Open issues and beyond 5G

Marco Chiani, Enrico Paolini, Franco Callegati

Abstract
Fifth generation mobile networks (5G) will change our life and society, thanks to
the introduction of recently developed technologies which will enable the develop-
ment of new services and applications. While 5G is appearing with her novelties,
we need to start thinking about the next steps that wireless cellular systems have
to tackle, always keeping in mind that the final goal is the benefit of humanity. In
this chapter we discuss some issues that 5G will still leave open, and the possible
evolution towards the next generation (6G) of wireless communication systems.

1 Introduction

The last two decades have witnessed an extremely fast evolution of mobile cellular
network technologies, from 1G to 4G, with 5G networks expected to be operational by
2020. This chapter will briefly review the path from the first to the latest generation of
mobile cellular systems, will discuss some issues not fully addressed in 5G systems,
and finally will sketch a vision of what we may expect beyond 5G. For the path
beyond 5G, we envision that the mobile network will become more intelligent,
with learning mechanisms to modify itself based on users’ experience; situation-
awareness will lead decision making and networking; this will allow fast and flexible
spectrum reallocation, with consequent large bitrates available to the users; other
human senses will be communicated, and 3D / holographic type communication
will improve the quality of the tele-interaction; users will not necessarily need to
bring a smartphone but will benefit of wireless-devices-as-a-service, with distributed
devices available to anyone; the devices battery life will be substantially extended.
Among the technologies to reach these goals we count machine learning, dynamic
spectrum allocation, wireless energy transfer, free-space optical communication, use
of bands beyond 100 GHz, massive use of multiple antenna systems, new access

CNIT/University of Bologna, e-mail: {marco.chiani, e.paolini, franco.callegati}@unibo.it

1
2 Marco Chiani, Enrico Paolini, Franco Callegati

Fig. 1 The network will learn from experience.

schemes allowing an increase in node density, cybersecurity for quantum attacks,

high accuracy indoor localization, massive virtualization of network functionalities
and novel, software driven, network control architectures.

2 The evolution from 1G to 5G

Table 1 provides an overview of the main features of each cellular network generation,
from the first generation to the fifth one, including regulation, services, innovation
with respect to the previous generation, and some ancillary information.
Concerning regulation, we witnessed a profound paradigm shift across the several
generations, from 1G state-owned monopoly operators, very often obtaining the
use of spectrum free of charge, to open-market auctions starting with 3G, to 5G
spectrum sharing. Concerning open-market auctions, 3G systems spectrum was
licensed at extremely high prices, with up-front payments years before the first
universal mobile telecommunications system (UMTS) terminal appeared on the
Open issues and beyond 5G 3

market. This regulation model contributed to late (and in some cases limited) profits
seen by operators and, consequently, to a limited availability to excessive up-front
payments for 4G spectrum licenses. The possibility of extending spectrum licenses
up to 25 years, to apply spectrum sharing mechanisms, and to benefit from some
regulatory flexibility in new millimeter wave (mmWave) frequency bands, are new
regulatory aspects characterizing 5G.
In terms of services, the initial basic voice-only calls featured by 1G mobile sys-
tems evolved into a multitude of different services with the subsequent generations,
from simple text messaging and basic high-latency data exchange to high-quality
video streaming and chatting services, to radically new services supported by 5G
networks. Services evolution was enabled by several factors, including ever-rising
supported bit rates, advances in air interface design, signal processing at physi-
cal layer, and MAC layer procedures, technological advances in mobile terminals
manufacturing, evolution of mobile internet protocols, cloud computing, advanced
networking control paradigms.
Second generation global system for mobile communications (GSM) cellular
networks initially provided digital voice service at bit rate 9.6 kbps. General Packet
Radio Service (GPRS) and ultimately Enhanced Data Rates for GSM Evolution
(EDGE) data services were subsequently introduced, with bit rates of a few tens
of kbps and up to 200 kbps, respectively. Push email was also introduced for the
first time on Blackberry devices. These bit rates were largely increased in next
generations. Third generation UMTS offered up to 2 Mbps bit rate (often 364 kbps)
initially, and then several tens of Mbps in downlink with High Speed Packet Access
(HSPA). Fourth-generation LTE features up to 300 Mbps in downlink, with a target
of 1 Gbps, and up to 50 Mbps in uplink. Fifth generation cellular networks are
expected to increase the bit rate significantly, up to 20 Gbps. These bit rates, end-
to-end latencies down to 1 ms, ultra reliability (packet error rate 10−5 or less), and
massive multiple access, will foster services such as enhanced mobile broadband,
device-to-device (D2D) communication, ultra-reliable and low-latency Internet of
Things (IoT) and machine-type communication (MTC), e-health, augmented reality
and tactile Internet, industrial control for the Industry 4.0, automated driving and
flying. D2D communication, consisting of establishing a direct link between nearby
devices without relaying information through a base station (BS), is emerging as a
key technology to achieve efficient resource allocation, higher spectral efficiency,
reduced latency.
Each generation of wireless cellular networks came with its own technical inno-
vations, both on the network, air interface, and user terminal side.
Multiple access. Multiple access schemes were constantly enhanced, from sin-
gle carrier per channel frequency division multiple access (FDMA), to frequency-
and-time division multiple access (FDMA/TDMA), code division multiple ac-
cess (CDMA), up to orthogonal frequency division multiple access (OFDMA) and
non-orthogonal multiple access (NOMA), with a progressive explosion of the net-
work capacity in terms of the number of users served at the same time with an
adequate quality of service (QoS).
4 Marco Chiani, Enrico Paolini, Franco Callegati

