Road to 5G: Key Enabling Technologies

Abstract

The research on 5G systems is intensively conducted in order to deliver systems by 2020, driven by the exponentially increasing internet traffic, in addition to the emergence of new services and use cases with specific sets of requirements. This paper is a literature review on 5G requirements, use cases and standardization picture. It presents existing and new enabling technologies towards future mobile systems and discusses challenges they face following two aspects: wireless technologies and network technologies.

Full text

Road to 5G: Key Enabling Technologies
Sanae El Hassani1, Abdelfatteh Haidine1, and Hayat Jebbar2
1 LTI Lab, ENSA El Jadida, Chouaib Doukkali University, El Jadida 24000, Morocco
2 EIPM Laboratory, Faculty of Sciences and Techniques, Moulay Ismail University, Errachidia 52000, Morocco
Email: { elhassani.s, haidine.a} @ucd.ac.ma; hayatjebbar25@gmail.com
Abstract—The research on 5G systems is intensively
conducted in order to deliver systems by 2020, driven by the
exponentially increasing internet traffic, in addition to the
emergence of new services and use cases with specific sets of
requirements. This paper is a literature review on 5G
requirements, use cases and standardization picture. It presents
existing and new enabling technologies towards future mobile
systems and discusses challenges they face following two
aspects: wireless technologies and network technologies.
Index Terms—5G, key technologies, 3GPP standards, IMT-
2020, advanced receivers.
I. INTRODUCTION
Mobile traffic has known a huge evolution in the last
few years. On the one hand, mobile subscriptions are
significantly increasing. For example, Ericsson Mobility
report establishes that 44 million new mobile
subscriptions were added globally in Q1 2019, resulting
in a total of around 7.9 billion subscriptions worldwide.
On the other hand, the mobile data traffic has also
significantly increased due to new emerging applications
and use cases. The growth is about 82% between Q1
2018 and Q1 2019 [1]. In particular, mobile video traffic
and social networking traffic are currently dominating on
smartphones and tablets, and it is forecasting a traffic
growth in volume, as predicted by Cisco in Fig. 1, but
also a redistribution of traffic dominance; smartphones
are currently predominant in terms of generated mobile
traffic, but the Machine-to-Machine (M2M) connections
represent the fastest growing device/connection category,
expected to grow from 11% to 31% of global mobile
devices and connections between 2017 and 2021[2].
This rising demand of data volume and the diversity of
requirements are leading the development of the fifth
generation (5G) of mobile systems. Following the rule of
a generation issuing every decade, 5G commercialization
is expected by 2020. Intensive research has been
conducted to make this technology available even before,
targeting among other enhanced mobile broadband
services and a wider range of Internet of Things (IoT)
applications. The variety of research domains linked to
5G makes it necessary to situate actual needs and
research choices about 5G technologies.
Manuscript received April 5, 2019; revised September 29, 2019.
Corresponding author email: elhassani.s@ucd.ac.ma
doi:10.12720/jcm.14.11.1034-1048
This paper is a review of the 5G research trends from
requirements and use cases to enabling technologies. The
second section illustrates 5G standardization activities,
objectives and applications. The third section elaborates
key enabling techniques of wireless 5G access, and the
fourth section gives an overview of 5G network enabling
techniques. The last section gives some conclusions about
current and future 5G research trends.
Fig. 1. Cisco forecast for mobile data traffic by 2022 (Source [2])
II. 5G REQUIREMENTS AND USE CASES
From a basic mobile voice system to advanced
broadband services, mobile systems have known many
evolutions, and the actual landscape is deploying 4G and
pre-5G systems while developing 5G systems. In fact,
Long Term Evolution (LTE) and LTE-Advanced (LTE-A)
are deployed in many regions, and already provide better
data rates. However, an increasing mobile data traffic
with more user centric concept is needed. 5G research has
been thus launched in 2013 and enhancements of current
concepts and technologies, as well as new ones, are being
proposed. A possible definition of 5G is given by the
Next Generation Mobile Networks (NGMN) Alliance and
describes 5G as “an end-to-end ecosystem to enable a
fully mobile and connected society. It empowers value
creation towards customers and partners, through existing
and emerging use cases, delivered with consistent
experience, and enabled by sustainable business models”
[3].
A. International Mobile Telecommunications-2020
(IMT-2020)
The 3rd Generation Partnership Project (3GPP)
standardization has issued LTE Releases 8 and 9 which
fulfill IMT-Advanced requirements and were considered
as 4G, then downgraded to 3.9G while LTE-Advanced
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1034

were considered as 4G [4]. As depicted in Fig. 2, next
releases brought each an evolution for 4G, and LTE
Release 13 has been almost frozen in March 2016 [5].
