to radio-frequency sys-
most 5G devices utilizing silicon wafer substrates, but their tems and consume much less power and mass. In this context,
adoption in commercial networks may still be realized for atmospheric perturbations and interference from sunlight can
6G. Additionally, the availability of multi-layered satellite be mitigated by wavefront correctors and deformable mir-
networks, e.g., LEO and GEO constellations, makes it possible rors, which compensate the signal distortion after propagating
to obtain better spatial/temporal coverage. Nevertheless, a real through the atmosphere, and advanced modulation schemes.
integration between terrestrial and non-terrestrial networks still Error control coding also improves the performance of the
seems far in the future, and standardization activities are optical link by making use of Turbo and convolutional codes.
scheduled within the timeframe of 6G. Nevertheless, despite this potential, standardization bodies
UAV technology has also improved recently. Solid-state have not yet considered inclusion of optical solutions in the
lithium batteries, in particular, make it possible for UAVs to NTN standard, and will be targeting beyond-5G use cases.
work twice as long compared to today’s aerial devices, and
are being considered as a safer and more efficient alternative C. Antenna advancements
compared to standard lithium-ion batteries. Furthermore, UAV
swarms, combined with HAPs and satellites, can operate Aerial/space devices can be equipped with reconfigurable
together to support more robust information broadcasting phased antennas offering electronic beam-steering to achieve
compared to a standalone deployment by adding redundancy lower power consumption and reduced size and weight com-
against single points of failure in the path. pared to typical mechanical antennas. Programmable environ-
Architecture optimization is also favored by the transition ments enabled by metasurfaces and intelligent structures are
to Software Defined Networking (SDN) [12] which, in com- another revolutionary element of the 6G ecosystem to realize
bination with network slicing, facilitates the deployment and antenna component miniaturization, improved directivity, low
management of Virtualization Network Functions (VNFs) onto sidelobes and fine beamwidth control [15]. Future trends in
the same physical platform. Furthering a trend already started the antenna domain further suggest the use of inflatable (i.e.,
in 5G, 6G will contribute to the design of a disaggregated made with flexible-membrane materials) and fractal antennas
architecture that can operate in view of the competitive na- with unique geometrical designs to obtain high directivity in
ture of the non-terrestrial environment to guarantee improved dynamic scenarios. Additionally, UAVs and/or nano-satellites
flexibility, automation, and agility in the delivery of services (e.g., in the LEO orbit) can be deployed in swarms to obtain
to terrestrial terminals. Satellite payloads can be realized a distributed coherent antenna array to realize extremely nar-
in software to flexibly adapt beam patterns, frequency, and rowbeam transmissions. Such solution offers maintainability
power allocation, and react to the dynamics foreseen in future and scalability, as elements can be easily arranged without
wireless traffic. Moreover, hybrid payload implementations, in affecting system operations, and robustness to single points
which the burden of signal processing is split between the of failure.
on-the-ground gateway and the non-terrestrial station, have Advanced antenna solutions allow the implementation of
been recently studied to achieve better trade-offs between multi-beam architectures that send information to different
performance and payload complexity. spots on the ground through a plurality of beams, thereby
maximizing spectrum efficiency through spatial diversity. The
B. Spectrum advancements multi-beam approach is further favored by operations in the
mmWave and optical domains, where the wavelength is so
Non-terrestrial devices have typically been operated in the small that it becomes practical to build large antenna arrays
legacy frequency bands below 6 GHz which, however, may in a small space while maximizing antenna gains through
not satisfy the boldest data rate requirements of future beyond- beamforming.
5G services. Capacity issues can be solved by transitioning to
high-frequency communications in the mmWave and optical
bands, where the huge bandwidths available may offer the D. Higher-layer advancements
opportunity of ultra-fast connections. However, while the NTNs come with their own set of challenges compared to
adoption of the mmWave spectrum is being successful in standalone terrestrial systems, which might make standard
the 5G market for both cellular and vehicular networks, it is transmission protocols, including congestion control over
still unclear whether this technology can be used in the non- Transmission Control Protocol (TCP), less effective. Network
terrestrial environment. Solutions are being proposed towards operators have therefore developed acceleration techniques
the development of new waveforms and modulation schemes, that make transport protocols perform better. TCP spoofing,
e.g., impulse-based ultra-wideband (UWB) modulation where in particular, is used to send false TCP acknowledgements
information is encoded depending of the characteristics of the to terrestrial terminals from a spoofing entity (or software)
transmitted pulse, as a viable approach to reduce the non-linear nearby, as if they were sent from the aerial/space station,
signal distortion typically experienced at high frequencies [13]. thereby making it possible for the TCP control mechanism to
Moreover, cognitive spectrum techniques may enable dynamic quickly reach the maximum supported rate. TCP multiplexing
spectrum utilization in different bands, while minimizing in- is another solution that converts a single TCP session into
terference. multiple data flows, each of which can adjust its TCP
Optical wireless technology can also be used in the feeder parameters to match the characteristics of the non-terrestrial
link to achieve aggregate capacity in the order of terabits- connection.
per-second [14]. Optical transceivers, in fact, leverage higher
 20 ↵ = 10 ↵ = 50 ↵ = 90
20 ↵ = 10 ↵ = 50 ↵ = 90
80 20 ↵ = 10 Dense urban
↵ = 50 Rural = 90
↵
C [Gbps]