PHY layer enhancements. Fundamental innovations at the physical layer con-

tributed to the above-reviewed ever-increasing bit rates. Among them we can count
multiple-input multiple-output (MIMO), evolving into massive MIMO in 5G, as well
as advanced channel coding schemes such as 3G/4G turbo codes and 5G New Radio
traffic channel low-density parity-check (LDPC) codes and control channel polar
codes.
Frequency bands New frequency bands were exploited by each generation. First
generation advanced mobile phone system (AMPS) and total access communication
system (TACS) were operated in the 800 MHz and 900 MHz bands, respectively,
while second generation GSM was initially operated in the 900 MHz band, and
then also in the 1800 MHz and 1900 MHz (in North America) ones. Frequency
bands around 2 GHz were for the first time used by UMTS networks, while a
number of frequency bands are available worldwide for long term evolution (LTE),
based on regulatory aspects in different geographical areas (e.g., 450/800/900 MHz,
1800/2100 MHz, 2600 MHz in Europe). From a spectrum allocation viewpoint, the
main breakthrough introduced by 5G is the use of licensed, shared, and unlicensed
frequency bands in the mmWave band, above 24 GHz. The one around 60 GHz is
of particular interest for indoor very-high data rate applications, wireless backhaul,
and femtocell implementation.
Evolution of switching and networking The 1G network was fully circuit based,
resembling the traditional telephone network, a choice motivated by being voice
calls the target service of the networks. In 2G and 3G systems data oriented services
progressively became more important and a packet switched network dedicated
to the transport of data was set aside the existing circuit switched network. From
4G the evolved packet core (EPC) concept is adopted meaning that the transport
network is based on packet switching only and that such packet switching is IP
based. Therefore the transport technology of the mobile networks converges with
the transport technology of the Internet. The EPC definition brings along another
concept fundamental for the evolution to the 5G: the separation of the control and
data plane. The technologies to control the transport of the user data are logically
separated from the transport itself, which is essential to allow a separate evolution
path to guarantee the required network scalability.
Network softwarization The emergence of the cloud computing paradigm and
of the related virtualization technologies brought forward a key innovation in the
5G. Complex networking functionalities, traditionally requiring dedicated hardware
and management may be virtualized as pieces of software into the cloud. This is the
network function virtualization (NFV) paradigm, defined and standardized by ETSI
in the recent years. Meanwhile software defined networking (SDN) also emerged,
a novel approach that allows a very effective separation of the network control and
data plane, thus further extending the ideas behind the EPC. The OpenFlow protocol
is the key component of SDN, supporting the implementation of a communication
channel between the controller (the brain of the network) and the network nodes
(that carry data). NFV exploits virtualization, cloud computing and SDN to define
an architecture which supports the implementation of network services as subsets of
software functionalities, with an unprecedented degree of flexibility and adaptation
Open issues and beyond 5G 5

capabilities. Network programmability and split of control and data plane are key
enabling technologies towards a full implementation of the 5G. NFV and SDN fulfill
such expectation and are now getting into a state of advanced validation.