Actually, 5G standardization is based on functionally
frozen specifications before 2020, authorized after
evaluation as IMT-2020, and Release 13 and Release 14
are considered as a pre-5G or LTE-A Pro. 5G starts with
Release 14 with new fundamentals, study and
formalization of some selected specifications into
standards in Release 15 (Phase 1) and Release 16 (Phase
2). The freezing date for Release 14 Radio Access
Network (RAN) work program by 3GPP was in June
2017, and Release 15 standards are planned to be made
operational at least in 2020. The Release 15 has been
actually divided into three separate sections; an "early
drop" working on the bulk of the initial 5G standard,
including a "standalone" (SA) option for 5G in order to
allow operators to deploy 5G without an LTE network.
The second main drop of Release 15 was approved in
June 2018, and the “late drop” in April 2019. The next
release, Release 16, named also 5G phase 2, is expected
to be approved in December 2019, and to include a wider
range of technologies, like 5G in unlicensed spectrum and
specifications for IoT technologies. Release 17 might
include spectrum above 50GHz, enhancements for
Unmanned Aerial Vehicles (UAV), options for 5G
multicasts and broadcasts, IoT specifications for
industrial sensors, and Artificial Intelligence (AI) and
machine learning enablers for 5G network operations. It
can be thus interpreted as 5G-Advanced [6].
Note that more particularly, 5G NR (New Radio),
which is the New Radio Access Technology (RAT)
developed by 3GPP for the 5G mobile network, started in
2015 and ended in September 2018, and focused on
delivering the first set of 5G standards [6]. The first
specifications were published in late 2017.
Fig. 2.The radio evolution in the present decade (Source: [4])
Commercial 5G networks already started to appear in
some cities. The first ever 5G NR call on a commercial
network was made in September 2018 with simulated
smartphone device by Verizon, Ericsson and Qualcomm,
while Verizon and Nokia completed the first over-the-air
data transmission on a commercial 5G NR network in the
same period [7]. Other examples are the recent switch on
of 5G networks in many regions, such as cities in
Australia by Telstra and in UK by EE in May 2019, and
cities in Spain and Italia by Vodafone in June 2019 [8].
Estimations from [1] forecast that by 2024, 5G
subscriptions will reach 1.9 billion, and 5G coverage up
to 65% of the world's population.
B. 5G Requirements
5G is expected to multiply mobile data traffic per area
by a factor of 1000. The typical user data rate and number
of connected devices is targeted to be 10 to 100 times
higher, and battery life for low power devices to last 10
times longer with an End-to-End (E2E) latency reduced
by a factor of 5. These targets are adopted in order to
enable diverse use cases addressed by 5G, with
requirements identified among others by a focus group of
the International Telecommunication Union (ITU) for
2020 and beyond, hosted by ITU-Standardization Sector
ITU-T. It highlights three main categories of use cases,
based on the key considered services:
 Extreme mobile broadband (xMBB)
 Massive Machine-Type Communications (mMTC)
 Ultra-reliable Machine-Type Communications
(uMTC)
Requirements such as delay, throughput, reliability, etc.
are related to the targeted use cases. A summary of 5G
requirements, as described in [9], is given in Fig. 3. Table
I gives examples of applications related to requirements.
Fig. 3. Summary of key requirements for 5G
TABLE I: EXAMPLE OF 5G REQUIREMENTS (SOURCE: [10])
Requirement
s
Desired Value
Application example
Data Rate
1 to10Gbps
Virtual reality office
Data Volume
9GB/h (busy period)
500GB/month/user
Stadium, Dense urban
information society
Latency
Less than 5ms
Traffic efficiency and
safety
Battery Life
One decade
Massive deployment
of sensors and
actuators
Connected
devices
300000 devices/AP
Massive deployment
of sensors
Reliability
99.999%
Teleprotection in
smart grid network,
Traffic efficiency and
safety
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1035

C. 5G Use Cases
Current broadband services are expected to evolve
significantly, as well as other emerging use cases that
should be supported by 5G. NGMN presents eight use
case categories through representative examples [3], as
illustrated in Fig. 4. These categories can be described as
follows:
 Broadband Access in Dense Areas : Pervasive Video,
Smart Office, Operator Cloud Services, HD
Video/Photo Sharing in Stadium/Open-Air Gathering
 Broadband Access Everywhere : 50 + Mbps
Everywhere, Ultra-low Cost Networks
 Higher User Mobility : High Speed Train, Remote
Computing, Moving Hot Spots, 3D Connectivity:
Aircrafts
 Massive IoT: Smart Wearables, Sensor Networks,
Mobile Video Surveillance
 Extreme Real-Time Communications : Tactile
Internet
 Lifeline Communication : Natural Disaster
 Ultra-reliable Communications : Automated Traffic
Control and Driving, Collaborative Robots, eHealth,
Remote Object Manipulation (Remote Surgery), 3D
Connectivity (UAV), Public Safety
 Broadcast-like Services : News and Information,
Local Broadcast-like Services, Regional Broadcast-
like Services, National Broadcast-like Services
Note that answering these promised use cases implies
the enhancement of already used techniques as well as the
use of new ones.