80 h =300 km (LEO) Dense urban Rural

C [Gbps]
80 Dense urban Rural
Shannon capacity C [Gbps]

h =300 km (LEO) 15

Shannon capacity C [Gbps]

C [Gbps]

60 h =300 kmkm
h =10,000 (LEO)
(MEO)

C [Gbps]
15
60 h =10,000 km (MEO) 15
km (MEO)
capacity

60 h =10,000
h =36,000 km (GEO)

capacity
h =36,000 km (GEO) 10
capacity

40 h =36,000 km (GEO)

capacity
40 10
10
Shannon

40

Shannon
Shannon

20

Shannon
5
20 5
20 5
0
0 0
0 50 0 LEO0 LEO LEO LEO
50 100 100 50
150 200 150 50 0 LEO
6 GHz LEO
020 GHz LEO70 GHz LEO
150 GHz
50 100 150 100 0 LEO LEO LEO LEO
100 150 200 100 50 6 GHz 20 GHz 70 GHz 150 GHz
Frequency [GHz] 150 Grx [dBi] 6 GHz
Frequency [GHz]
150 200
Grx [dBi] Fig. 3: Shannon capacity vs. 20
↵ GHz 70 GHzscenario with
for an LEO-GND 150 GHz
h =300
Fig.Fig.
km,
3:
3: with Shannon
Shannon capacity
capacity vs. vs. α for
↵ for an LEO-GND
an LEO-GND
antenna-gain-to-noise-temperature
scenario
scenario
15.9with
with
dBi/K
h =300
h [9].
=300
Frequency [GHz]
Fig. 2: Shannon capacity vs. h, fc and Grx , with ↵ = 10 and for a dense km,km, Grx [dBi]
Fig.
with3:antenna-gain-to-noise-temperature
with Shannon capacity vs. ↵ for anGLEO-GND
antenna-gain-to-noise-temperature
Grx /T
rx G
=scenario
/T15.9
/Trx= = 15.9 dBi/K
dBi/K [9]. h Dense
withDense
[9]. Dense
=300
Fig. 2: Shannon capacity vs. h, f and G , with ↵ = 10 and for a dense urban (rural)
km,(rural)
with scenario for plain (striped)
antenna-gain-to-noise-temperature bars.
Grx /T = 15.9 dBi/K [9]. Dense
Fig. 2: Shannon
urban scenario. capacity vs. h, fc c and G rxrx , with α = 10° and for a dense urban
urban (rural) scenario
scenario for for plain
plain (striped)
(striped) bars.bars.
Fig. 2: Shannon
urban scenario. capacity vs. h, f c and G rx , with ↵ = 10 and for a dense
urban urban (rural) scenario for plain (striped) bars.
urbanscenario.
scenario.
↵ = 10 ↵ = 50 ↵ = 90
IV. N ON -T ERRESTRIAL N ETWORKS : ↵ = 10 ↵ = 50 ↵ = 90
IV.
IV. NN ON -T ERRESTRIAL N ETWORKS :
ON -T ERRESTRIAL N ETWORKS : ↵ =LEO-GND
10 ↵ = 50LEO-HAP-GND ↵ = 90
IV. N ONA -TAC C ASE S TUDY
ERRESTRIAL
ASE S TUDY
N ETWORKS : 10 LEO-GND LEO-HAP-GND
AACCASE ASESSTUDY
10

C [Gbps]
TUDY LEO-GND LEO-HAP-GND

Shannon capacity C [Gbps]

AsAsa acase casestudy,
study,ininthis thissectionsection weweassess assessthethefeasibilityfeasibility 10

C [Gbps]
AsAs a acase
ofofestablishing
establishingcasestudy, mmWave
study,
mmWave
ininthis this section
section we
communications
communications we assess between
assess
between
the feasibility
terrestrial
theterrestrial
feasibility
of and
ofestablishing
satellite
establishing mmWave
terminals,
mmWave communications
possibly through
communications
and satellite terminals, possibly through hybrid integration between
hybrid
between terrestrial
integration of of
terrestrial capacity
and and satellite
satellite terminals,
terminals, possibly
multiple aerial/space layers. This choice was driven by thethe
multiple aerial/space layers.
possibly Thisthroughchoice
through hybrid
was
hybrid integration
driven
integration by of
of
capacity
5
5
multiple
fact that
multiple aerial/space
the use
aerial/space of layers.
satellites
layers. This
operating
This choice
choice in was
thewas driven
mmWave
driven by the
bands,
by the
Shannon

fact that the use of satellites operating in the mmWave bands, 5

factfactthat
among that the
allthe use
the
use of ofsatellites
technologies
satellitesdiscussedoperating
discussedin
operating inininthe
the mmWave
Sec. III,III,currently
mmWave bands,
currently
bands,
Shannon

among all the technologies Sec.