3 5G Issues

This section points out some issues and challenges that, at present, have not been
fully addressed in 5G mobile networks and that are attracting further research efforts.
Coverage issues. With the advent of 5G we will witness a site densification pro-
cess as a primary means to increase network capacity. Site densification necessarily
poses economical issues that may slow down considerably spatial and temporal 5G
deployment, unless a substantial BS cost reduction is achieved over the time. It has
been reported how SDN and network virtualization may contribute to cut costs,
but there remains uncertainty about to what extent this will speed up 5G rollout.
Recent studies have shown that, under a business-as-usual model, in UK 90% of
the population will be covered with 5G not before 2027 and the that 100% coverage
will be extremely hard to reach due to prohibitively increasing deployment costs in
less populated areas [7]. Similar expectations have been reported for other countries
during discussions at the 2018 IEEE 5G World Forum.
Emerging applications challenges. Super-hype of 5G contributed to create un-
precedented expectations about the levels of QoS these mobile networks will be
able to provide. It is becoming clear, however, how some emerging applications will
push the required QoS to extreme levels that appear very challenging for currently
envisioned 5G architectures. Among them, driverless cars and vehicles and the tac-
tile Internet. For example, the set of requirements imposed by the tactile Internet
(end-to-end ultra-low latency not exceeding 1 ms, outage probability 10−7 or less,
network intelligence to support predictive actuation) to deliver actuation and senses
such as hearing, touching, and seeing, is still considered a 5G challenge especially
over long ranges beyond 100 km [9].
D2D challenges and vulnerabilities. There are aspects of D2D communication
in 5G that have not yet been addressed in a totally satisfactory manner. One is coex-
istence of cellular users (CUs) and D2D pairs, particularly, mitigation of D2D links
interference on CUs, when downlink resources are shared with D2D devices, and
also on the BS when uplink resources are shared. D2D interference management is
still a subject of research efforts, since existing interference mitigation techniques
(interference cancellation based on coding and signal processing, interference avoid-
ance based on orthogonal resource allocation, and interference coordination based
on scheduling and power control) are expected to be insufficient in ultra-dense node
deployment scenarios. Other issues concern D2D security and privacy. Direct or
relay-assisted communication may be established on device controlled links with
no control of the core network, a trusted party providing identification, authentica-
tion, and encryption. This makes D2D links potentially more vulnerable to privacy
violation, besides suffering from all of the attacks affecting other networks.
6 Marco Chiani, Enrico Paolini, Franco Callegati

Mobile edge computing issues. It is foreseen that moving computing, storage,

and networking resources to the edge of the radio access network (RAN) will be
a key ingredient to alleviate backhaul and core network and to allow executing
delay-sensitive and context-aware applications in the proximity of end users. This
paradigm, referred to as mobile edge computing (MEC), poses however concerns
[10]. Among them, the limited computing and storage resources per each MEC
platform, the necessity for MEC platforms of different provider to collaborate, chal-
lenges in user mobility support in small cells, problematic applicability of centralized
authentication protocols.
An open and smart RAN? We are currently witnessing efforts towards open,
interoperable interfaces, RAN virtualization, and RAN intelligence.There is a trend
to incentivize the use of commercial off-the-shelf (COTS) hardware, to explore open
source and open whitebox network by introducing virtualized network elements
with standardized open interfaces, to push embedded machine learning systems and
artificial intelligence back end modules for an enhanced network intelligence. These
are the main objectives of the ORAN Alliance. This trend is taking the first steps and
it is clear to what extent the envisioned features will be implemented in 5G networks.
Network orchestration and slicing. The NFV-SDN technologies, as briefly men-
tioned above, promise to support the implementation of a large variety of services
by means of softwarized functionalities hosted into cloud computing platform over
data centers implemented with standard hardware (COTS). In terms of Capex and
Opex this is a real revolution for network operators, that can move into the direc-
tion of more effective procurement and simplified operations. Moreover the same
resources (both and hardware and software) may be shared among different subsets
of customers, while keeping full isolation of the data paths and of the quality of
service thanks to the native capabilities of cloud computing platforms. This is the
"slicing" concepts, different subsets of customers may subscribe a service contract
with different operators (either real or virtual) and share in the end the same infras-
tructure, which paves the way to novel business models and opportunities for the
network providers. An effective management of NFV requires an orchestration plat-
form that will automate the deployment of the required functionalities and manage
operations. Some opensource implementations of such an orchestration platform are
currently under implementation, for instance Open source MANO (OSM)1, and the
open network automation platform (ONAP)2, and running proof of concepts are also
available such as the central office re-engineered as a data-center (CORD)3. These
technologies are not at the production level and some issues like multi-domain or-
chestration support are still open. Similarly, the slicing paradigm has still to be fully
understood and experimented to validate some critical aspects such as full quality of
service isolation and security guarantees.

1 https://osm.etsi.org
2 https://www.onap.org
3 https://opencord.org
Open issues and beyond 5G 7

4 A Vision of 6G

Some possible facts which will arise are summarized below.