Fig. 4. 5G use case families and related examples (Source [3])
III. WIRELESS ENABLING TECHNOLOGIES
Fig. 5. Breakthrough in three different dimensions (Source: [11])
The fulfillment of 5G requirements implies a dramatic
change in the design of cellular architecture and wireless
technologies. The massive network capacity as well as
the increased user experience required for 5G systems
suggests researches in numerous dimensions, as
illustrated in Fig. 5.
Following the requirements addressed for 5G, the
RAN appears as one of the biggest challenges [12]. For
instance, the increased required bandwidth leads to a
reconsideration of the current use of already adopted
mobile frequency bands and a better use of unlicensed
spectrum, as the one of 5GHz, in addition to the move to
millimeter wave (mmWave) bands. This implies a
number of challenges allocating and re-allocating
bandwidth, as the new system should provide a
ubiquitous high-rate low-latency experience for network
users. On the other hand, increased spectral efficiency
and spectrum utilization, through new waveforms and
advanced Multiple Input Multiple Output (MIMO)
among others, is to be adopted. Extreme densification and
offloading are also key factors of capacity enhancement.
All these have taken into consideration the energy
efficiency issue for all the system’s components, as it
becomes critical when the network goes denser. Network
densification also positions solutions for virtualization
and softwarization of the network as a must.
Note that 5G systems are meant to represent a
convergence of existing systems with the integration of
new techniques and components. As discussed in the
previous section, the coming 3GPP releases, named as
beyond 4G, include some key enabling techniques to
address 5G requirements. Fig. 6 illustrates techniques
evolution towards 5G.
Fig. 6. Evolution to 5G (Source: [8])
In the following, 5G wireless enabling techniques are
discussed through spectrum utilization, spectrum
efficiency and densification issues.
A. Spectrum Utilization
5G is a convergence of existing and new systems with
new air interface. New waveforms, multiple access
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1036

technologies, modulation and coding schemes, etc. are
being proposed.
a) Spectrum allocation
5G is a heterogeneous network and its spectrum is
expected to be a combination of established and new
bands, with a major challenge being to integrate the
various bands. The focus is then on spectrum utilization
in addition to spectrum efficiency. Some of the spectrum
below 6GHz is thus being re-purposed for use with newer
technologies, in particular for Non-Line-Of-Sight (NLOS)
requirements.
Fig. 7. 5G will aggregate sub 6GHz and above 6GHz
The new spectrum below 6GHz has been allocated for
mobile communication at the WRC 2015, and is the
primary band of 5G, while the band above 6GHz,
considered as complementary 5G band, is expected to be
allocated at WRC 2019 [13] (October), as illustrated in
Fig. 7. In WRC-15, ITU through its ITU-R, studied 5G
mobile broadband systems in the 24.25-86 GHz range
and prepared the decisions for WRC-19. On the 3GPP
Release-15, the NR-U work item supports both the
existing 5GHz unlicensed band and the new "greenfield"
6GHz unlicensed band, and other unlicensed and shared
spectrum bands, including mmWave, can be expected in
coming releases [14].
In fact, frequency bands for 5G present different
propagation characteristics and available bandwidths,
which must be considered for system design. Decisions
taken at WRC-15 do not consider though studies of bands
below 24.25 GHz, while the 27.5-29.5 GHz band is not
on the ITU-R list for WRC-19, even though several
countries were intending to use all or parts of it. As a
result, mobile industry ought to find solutions for early
deployments of 5G, where parts of the range 3.1-4.2 GHz
are actually seen as such. The 600 MHz (US) and 700
MHz (Europe) bands are also being considered in some
countries for early 5G services deployment by Mobile
Network Operators (MNOs).
Note that the backhaul spectrum for 5G networks also
needs to be considered; the increasing capacity needed
for backhauling has already driven the shift towards
higher frequencies about a decade ago, when the 26 GHz,
28 GHz, and 32 GHz bands were introduced. 70-80 GHz
band is also interesting and enables capacities in the order
of 10 Gbps or more over distances of a few kilometers.
The appropriate band depends, among others, on the
national spectrum regulation authorities and on the
regional climate. In fact, microwave backhaul has
traditionally used frequencies from about 6GHz to
86GHz, while mobile broadband networks frequency
bands are, or will be, ranging from about 450MHz to
about 5GHz, for which WRC-15 added some new bands
for 4G mobile broadband use. The emerging techniques
of spectrum sharing and flexible spectrum use between
mobile radio access and microwave backhaul are thus key
enablers for 5G.
b) Carrier Aggregation (CA)
As stated, 5G is expected to present a convergence of
existing and new systems. The multi Radio Access
Technologies (multi-RAT) requirements for 5G imply the
need to aggregate, for the same end user, RATs possibly
operating on different bands. More efficient use of the
fragmented and crowded spectrum can be made that way
and needs to use coordination and load balancing
between different RATs. Furthermore, asymmetric uplink
and downlink aggregation provides carrier allocation
flexibility. The concept of CA has already been proposed
for LTE since 3GPP Release 10 in the case of LTE-LTE
aggregation of inter and intra band contiguous carriers.