represents
among
amongallall
represents oneone
the oftechnologies
themost
theoftechnologies
the mostpromisingpromising
discussed
discussed innovations
inin Sec.
innovations Sec. III, III,(as
(as already
currently
currently
already
successfully
represents
representsone
successfully demonstrated
one ofofthe
demonstrated themost most inin thecellular
promising
promising
the cellular andvehicular
innovations
innovations
and vehicular (as fields)
already
(asfields)
already
toto offerhigh-capacity
successfully
successfully
offer high-capacity
demonstrated
demonstrated broadcasting
broadcasting ininthe thecellular capability
cellular
capability and in
andinvehicular NTNs.fields)
vehicular
NTNs. fields) 0 0
200 200 300 300 600 600 1200 1200
totoIn Inoffer
offer
our high-capacity
ourhigh-capacity
simulations,a broadcasting
simulations, broadcasting
aterrestrial
terrestrialterminal capability
terminal
capability inin NTNs.
communicates
communicates NTNs.with with 0
200 300
LEO orbit altitude
LEO orbit altitude h [km] h 600
[km] 1200
aasatellite
satellite
In Inour placedatatdifferent
oursimulations,
simulations,
placed different
a aterrestrial
terrestrial altitudes
altitudes terminal
terminalh,h,and and wewe
communicates
communicates consider
consider dif-
with
with
dif- LEO orbit altitude h [km]
asatellite
satellite
ferent
aferent placed
elevation
elevation placed at different
angles
angles different
↵↵22{10 altitudes
{10 altitudes
, .,... ., .90,h, },and
90h, and
},andwepropagation
and consider
we propagation
considerdif- Fig.Fig.
GHz.
4: Shannon
4: Shannon capacity
capacity vs. vs.
↵ and↵ andh forh afordense
a dense urban urban scenario
scenario at fcat=fc20= 20
GHz.
Fig.We
Fig. 4:We
4: compare
compare
Shannon
Shannon an an LEO-GND
LEO-GND
capacity vs. ↵
α and(plain
(plain hbars)
forbars)
for avs.
a avs.
dense
dense aurban
multi-layered
multi-layered
urban scenario
scenario LEO-HAP-
LEO-HAP-
at
atffcc==20 20
ferent elevation
scenarios.
different
scenarios. Thechannel
elevation
The angles
channel
angles isis αmodeled
2∈ {10
↵modeled {10°,as , ..as ..,described
..described
,90
90°} and
}, and byby propagation
the
propagation
the 3GPP3GPP GND GND
GHz. scenario
scenario
We compare (striped
(striped an bars)
bars) in
LEO-GND in which
which a
(plain abars)
HAP HAP deployed
deployed
vs. a at 20 at km
multi-layered 20 acts
km acts
as a as a
LEO-HAP-
GHz. We compare an LEO-GND (plain and a multi-layered LEO-HAP-
scenarios.
inin [9]and
scenarios.
[9] TheThechannel
andsummarized channelinisin
summarized is[5, modeled
[5,
modeled
Sec.Sec.III]: asasspecifically,
III]: described
specifically,
described byby the
the
thethe 3GPP relay.
signal
3GPP
signal relay.
GND scenario
GND scenario (striped bars) in which aa HAP HAP deployed
deployed atat 20 20 km km actsacts asas aa
in[9][9]and
undergoes
undergoes
in andsummarized
summarized
severalstages
several stages ininof [5,
of
[5, Sec.III]:
attenuation
attenuation
Sec. III]:specifically,
specifically,
duedueto toatmospheric the signal
atmospheric
the signal relay.
relay.
undergoes
gases
gases
undergoes and severalstages
andscintillation.
scintillation.
several stages ofof attenuation
Terrestrial
Terrestrial attenuation
stationsdue
stations due
are toto
areequipped atmospheric
equipped
atmosphericwithwith
gasesand
directional
directional
gases andscintillation.
scintillation.
antennas
antennas offering
offering Terrestrial
a again
Terrestrial gainG stations
G
stations
tx tx == 39.7 aredBi
39.7
are equipped
dBi [9][9]
equipped with less,
while,
while,with less, gigabits-per-second
gigabits-per-second capacities
capacities cancan stillstill be reached
be reached if the if the
fordirectional
for satellite
directional antennas
satelliteantennas
stations,the
stations, offering
the gain
offering gainGaG again
rx gain
rxis is Gvaried
variedGtxtx= =to
to consider
consider
39.7
39.7 dBidifferent
dBi [9]
[9] different
while, less,
while, satellite
h, i.e.,gigabits-per-second
satellite station
station forms
transitioning formsvery very
from capacities
sharp sharp
LEO beams, can
tobeams,
GEO still
thereby beboosting
thereby
satellites.reached
boosting if the
the
Neverthe-
for forsatellite
antenna
antenna satellite stations,the
architectures.
architectures.
stations, the gainGGrxrxcommunication
Satellite
Satellite
gain isvaried
variedtotoconsider
iscommunication consider
leverages different
leverages
different satellite
performance
a a performance
less, station
through
gigabits-per-second forms
through very
massive
massive sharp
capacities beams,
beamforming.
beamforming. can still thereby
This This
be boosting
is
reachedpractically
is practically if the