More and more data. As of today (2018), about one million Terabytes (i.e.,
1 Exabyte = 1018 bytes) of data per day are exchanged over the mobile networks
all over the world (see Fig. 2). The amount of data exchanged by mobile users will
continue to increase, on one side due to the increasing number of non-human devices
connected (including vehicles, UAVs, and autonomous systems), and to the enhanced
quality 3D video / holographic type communication that will be used by humans.
Network Intelligence. To deal with increased traffic, the mobile network will
become more intelligent, with learning mechanisms to autonomously modify itself
based on users’ experience, and situation-aware decision making and networking.
Fast spectrum reallocation. Network intelligence will be used to allow fast and
flexible spectrum allocation/reallocation, with consequent large bitrates available to
the users.
Enhanced senses. Other human senses will be communicated to improve the
quality of the tele-interaction, including 3D / holographic type communication,
taste, smell, touch.
Wireless-devices-as-a-service. Users will not necessarily need to bring a smart-
phone but will benefit of wireless-devices-as-a-service, with distributed devices
available to anyone. All information being in the cloud, users will just need to be
authenticated and then access the network by using any available device.
Battery duration & energy. The need to put devices on recharge will be dramat-
ically reduced, so that the battery life will be substantially extended.
Quantum computers & quantum networks. We will see the appearance of
quantum computers, capable to solve problems that are not solvable with non-
quantum computers. Also, quantum communication and networks will be available,
e.g., for cryptographic key exchange, also from satellites.
Privacy, security, data manipulation. The need for privacy and security for a
proper management of personal data will be of paramount importance. The appear-
ance of quantum computers will force to re-think the cryptography and security
mechanisms.
Security and safety. The IoT and the Industry 4.0 will bring the network very
close to the real infrastructures. Therefore a security breach on the networking side
may quickly become a very important safety issue in the real life. This already
happened with the failed attack to a petrochemical plant in Saudi Arabia in August
2017, that is widely believed was aimed to make the plant explode by attacking the
IT control infrastructure.
Virtual Operators explosion. The NFV-SDN technologies and the related slic-
ing capabilites will boost the emergence of virtual operators that will compete by
providing added values services to the users by deploying innovative network func-
tionalities in the cloud. This may lead to a significant innovation also in the business
models and commercial strategies in the field.
8 Marco Chiani, Enrico Paolini, Franco Callegati

5 Technologies

Among the technologies in the evolution path, we cite the following.

Machine Learning & AI. The complexity of the network and the number of
connected devices will lead to a network which will learn from the experience to
modify itself and accomodate new services. A suitable policy should be investigated
to avoid the “Skynet” of the movie “The Terminator”.
Dynamic spectrum allocation. The precious radio spectrum in the lower bands
will be used more effectively by allocating the frequencies every second or so, based
on the context.
Wireless energy transfer. This could be in some situations a viable way to extend
the battery life, avoiding frequent recharges.
Free-space optical communication. Optical free-space communication will al-
low high data rate, for both outdoor and indoor scenarios, so releasing the lower
band radio spectrum for long range uses.
Sub-Terahertz and Terahertz communications. The availability of large bands
beyond 100 GHz (D-band, 110 GHz to 170 GHz,...) will lead to transmission systems
at high data rate over short distances, so releasing the lower band radio spectrum for
long range uses (see Fig. 3).
Massive use of multiple antenna systems. The increase in the frequency will
require multiple antenna systems able to exploit the multi-rays propagation, for larger
throughput but also for precise localization and for energy transfer.
High accuracy indoor localization. Context awareness needs a precise user’s
localization. For this item see the chapter about localization in this book.
Access schemes for massive wireless networks. New access schemes will
be needed to handle a massive number of non-orthogonal users (more than 10
devices/m2 ) in an efficient and scalable way.
Cybersecurity. The possibility to use non-personal devices for personal commu-
nication will impose new challenges on biometric authentication and privacy. New
schemes for wireless security, capable to deal with attacks of quantum computers,
will be needed.