Intra-band non-contiguous CA has been presented in
Release 11, and CA using both Frequency Division
Duplex (FDD) and Time Division Duplex (TDD) band
for Release 12. A CA method using 5 GHz WiFi bands,
called Licensed Assisted Access, is presented in Release
13, and 5G standardization starting from Release 14
defines more cases following the 5G adopted bands [15].
Some challenges have to be solved in this case for
configuring the Internet Protocol (IP), coordinating
interference and time synchronization. In addition,
resource scheduling over the aggregated bandwidth has to
be optimized and may use advanced tools, such as game
theory [16].
c) MmWave
New spectrum is allocated to face expected congestion
of wireless technologies. MmWave frequencies could be
used to augment the currently saturated 700 MHz to 2.6
GHz radio spectrum bands for wireless communications
[17], [18]. Compared to currently used bands, little
knowledge about cellular mmWave indoor and outdoor
propagation environments with high users’ density is
available. In fact, the rather hostile propagation
environment at these frequencies, in addition to hardware
equipment costs, limited their use so far. The emerging
techniques using large antenna arrays should allow
though narrow beams to communicate, and novel
transceivers hardware designs are needed to realize such
functions and enable mmWave use. As an example, New
York University (NYU) and NYU-Poly through their
NYU Wireless research center are conducting research to
create new technologies and fundamental knowledge for
future mmWave wireless devices and networks [19]. Fig.
8 illustrates the position of 5G mmWave expected
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1037

frequencies in regards to future bands (24.25-27.5 GHz,
27.5-29.5 GHz), in addition to first phase lower 5G
frequencies.
Fig. 8. MmWave frequencies for 5G
Note that Samsung demonstrated the world’s First 5G
mmWave Mobile Technology in May, 2013 using
adaptive array transceiver technology over mmWave
frequency bands for outdoor cellular and studying device
feasibility and densification [12]. Many other mmWave
systems have also been demonstrated since then, as it is
the case for Huawei [20]. A more accomplished 5G pilot
system was demonstrated in 2018 Winter Olympics [21].
B. Spectrum Efficiency Issues
The enhancement of existing techniques, in addition to
new ones, such as new waveforms, multiple access
technologies, modulation and coding schemes, etc. are
being proposed to meet 5G requirements.
a) Massive MIMO
Massive MIMO is considered as one of the most
important techniques for 5G Radio Access. Based on
multiple antennas, MIMO is already used in LTE and
LTE-A, and provides considerable diversity/capacity
gains. The use of multiple antennas at the Base Station
(BS) helps to focus the energy to form beams towards the
device. Unlike LTE and LTE-A, the large amount of
antennas, defining massive MIMO, can make it possible
to serve many users in an accurate way. As described in
[22], when combined with other 5G key enablers,
essentially small cells and mmWave, massive MIMO can
provide significant capacity gains to meet the expected
explosion of data traffic demand. A new architecture is
required then to realize the Full-Dimension MIMO (FD-
MIMO) [23], and the use of massive MIMO can enable
other key proposed technologies such as Beam division.
As illustrated in Fig. 9, massive MIMO can be used to
serve users but also for backhauling [24]. Note that the
equipment of BSs with a large number of antennas may
not be applicable to all types of antennas. On the other
hand, BSs with a very variable number of antennas will
have to coexist, and even take advantage of this diversity
to prevent interference. The channel estimation is also
critical for massive MIMO as the channel state
information (CSI), though important for beamforming,
can imply large overheads.
b) Channel modeling
In order to enable many of the physical layer
techniques for 5G, accurate radio propagation models are
needed. The expected bands to be used, especially above
6GHz, are not addressed by current models. Thus, new
field measurements and ray tracing are needed for
channel modeling, and research is already conducted for
mmWave characterization in several environments [18].
Fig. 9. BS using Massive MIMO to serve UEs and relays through beams
As highlighted in [25], the new channel model is an
extension of the existing 3GPP 3D channel model. In
addition, the frequency range handled by the model
should be up to 100 GHz with a multi-band
characteristics evaluation, and supports different mobility
cases and large channel bandwidths. Many research
activities are conducted to respond to these requirements.
In particular, 3GPP RAN Meeting held in Busan Korea in
June 2016 approved the first standard for the mobile
broadband 5G high-frequency (6-100 GHz) channel
model, considering existing and new scenarios. The
ongoing works use new coordinate systems and antennas
models, and consider mainly modeling of Pathloss, LOS
probability, penetration, fast fading and blockage. A
survey on approaches, models and measurements for 5G
mmWave channels, considering massive MIMO channels
and human body blockage, can be found in [26].
c) Signal waveforms
LTE is based on multicarrier modulation, namely
Orthogonal Frequency Division Multiplexing (OFDM)
and Single Carrier Frequency Division Multiplexing (SC-
FDMA) [27], which allowed a good spectrum efficiency
over past techniques. However, the use of Cyclic Prefix
(CP) reducing spectral efficiency, the large sidelobes
caused by the rectangular pulse shaping and implying
considerable Out Of Band Emission (OOB) as well as the
high Peak to Average Power Ratio (PAPR) all suggest
possible enhancement of multicarrier current techniques.