the
antenna
bandwidth
bandwidth
antenna architectures.
WW that
architectures.thatdependsdepends Satellite
Satellite ononthe communication
thefrequencies
communicationfrequenciesfleverages leverages
cf: cwe: wesetset a performance
feasible since through
GEO massive
satellites
satellites beamforming.
a satellite station forms very sharp beams, thereby boosting the
feasible since GEO are are stationary
stationary This is
relative
relative practically
to to
the the
WW bandwidth
bandwidth== 2020W MHzMHz
Wthat that
forfor depends
fcfc on
depends 6 6onGHz,
GHz,
the thecarrier
W frequencies
W =frequency
=800 800MHz fcMHz: fwe forfor feasible
Earth’s
set Earth’s surface
performance sincethrough
surface andGEO
anddo do satellites
not not
massive require
require are
beam stationary
beam
beamforming. This relative
re-alignment.
re-alignment. to the
is practically
c : we
66W
set <<W fcf=c= 20 MHz
20GHz,
6060 MHzand
GHz, for
and
forW fW cf= =2 2GHz6 GHzGHz, forforfW cf> c >=6060 GHz.
800 GHz. MHz for Earth’s
Similar
for feasible
Similar surface
conclusions
since and do
conclusions
GEO can not
canbe require
satellites drawn
be drawn beam
are from from
stationary re-alignment.
Fig. 3,relative
Fig. where
3, where weto we the
c ≤ 6 GHz, W = 800 MHz
6InIn
< Fig.
Fig.
f 2 2we
60we GHz,
plot
plot the and
the Shannon
6 < fc ≤ 60 GHz, and W = 2 GHz for fc > 60 GHz.
c Shannon
W = 2 GHz
capacity
capacity forC, fwhich
C, c which
> 60 GHz.
represents
represents a a observeSimilar
observe that that conclusions
the system
the system can
performance be
performance
Earth’s surface and do not require beam re-alignment. drawn
decreases from
decreases Fig.
at low
at 3, where
elevation
low elevation we
InInFig.
commonly
commonly Fig.2accepted
2we weplotplotthe
accepted the
metric
metricShannon
Shannon totofacilitatecapacity
facilitate
capacity accurate
accurate which
C,C,which represents aa duedue
benchmarking
benchmarking
represents observe
to tothethat
Similar themore the
more system
severe
conclusions severe performance
impact
can impact of of
be drawn decreases
scintillation
scintillation
from Fig. atabsorption
low elevation
3, absorption
where we
ofof wireless
commonly
wireless networks,
accepted
networks, as a
metric
as afunction
to of
facilitate
function ofh, h,f and
accurate
f and G . First,
benchmarking
G . First,we we (which
due
(which is
to caused
the
is more
caused by sudden
severe
by sudden changes
impactchanges ofin the
in refractive
scintillation
the refractive index
absorption index
commonly accepted metric to facilitate accurate benchmarking observe that the system performance decreases at low elevation
c c rx rx
observe
of ofwireless
wireless
observe that
that satellite
satelliteoperations
networks,
networks, asaafunction
operations
as function ininthethe
ofofh, bandwidth-constrained
andG Grx.. First,
h,bandwidth-constrained
ffccand we due
First, we due
due to totheis
(which to the thevariation
caused
variation
more byofsudden
severe temperature,
of changes
temperature,
impact ofwaterinwaterthevapor
scintillation vapor content,
refractive index
content,
absorption
rx
below-6
observethat GHz
that spectrum
satellite offer
operations limited capacity (i.e., andduebarometric
to the pressure),
variation ofas the signal
temperature, has to
water transit
vapor longer
content,
below-6 GHz spectrum offer limitedininthe thebandwidth-constrained
bandwidth-constrained
capacity (i.e., < 500 (which and barometricis causedpressure), as changes
the signalinhas the torefractive
transit longer
< 500 by sudden index
observe satellite operations
Mbps),
below-6
Mbps), which GHz
which mightspectrum
might bebeinsufficient
offer limited
insufficient toto satisfy
satisfy thethe
capacity mostmost demand-
(i.e., < 500 through
demand- through
due to
thethe
and barometric
the
atmosphere.
atmosphere.
variation
Moreover,
pressure),
of as theFig.
Moreover,
temperature, signal3 exemplifies
Fig. 3has to transit
exemplifies
water vapor
thatthat the
longer
content, the
below-6 GHz spectrum offer limited capacity (i.e., < 500
ing
ing beyond-5G
Mbps), which might
beyond-5G useuse cases.
cases. The
be insufficientThe performanceto satisfycan
performance the
can bemost
be demand- increased
improved
improved throughprobability
increased
and
theprobability
barometric
atmosphere. of of
pressure),
path pathblockage
Moreover,
as blockage
the signal
in the
Fig. in has
urban
3 exemplifies
the to urban scenario
transit
that the
scenario
longer
Mbps), which might be insufficient to satisfy the most demand- may reduce the achievable capacity by more than 60% at high
by ingconsidering
beyond-5G mmWave