References

1. IEEE Future Networks Initiative. URL: https://futurenetworks.ieee.org/.

2. Ericsson Mobility Report, Q2, 2018. URL: https://www.ericsson.com/en/mobility-report.
3. IEEE 5G World Forum, July 2018. URL: http://ieee-wf-5g.org/past-events/2018-2/.
4. ITU Focus Group on Technologies for Network 2030. URL: https://www.itu.int/en/ITU-
T/focusgroups/net2030.
5. K. David and H. Berndt. 6G Vision and Requirements: Is There Any Need for Beyond 5G?
IEEE Vehicular Technology Magazine, 13(3):72–80, 2018.
6. S. Kim, W. T. Khan, A. Zajić, and J. Papapolymerou. D-Band Channel Measurements and
Characterization for Indoor Applications. IEEE Transactions on Antennas and Propagation,
63(7):3198–3207, July 2015.
Open issues and beyond 5G 9

7. E. J. Oughton and Z. Frias. The cost, coverage and rollout implications of 5G infrastructure in
Britain. Telecommunications Policy, 42(8):636–652, 2018.
8. P. H Pathak, X. Feng, P. Hu, and P. Mohapatra. Visible light communication, networking,
and sensing: A survey, potential and challenges. IEEE communications surveys & tutorials,
17(4):2047–2077, 2015.
9. M. Simsek, A. Aijaz, M. Dohler, J. Sachs, and G. Fettweis. 5G-enabled tactile Internet. IEEE
J. Sel. Areas Commun., 34(3):460–473, March 2016.
10. T. X. Tran, A. Hajisami, P. Pandey, and D. Pompili. Collaborative mobile edge computing in
5G networks: New paradigms, scenarios, and challenges. IEEE Commun. Mag., 55(4):54–61,
April 2017.
11. Y. Xing and T. S Rappaport. Propagation Measurement System and Approach at 140 GHz-
Moving to 6G and Above 100 GHz. arXiv preprint arXiv:1808.07594, to appear in IEEE
Globecom 2018, 2018.
 10

Generation Regulation Services Innovations Notes

1G monopoly voice lightweight mobile terminals humans,  10% popul.
2 G / GSM competing operators cho- voice, SMS, data 9.6 − 200 FDMA/TDMA, smaller mobile humans, > 90% popul., M2M
sen by regulators kbps (GPRS-EDGE), push email terminals, first cameras, GPS,
(Blackberry) content (iPhone), access to
WWW (WAP, not great succ.)
3 G / UMTS granting licenses, open- high data rates 364 kbps - 10 CDMA, camera, GPS superhype, limited success due to
market auction Mbps, email, file/image competition with cheaper WiFi,
mobile devices too heavy, low bat-
tery duration compared to GSM
4 G / LTE less up-front money from downlink ∼ 300 Mbps, uplink up OFDM, MIMO, D2D competitive with WiFi, elimina-
operators, elimination of to ∼ 50 Mbps, mobile video, mul- tion of roaming charges in Europe
roaming charges in Eu- timedia messaging (WhatsApp,
rope Instagram, Telegram, Twitter,...)
5G spectrum licenses up to bit rates ∼ 20 Gbps, lower laten- frequencies up to 60 GHz for superhype
25 years, spectrum shar- cies ∼ 1 ms, D2D, massive M2M extremely high-speed transmis-
ing, reuse, new frequency for IoT sion, massive MIMO, LDPC, Po-
bands mmWave lar Codes, network function vir-
tualization, software-defined net-
working, network slicing

Table 1 Evolution from 1G to 5G

Marco Chiani, Enrico Paolini, Franco Callegati
 Traffic growth is driven by both the rising percentage change for mobile data.
number of smartphone subscriptions This curve illustrates a typical moderating
and expanding average data volume per growth rate as the market evolves over time.
subscription, fueled primarily by more In Q2 2018, mobile data traffic grew
viewing of video content at increasingly by around 52 percent year-on-year and
higher resolutions. 9 percent quarter-on-quarter. There are
The graph below shows total global large differences in traffic levels between
monthly data and
Open voiceand
issues trafficbeyond
from Q2 2013
5G markets, regions and operators. 11
to Q2 2018, along with the year-on-year

Voice Data Year-on-year growth

18 180

16 160

14 140
Total (uplink + downlink) traffic (exabytes per month)

12 120

Year-on-year growth (percent)

10 100

8 80

6 60

4 40

2 20

0 0
Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2
2013 2014 2015 2016 2017 2018

Source: Ericsson traffic measurements (Q2 2018)

Fig. 2 Mobile data

Ericsson traffic (doesThe
notcontent
include DVB-H,
of this documentWi-Fi)
is subject[2].
to EAB-18:008331 Uen, Revision A
SE-164 80 Stockholm, Sweden revision without notice due to continued © Ericsson 2018
Telephone +46 10 719 0000 progress in methodology, design and
www.ericsson.com manufacturing. Ericsson shall have no
liability for any error or damage of any kind
10 2 resulting from the use of this document

10 1
Specific Attenuation (dB/km)

10 0

10 -1

-2
10
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
Frequency (GHz)

Fig. 3 Specific attenuation of atmospheric gases at sea-level.