A first example of waveform candidates is filtered
OFDM, which suggests filtering the signal in order to
remove OOB emissions [28]. Filter Bank based
MultiCarrier-Offest Quadrature Amplitude Modulation
(FBMC-OQAM) modulation is another adaptation of
OFDM, where no CP is required, enhancing thus
spectrum efficiency [29]. Filtering in FBMC-OQAM is
applied on a sub-carrier basis both at transmitter and
receiver which, along with OOB reduction, makes the
signal more robust against Inter-Symbol Interference
(ISI), Doppler effect, synchronization imperfections,
spectrum fragmentation, etc. Other resulting properties
include self-equalization, which can reduce the number of
unused subcarriers, and blind channel tracking, reducing
pilot contamination and thus enhancing massive MIMO
functioning. On the other hand, the orthogonality is
relaxed from the complex field using QAM to the real
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1038

field using OQAM, which may re-question some
orthogonality techniques used for CP-OFDM, such as
Alamouti Space time coding.
Universal Filtered Multi Carrier (UFMC), also called
UF-OFDM, applies a subband filtering instead of
subcarrier filtering as in FBMC-OQAM [30]. The choice
of the subbands and of the filter is flexible so to achieve
the desired OOB radiation and in-band distortion. As a
result, the use of a prefix depends on filters design.
Similarly, Generalized Frequency Division
Multiplexing (GFDM) also aims to provide a flexible
multi-carrier technique through additional degrees of
freedom compared to OFDM. It uses adjustable pulse
shaping filters on a subcarrier basis to control OOB
emissions. Furthermore, data symbols are grouped both in
frequency and time dimensions, with a group size being
also an adjustable parameter. The use of a CP may be
needed in this case. The filtering flexibility implies
however self-created interference, which is dealt with
using receiver methods and iterative interference
cancellation [31].
Comparison of such techniques has been elaborated
through different studies, as by METIS project in [32]. A
qualitative comparison of mentioned waveforms is given
in Table II.
TABLE II. COMPARISON BETWEEN GFDM, FBMC, UFMC
GFDM
FBMC
UFMC
PAPR
Low
High
Medium
OOB
Very low
Low
Low
Spectral efficiency
Medium
High
High
Processing complexity
Medium
High
High
CP
Yes
No
No
Orthogonality
No
Yes
Yes
ISI/Multipath distortion
Medium
High
High
Synchronization requirement
Medium
Low
Low
Effect of frequency offset
and phase noise
Medium
Medium
Medium
Latency
Short
Long
Short
Ease of integration with
MIMO
Yes
Yes
Yes
Note that 5G NR adopted still OFDM based
waveforms with 5 different subcarrier spacings: 15 kHz,
which is same as LTE, 30 kHz, 60 kHz, 120 kHz, and
240 kHz, with possible use of either normal CP or
extended CP.
d) Multiple access
As OFDMA was successfully used for 4G as a
multiple access technique, it may survive for 5G, but
considering enhancement of the multicarrier modulation
to be adopted. With the use of massive MIMO in a dense
network context, multi-user MIMO (MU-MIMO) can be
enabled through Spatial Division Multiple Access
(SDMA) that uses multipath properties to multiplex users
in the spatial dimension. The communication between the
BS and the User Equipment (UE) can use an orthogonal
beam, dividing antenna beams according to locations of
the UEs [33]. In order to achieve higher capacity, Non-
Orthogonal Multiple Access (NOMA) schemes are also
proposed in [34]–[36], which allows more than one user
to share the same subband without coding/spreading
redundancy, breaking thus orthogonality. The receiver
implements then joint processing to separate users’
signals, aided by subcarrier and power allocation
algorithm to maximize the offered rates [37]. NOMA
flavours include among others Resource Spread Multiple
Access (RSMA) and Sparse Code Multiple Access
(SCMA) [5], [9], [38]. In fact, RSMA, based on Low
Density Spreading CDMA, spreads user signal with
appropriate channel coding, instead of simple repetition,
depending on the demanded rate. On the other hand,
SCMA uses a low density spreading with multi-
dimensional constellations. The dimensions are partially
used by one user and zero padded for the resources to be
used by other users [39]. For both cases, NOMA principle
of multiuser interference cancellation is used at the
receiver. Up to now, 3GPP NR has specified flexible
technology framework that can be tuned to enable a wide
range of 5G scenarios, and more specifications are
expected for Release 16 [40].
e) Channel coding
The use of new spectrum, new waveforms and multiple
access techniques, the diversity of use scenarios and
devices needing rate adaptability and decoder flexibility
implies the need of appropriate channel coding in 5G
context. The turbo codes used in 3G and 4G cellular
systems have been rediscussed at 3GPP level, and were
replaced in 5G NR by Low Density Parity Check (LDPC)
and polar code in 5G, following the channel conditions.