use cases. transmissions,
The performance thanks to the
can be massive
improved increased
may reduce probability
the achievable of path blockage
capacity by in thethan urban60%scenario
ingby considering
beyond-5G mmWave
use cases. transmissions,
The performance thanks cantobethat the massive through
improved elevation,
the atmosphere.
compared to atorural
Moreover,
scenario.
Fig.more 3 exemplifies at high
that the
bandwidth
by considering
bandwidth available
availablemmWave atathigher
higher frequencies,
transmissions,
frequencies, provided
thanks
provided to thethat high-
massive
high- may reduce
elevation, the achievable
compared a capacity
rural scenario.by more than 60% at high
by considering mmWave transmissions, thanks to the massive increased probability of path blockage in the urban scenario
gain directional
bandwidth
gain directional antennas
available antennasat higher(i.e., G >>5050dB,
frequencies,
(i.e.,frequencies, asasis is
provided
dB, typical in in Despite
that high-
typical elevation,
Despite these thesepromising
compared to aresults,
promising rural
results,better
scenario.
betterwireless
wireless coverage can can
G thancoverage
rxrx
bandwidth
current available
satellite antenna at higher
technologies) areare provided that high- may reduce the aachievable capacity by is more 60% at high
gain directional
current satellite antennas
antenna (i.e., Grx >
technologies) 50employed,
dB, as istotypical
employed, cover
to coverin be be
provided
Despite
provided when
these
when standalone
promising
a standalone space
results,
space layer
betterlayer assisted
wireless
is assisted by HAPs
coverageby HAPs can
gain
very directional
long antennas
transmission (i.e.,
distances. G Fig. > 50
2 arealsodB, as
makes is typical
thethe in elevation, compared to a rural scenario.
current
very long satellite
transmissionantenna technologies)
distances.
rx
Fig. 2 employed,
also makes tocase
cover
case operating
be provided
operating in in thewhenstratosphere,
the a standalone
stratosphere, as already
space
as already discussed
layer in Sec.
is assisted
discussed in by
Sec.III.
HAPs III.
current
that satellite antenna technologies) are employed, to cover A performance
Despite these promisingbetween results, better wirelessLEO coverage can
veryfurther
that further increasing
increasingfcdistances.
long transmission fcbeyond
beyond70 70GHz
Fig. 2GHzalsowould makes
would decrease
the case A
decrease operating
performance in thecomparison
stratosphere,between
comparison as alreadya standalone
a discussed inLEO
standalone sce-sce-
Sec. III.
very
the long transmission distances. Fig. 2 also makes the case be provided when a standalone space layer is assisted by HAPs
thatShannon
the further capacity
Shannon capacitydue
increasing due fctoto the
beyondtheincreasingly
70 GHz would
increasingly harshharshimpact impact nario
decrease (LEO-GND)
A performance
nario (LEO-GND) andand a multi-layered
comparison between scenario
a multi-layered a standalone
scenario (LEO-HAP- LEO sce-
(LEO-HAP-
that
ofof
the further
atmospheric
Shannon increasing
absorption
capacity fduecinbeyond
the
to higher
the 70 GHz would
mmWave
increasingly spectrum.
harsh decrease
impact operating
GND) in in thea stratosphere,
which HAP bridges asthealreadyLEO discussed
communications in Sec. III.
atmospheric absorption in the higher mmWave spectrum. nario (LEO-GND) and a multi-layered
GND) in which a HAP bridges the LEO communications scenario (LEO-HAP-
the ofAsShannon
expected, capacity due in to the the higher
increasingly harsh impact A performance comparison between aLEO standalone LEO the sce-
As expected,Cabsorption
atmospheric Cseverely
severelyreduces reduces forforincreasing
increasingvalues
mmWave valuesof of towards
spectrum. GND)
towards theinthe
ground
which
ground isaplotted
isHAP
plotted in Fig.
bridges
in Fig.4. the
It4.appears
It appears clear that
communications
clear that the
of
h,h,atmospheric
i.e.,
As transitioning absorptionfrom in
LEO the to higher
GEO mmWave
satellites. spectrum.
Neverthe- nario
intermediate (LEO-GND)HAP offersand a multi-layered
improved capacity scenario
by amplifying (LEO-HAP- the the
i.e.,expected,
transitioning C severely
from LEO reduces to GEO for increasing
satellites. Neverthe- values of intermediate
towards the ground HAP offersis plottedimprovedin Fig.capacity
4. It appears clear that
by amplifying
h,Asi.e.,expected,
transitioningC severely from LEO reduces to GEO for increasing
satellites. values Neverthe- of GND)intermediate in which HAPaoffers HAPimprovedbridges capacitythe LEObycommunications amplifying the
 TABLE II: Open challenges for non-terrestrial networks.

Open challenge Explanation

Channel Modeling Missing adequate characterization of mmWave second order statistics, Doppler, fading, multipath
Spectrum co-existence Spectrum sharing is required to provide isolation among different non-terrestrial services
Design of flexible numerology to compensate for large Doppler shift
Non-linear payload distortions may complicate signal reception
PHY procedures
Large RTTs increase the response time for ACM scheme
Large RTTs make it infeasible to operate in TDD
HARQ Large RTTs may exceed the maximum possible number of HARQ processes
Synchronization Large non-terrestrial station’s footprint creates a differential propagation delay among users in the cell
Initial access Channel dynamics may result in obsolete channel estimates
Mobility management Directionality complicates user tracking, handover, and radio link failure recovery
Non-terrestrial stations may need to serve a very large number of users
Constellation of non-terrestrial stations is necessary to maintain ubiquitous service continuity
Constellation management
High cost of satellite launches complicates deployment of dense constellations
Wireless coordination among air/spaceborne vehicles complicates constellation management
Channel dynamics result in obsolete topology information
Higher-layer design Large RTTs result in longer duration of the slow start phase of TCP
Channel dynamics result in decreased resource utilization due to sudden drops in the link quality
Unclear where to distribute SDN planes
Architecture technologies Long RTTs prevent long duration of batteries
Design of central authority making secure network/communication decisions