In fact, turbo codes must have component codes finely
designed to maximize free distance, in addition to a good
interleaver design, while LDPC codes are patent-free and
less complex per iteration. LDPC codes don’t require a
random interleaver and have a lower error floor. In
addition, the use of non-binary version of LDPC codes
can enhance the performance especially for small and
medium packet lengths, even though they are more
complex [41]. Enhancement of LDPC codes can also
include spatially coupling (SC-LDPC) using convolution
and a windowed decoding for complexity reduction. The
other adopted codes, polar codes, were relatively recently
proposed by Arikan [42]. These codes give a constructive
technique to achieve channel capacity with bounded
complexity. They are suitable for multi-terminal
environment and are constructed recursively via
Kronecker products. However, the disadvantage of polar
codes however lays in its rather high latency due to the
inherent nature of the code construction. Note that polar
codes were pushed for 5G systems notably as they were
adopted for Huawei demonstration [13], while sparse
regression codes were subject of a European project that
ended in 2018 [43]. Huawei announced achieving
27Gbps using polar codes [44], and discussions in 3GPP
identified it as 5G enhanced Mobile BroadBand (eMBB)
channel coding scheme in the RAN1 (wireless physical
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1039

layer) 87 meeting [45]. Generally, as compared in [46],
the interest of these advanced codes depends on the
context use; instead of considering capacity-approaching
for point-to-point Additive White Gaussian Noise
(AWGN) channel gains as in 3G/4G, transmission
techniques, such as the waveforms, multi-antenna
processing and network coding are considered for 5G.
Nonetheless, optimization in accordance with multiple
aspects is sometimes conflicting [47].
C. Network Densification
a) Ultra Dense Network (UDN)
Future 5G networks are meant to support the extremely
heavy traffic forecasted for the coming years. This leads
to a rethinking of the system design. Not only spectrum
efficiency will have to be improved, as for previous
evolutions from generation to generation, but a
substantial network architectures redesign, through
densification, will be necessary.
Cellular infrastructure densification was applied since
the second-generation voice-oriented systems in order to
enhance capacity or performance of an already deployed
system through cell splitting and sectorization. Ultra
densification for future networks has a more important
role, while providing an extreme and user-centric reuse of
system bandwidth on the spatial domain, and improving
propagation conditions by reducing the distance between
the end user and the BS.
Fig. 10. UDN infrastructure composed by operator- and user-deployed heterogeneous serving ANs (source [41])
The densification considers operator-deployed
infrastructure elements as well as user-deployed access
nodes (ANs) and mobile user devices, as illustrated in Fig.
10, where multiple types of users and machines act as
served nodes, and disruptive devices act as prosumers.
For a given area, the number of serving and user nodes is
of the same order. UDN imply an increased spatial reuse
of system resources, a large node density and irregular
deployment. This deployment is challenging in terms of
interference, design and choice of density, network-wise
coordination, load balancing, etc.
In fact, it is critical while densifying preserve the cell-
splitting gains, in particular with the constraint of low
power BSs. The limitation caused by interference in such
networks has also to be carefully handled, especially for
mmWave channels. In addition, there should be a fine
optimization of users and BSs association in such
heterogeneous multi-RAT using performant concepts,
such as game theory, and models for mobility support
such as virtual cells. The localization of nodes has thus
to be handled as it is an input for that optimization. The
deployments should be designed taking into account the
possibly high cost of a dense infrastructure.
Cell-edge effect is an issue for UDN. In fact,
densification challenges mobility and interference, and
solutions adopted for LTE (Coordinated MultiPoint
(CoMP)), CDMA (soft handover) or Wimax (fast cell
selection) have to be replaced by an adequate system with
ideal backhaul. A possible solution, among others, is the
smooth cell virtualization [48].
The backhauling in a dense network is also a
challenging issue, as discussed in [24], and it is proposed
to use self-backhauling as a cost-effective solution. Given
a dense deployment network, self-backhaul nodes can
also provide traffic offloading and cell split using content
prediction and caching techniques.
b) Cloud-Based Radio Access Network (C-RAN)
Network densification is a key enabler of 5G networks.
This implies a heavy radio processing to be handled by
the BSs. Cloud-based Radio Access Networks (C-RANs)
is adopted in 5G RAN to handle such amount of
processing. In fact, C-RAN decouples the BaseBand
processing Unit (BBU) from the Remote Radio Heads
(RRHs), allowing thus a centralized processing and
assignment of radio resources. The BBU part can be
moved to a pool, or more generally to the cloud, in order
to optimize resource utilization. The transmission of radio
signal is then performed by RRHs based on computed
signals from the cloud. The BBU pool is a virtualized
cluster connected to the core network through the
backhaul. Each RRH is connected with the cloud BBU
pool via a fronthaul. Cloud based computation makes
RRHs cheaper and simpler, allowing more scalability and
densification opportunities. It also reduces the delay for
inter-cell coordination and permits more joint
optimization of processing, e.g. for cooperative
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1040

interference management, scheduling, CoMP, etc. [49].