towards the ground is plotted in Fig. 4. It appears clear that the PHY procedures. In the non-terrestrial case even the highest
intermediate HAP offers improved capacity by amplifying the available sub-carrier spacing in the frame structure may not
signal from the upstream satellite before forwarding it to the be enough to compensate for the large Doppler experienced
ground, while ensuring a quicker deployment and lower costs considering the high speed of aerial/space stations. Moreover,
compared to spaceborne stations. The benefits are particularly the large propagation delays in NTNs may create a larger
evident when h = 1, 200 km and α = 10°, i.e., when the response time for the Adaptive Modulation and Coding (AMC)
resulting longer propagation distance from the LEO satellite scheme loop and requires a margin to compensate for the
may deteriorate the signal quality below detectable levels on possible outdated control signals exchanged during channel
the ground, with a performance boost of +250%. estimation. Notably, in an integrated terrestrial/non-terrestrial
framework, different network elements on the end-to-end com-
V. N ON -T ERRESTRIAL N ETWORKS : munication path may process the information at different rates,
O PEN C HALLENGES thus contributing to the overall communication delay. Addi-
Despite current standardization efforts towards the develop- tionally, Time Division Duplexing (TDD), which is frequently
ment of NTNs, there remain several open issues for proper considered in terrestrial networks, may be infeasible in non-
protocol design which call for long-term research, as high- terrestrial networks since guard times must be proportional to
lighted below and summarized in Table II. the propagation delay.
Channel modeling. Even though the 3GPP has specified how HARQ. The long Round Trip Time (RTT) experienced in
to characterize mmWave propagation for the satellite chan- non-terrestrial networks may exceed the maximum possible
nel [9], it is currently not investigating second order statistics number of Hybrid Automatic Repeat reQuest (HARQ) pro-
(including therefore correlation in both space and time), nor cesses that are typically supported in 5G NR systems. In this
the impact of Doppler, fading, and multipath components, regard, simply increasing the number of processes may not be
which is critical at high frequencies. Moreover, a general and feasible due to memory restrictions at the mobile terminal’s
accurate model of a fully-layered space-air-ground channel is side. Long RTTs also require large transmission buffers, and
still lacking. potentially limit the number of retransmissions allowed for
Spectrum co-existence. As non-terrestrial systems move each transmission.
into the mmWave bands, where other systems have been Synchronization. Non-terrestrial systems are fast-moving,
operating for many years (e.g., satellites offering weather and typically feature larger cells compared to terrestrial net-
forecasting services), consideration needs to be given to the works. At low elevation angles, this may create a very large
co-existence among different networks. The main challenge is differential propagation delay between users at the cell edge
the development of flexible spectrum sharing techniques that and those at the center (up to 10 ms for GEO satellites [9]),
maintain adequate isolation among different communications thereby raising synchronization issues.
while ensuring reasonable licensing costs. Initial access and channel estimation. Initial access makes
on-the-ground terminals establish a physical connection with Furthermore, due to the large distances involved in non-
a non-terrestrial station by detecting synchronization signals. terrestrial operations and the resulting severe path loss expe-
This is particularly challenging in non-terrestrial applications, rienced, the transmit power is typically to be set as close as
where the channel may vary quickly over time, as the initial possible to the saturation point. This could reduce the duration
estimate may rapidly become obsolete. Also, in space-ground of batteries, which is particularly critical in scenarios where
integrated networks, each intermediate node tends to associate aerial devices are used to support IoT applications.
to a gateway based on its own unilateral benefit, neglecting the Finally, an integrated space-air-ground architecture should
potential disadvantages on the whole network performance. envision the existence of a trusted central authority making
secure network topology and communication decisions to
Mobility management. When operating at mmWaves to
prevent malicious nodes from being selected as a gateway.
maintain high-capacity connections, directionality is required
in order to achieve sufficient link budget. In this case, fine
alignment of the beams has severe implications for the design VI. C ONCLUSIONS
of control operations, e.g., user tracking, handover, and radio Non-terrestrial networks are being investigated as a key
link failure recovery. These challenges are particularly critical component of the 6G framework to support global, ubiqui-
in the non-terrestrial domain, where the very high speed of tous and continuous connectivity, and to overcome the cov-
aerial/space platforms could result in loss of beam alignment erage limitations of envisioned 5G networks. In this paper
before a data exchange is completed. The increased Doppler we overviewed recent advancements that will make non-
encountered at high speed could also make the channel non terrestrial networks a reality, including the development of
reciprocal, thus impairing the feedback over a broadcast chan- new aerial/space architectures, and innovative spectrum and
nel. antenna technologies. As a case study, we demonstrated that
Constellation management. A non-terrestrial station has a the mmWave frequencies can be used to establish high-
larger footprint than a terrestrial cell and is required to serve capacity connections between on-the-ground terminals and
a larger number of on-the-ground terminals. This may result in satellite/HAP gateways, provided that sharp beams are formed.
saturation of the available bandwidth, with strong implications Despite such promises, we also summarized current open
for latency and throughput performance. challenges for the deployment of non-terrestrial networks,