C-RANs can also be self-organizing, where intelligent
smart cells are deployed in the area covered by a macro
BS, and can also enable advanced clustering and
coordination schemes to enhance mobility and handovers.
Moreover, C-RAN has to adapt to Heterogenous Network
(HetNet) context, leading to H-CRAN. In such networks,
different types of nodes, including Low Power Nodes
(LPN, e.g., pico BS, femto BS, etc.) and high power
nodes (HPN, e.g. 4G BSs) may have to cooperate using
the same computing pools, and solutions to solve inter-
tier interference improve joint processing gains. The H-
CRAN is illustrated in Fig. 11.
Note that CoMP, already used in LTE, is adopted and
enhanced for 5G, as it is a tool for interference combating
in case of multiuser, very likely to happen in a UDN, in
addition to the enhancement of coverage, cell edge
throughput, and/or system efficiency [51], [52].
Fig. 11. H-CRAN: Heterogeneous cloud radio access networks
(source ( [50])
D. Energy Efficiency
Network densification is an effective method of
increasing system capacity and coverage. In parallel,
increased power consumption is expected. Thus, energy
efficiency is a central challenge for 5G systems, due to
the diversity of expected applications. In fact, the device
battery life has to be lengthened and the overall energy
efficiency of the network has to be improved. A solution
can be energy harvesting from environmental energy
sources for example. The variability of available energy
levels over time, locations, weather conditions etc.
presents however a challenge for Quality of Service (QoS)
constrained wireless applications. Another solution can
be the harvesting of energy from ambient radio signals:
RF-powered energy-harvesting networks (RF-EHN) [53],
[54] are actually a hot topic. The energy efficiency from a
network side is also to be optimized. For example, it is
proposed to reduce the energy consumption of BS in a
smarter way, through cellular partition zooming for
example, where the BS can zoom in to maintain the
coverage area according to actually present users, and
zoom out to sleep mode to save energy [55]. Other
optimizations are proposed for the whole network,
including core, fronthaul, backhaul, etc. [56].
IV. NETWORK ENABLING TECHNOLOGIES
Given the diverse 5G requirements, 5G networks need
to be carefully designed in order to provide enough
flexibility to meet all these requirements in an efficient
and cost-effective fashion. This is to say, emerging use
cases and support for vertical markets, multi-tenancy and
multi-service, fixed-mobile access networks convergence,
etc. imply a new ecosystem and need thus orchestration
functions to serve resources to multiple logical networks.
Many challenges arise when insuring evolutivity of the
network to multiple future technologies along with a
flexible resource sharing through cross-domain
orchestration. In fact, multiple logical networks, known
as network slices, are needed for 5G. This new network
view can be possible using Mobile Edge Computing
(MEC) technique for concurrent instantiations of the
network. Required capabilities can be enabled through
Software Defined Network (SDN) and Network Function
Virtualization (NFV). In addition, the expected 5G
architecture positions security and privacy as a major
issue. Views and techniques being developed to meet 5G
network requirements can be found in the white paper
issued by the 5G Infrastructure Public Private Partnership
(5G PPP) Architecture Working Group [57], where the
5G PPP is a joint initiative between the European
Commission and European ICT industry. This section
presents some logical and functional aspects and
techniques of 5G architecture.
A. Network Architecture
Cellular networks are evolving to a multi-radio access
technology (multi-RAT) and multi-layer heterogeneous
network. At the same time, emerging mobile use cases
make the current network architecture relatively
incompetent to address current and future requirements.
The network flexibility is a central objective in 5G
architecture design, which suggests, as pointed in [58],
the usage of a multi-service and context-aware adaptation
of network functions, adaptive decomposition and
allocation of mobile network functions, software-defined
mobile network control, and joint optimization of mobile
access and core network functions. A split in logical core
and RAN is also necessary to allow evolution of both
RAN and Core and to insure more deployment flexibility.
This is illustrated by the adoption by the European
Telecommunications Standards Institute (ETSI) of
separation of User Plane (UP) functions and Control
Plane (CP) functions, and modularization the function
design, e.g. to enable flexible and efficient network
slicing [59].
Similarly, NGMN presents a 5G architecture
comprising three layers and a management entity: the
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1041

infrastructure resources layer, the business enabling layer
and the business application layer, in addition to an E2E
management and orchestration entity to configure and
supervise all three layers according to the demands of the
requested services or business model. The architecture is
based on central cloud and comprises typically multiple
data centers which may be very distant, connected to each
other or to the central cloud or edge clouds through wide
area network (WAN) among others. Fig. 12 illustrates
this topology as viewed by the project 5G NOvel Radio
Multiservice Adaptative network architecture (5G
NORMA), which is one of the 5G PPP projects under the
horizon 2020 framework [60]. Note that control and data
planes have to be separated as network densification
increases overhead of control signaling while requiring a
high capacity of the backhaul, especially for small cells
and Device-to-Device (D2D). As suggested in [61], low
layer functionalities will be provided by Radio logical
network while a network cloud provides all higher layer
functionalities in a way that avoids functionalities
redundancy. In addition, functions can be dynamically
deployed in the network cloud through SDN/NFV.