thereby stimulating further research in this domain. Most
Additionally, air/spaceborne vehicles move rapidly relative
importantly, our future studies will be dedicated to exploring
to the Earth’s surface and may create regions where coverage
the relationship between capacity performance and energy
is not continuously provided. A constellation is thus necessary
efficiency in the non-terrestrial ecosystem.
to maintain ubiquitous service continuity. When configuring
multiple satellites that move in different orbits to operate
in an integrated fashion, however, constellation management ACKNOWLEDGMENTS
is hindered by handovers and load balancing among the Part of this work was supported by the US Army Research
different layers. Office under Grant no. W911NF1910232 “Towards Intelligent
Moreover, while in the terrestrial scenario coordination Tactical Ad hoc Networks (TITAN)”.
between base stations is possible through fiber connections
or via a central entity, coordination among non-terrestrial R EFERENCES
air/spaceborne stations has to be implemented wirelessly, thus
further complicating constellation management. [1] M. Giordani, M. Polese, M. Mezzavilla, S. Rangan, and M. Zorzi,
“Toward 6G Networks: Use Cases and Technologies,” IEEE Commu-
Higher-layer design. Current network/transport protocols nications Magazine, vol. 58, no. 3, pp. 55–61, March 2020.
[2] E. Yaacoub and M. Alouini, “A Key 6G Challenge and Opportunity –
may show low performance when NTNs are involved. First, Connecting the Base of the Pyramid: A Survey on Rural Connectivity,”
topology information may quickly become obsolete (especially Proceedings of the IEEE, vol. 108, no. 4, pp. 533–582, Mar. 2020.
considering unpredictable mobility, e.g., for UAV swarms) [3] M. Polese, M. Giordani, T. Zugno, A. Roy, S. Goyal, D. Castor, and
M. Zorzi, “Integrated Access and Backhaul in 5G mmWave Networks:
and must constantly be refreshed, thus increasing the com- Potential and Challenges,” IEEE Communications Magazine, vol. 58,
munication overhead. Second, a large RTT results in a longer no. 3, pp. 62–68, March 2020.
duration of the slow start phase of TCP, during which the [4] M. Mozaffari, A. T. Z. Kasgari, W. Saad, M. Bennis, and M. Debbah,
“Beyond 5G with UAVs: Foundations of a 3D wireless cellular network,”
sender may take inordinately long before operating at full IEEE Transactions on Wireless Communications, vol. 18, no. 1, pp. 357–
bandwidth. Third, sudden drops in the link quality, which may 372, Nov 2018.
be common in NTNs, make the sender reduce its transmission [5] M. Giordani and M. Zorzi, “Satellite Communication at Millimeter
Waves: a Key Enabler of the 6G Era,” IEEE International Conference
rate, thus leading to a drastic decrease in resource utilization. on Computing, Networking and Communications (ICNC), Feb 2020.
Finally, when a multi-layered integrated network is considered, [6] F. Babich, M. Comisso, A. Cuttin, M. Marchese, and F. Patrone,
different network devices may support different (and some- “Nanosatellite-5G Integration in the Millimeter Wave Domain: A
Full Top-Down Approach,” IEEE Transactions on Mobile Computing,
times conflicting) communication protocols, thus complicating vol. 19, no. 2, pp. 390–404, Feb 2020.
network management. [7] M. Boschiero, M. Giordani, M. Polese, and M. Zorzi, “Coverage
Analysis of UAVs in Millimeter Wave Networks: A Stochastic Geometry
Architecture technologies. It is still unclear where to dis- Approach,” 16th IEEE International conference on Wireless Communi-
tribute SDN planes for proper service delivery; the choice cations & Mobile Computing (IWCMC), 2020.
[8] J. Liu, Y. Shi, Z. M. Fadlullah, and N. Kato, “Space-Air-Ground Inte-
depends on different factors, like the available processing grated Network: A Survey,” IEEE Communications Surveys Tutorials,
capabilities or the achievable transmission rate. vol. 20, no. 4, pp. 2714–2741, May 2018.
 [9] 3GPP, “Solutions for NR to support Non-Terrestrial Networks (NTN),”
TR 38.821 (Release 16), 2020.
[10] F. Zhou, R. Q. Hu, Z. Li, and Y. Wang, “Mobile edge computing in
unmanned aerial vehicle networks,” IEEE Wireless Communications,
vol. 27, no. 1, pp. 140–146, Jan 2020.
[11] J.-L. Muraro, G. Nicolas, S. Forestier, S. Rochette, O. Vendier, D. Lan-
grez, J.-L. Cazaux, M. Feudale et al., “GaN for space application:
Almost ready for flight,” International Journal of Microwave and
Wireless Technologies, vol. 2, no. 1, pp. 121–133, Apr 2010.
[12] A. Abdelsalam, A. Bujari, M. Luglio, D. Munaretto, C. E. Palazzi,
M. Quadrini, S. P. Romano, C. Roseti, and F. Zampognaro, “Implemen-
tation of Virtualised Network Functions (VNFs) for Broadband Satellite
Networks,” in European Conference on Networks and Communications
(EuCNC), June 2019.
[13] M. De Sanctis, E. Cianca, T. Rossi, C. Sacchi, L. Mucchi, and R. Prasad,
“Waveform design solutions for EHF broadband satellite communica-
tions,” IEEE Communications Magazine, vol. 53, no. 3, pp. 18–23, Mar
2015.
[14] H. Kaushal and G. Kaddoum, “Optical Communication in Space:
Challenges and Mitigation Techniques,” IEEE Communications Surveys
& Tutorials, vol. 19, no. 1, pp. 57–96, Firstquarter 2017.
[15] Y. Rahmat-Samii and A. C. Densmore, “Technology Trends and Chal-
lenges of Antennas for Satellite Communication Systems,” IEEE Trans-
actions on Antennas and Propagation, vol. 63, no. 4, pp. 1191–1204,
April 2015.

Marco Giordani [M’20] received his Ph.D. in Information Engineering in

2020 from the University of Padova, Italy, where he is now a postdoctoral
researcher and adjunct professor. He visited NYU and TOYOTA Infotechnol-
ogy Center, Inc., USA. In 2018 he received the “Daniel E. Noble Fellowship
Award” from the IEEE Vehicular Technology Society. His research focuses
on protocol design for 5G/6G mmWave cellular and vehicular networks.

Michele Zorzi [F’07] is with the Information Engineering Department of the

University of Padova, focusing on wireless communications research. He was
Editor-in-Chief of IEEE Wireless Communications from 2003 to 2005, IEEE
Transactions on Communications from 2008 to 2011, and IEEE Transactions
on Cognitive Communications and Networking from 2014 to 2018. He served
ComSoc as a Member-at-Large of the Board of Governors from 2009 to 2011,
as Director of Education and Training from 2014 to 2015, and as Director of
Journals from 2020 to 2021.