Fig. 12. 5G NORMA topology view on architecture (Source [51])
B. NFV/SDN
Virtualization facilitates resource sharing among many
operators. As there is a need to decouple hardware from
software and to move network functions towards software,
and assuming separation between control and data, NFV
is one of main 5G architecture drivers, ensuring network
adaptability and making it easily scalable [59]. With
simpler operation, new network features are likely to be
deployed more quickly. NFV enables in fact sharing
common resources through abstraction of physical
resources to a number of virtual resources. It uses
automation mechanisms in order to accelerate scalability
following the increasing network demand. An example of
wireless network virtualization is illustrated in Fig. 13.
Fig. 13. An example of wireless network virtualization (Source [62])
Basically, NFV installs network function software in
virtual machines deployed in a virtualized commercial
server, not in dedicated network equipment individually.
RAN works thus as edge cloud while Core works as core
cloud. Connectivity among virtual machines located in
edge and core clouds are provisioned using SDN.
Actually, NFV/SDN allow network linking to the cloud
to enable the suggested ultra-flexible network
architecture, as illustrated in Fig. 14.
Fig. 14. NFV Vs. SDN (Source [63])
SDN fits into the NFV paradigm and can enable the
orchestration of the NFV physical and virtual
infrastructure resources, through supporting provisioning
and configuration of network connectivity and bandwidth.
An example of SDN components is illustrated in Fig. 15.
Note that in addition to programmability, flexibility,
adaptability and capabilities, NFV/SDN can allow among
others Operational EXpenditure (OPEX) and Capital
EXpenditure (CAPEX) reduction, speedy service creation
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1042

and deployment, efficient service life-cycle management,
energy consumption reduction, and improved quality of
experience for users. Along with SDN, NFV can offer, in
addition to an increased throughput, a better Quality-of-
Experience (QoE) for the end users, improving thus
capital efficiencies vs. dedicated hardware
implementation solutions, flexibility in assigning
functions to hardware, scalability, resilience and resource
sharing, operational efficiencies, power efficiency,
service innovation through software based deployment,
migration to newer technologies by isolating part of the
network, granularity of assigned resources, workload
migration, automation and operating procedures
mutualization, decoupling of functionality from location
[64].
Fig. 15. SDN Components descriptions (Source [54])
C. Network Slicing
The new 5G ecosystem needs to provide
communication services for multiple applications with
different requirements. In fact, 5G is meant to serve a
variety of devices with different characteristics and needs,
like mobile broadband, massive IoT, and mission-critical
IoT, each requiring different types of features and
networks in terms of mobility, charging, security, policy
control, latency, reliability, etc. Using the same physical
infrastructure, multiple logical networks will be needed,
known as network slices. Network Slicing is a network
technique, initially for the core network but extended to
an E2E concept including also RAN part. Network slicing
enables the delivery of services with differential
performance characteristics.
For example, massive IoT service connecting
immobile sensors measuring temperature, humidity,
precipitation, etc. to mobile networks does not require
advanced mobility features, critical in serving mobile
phones. On the other hand, mission-critical IoT service,
like remote controlled robots or autonomous driving,
requires, unlike mobile broadband service, a very low
E2E latency. It can’t be cost-effective to build a
dedicated network for each service. Instead, network
slicing can define multiple logical networks over a single
physical network. The physical network is sliced up into
different virtual E2E slices following on the service
requirements. Each slice is made of a number of network
functions and given radio access technology
parametrization, combined together for a given use case
and/or business model. The slice is logically isolated and
is assigned dedicated resources. The isolation of slices
separates error propagation between slices. Note that
virtualized network functions are placed in different
locations in each slice (i.e. edge or Core cloud) depending
on services, and therefore customized by operators.
D. Mobile Edge Computing
The expected network density and application diversity,
which may lead to congestion, can be alleviated through
Mobile Edge Computing (MEC), mainly standardized by
ETSI, 3GPP and ITU-T. The management of cloud-based
applications, either for 4G or future 5G networks, can
also be improved through MEC. In fact, MEC provides
Information Technologies (IT) and cloud computing
capabilities within the RAN in close proximity to mobile
subscribers, and thus accelerates responsiveness for
content, services and applications demand, enabling ultra-
low latency, high bandwidth and real-time radio network
information, particularly useful for context-related
services. Examples of MEC use cases include [56]:
 RAN-aware content optimization in terms of cell
load, link quality, etc. so to increase optimization
and QoE.
 Dynamic content optimization, particularly video,
distributed video analytics and video management.
 Distributed Content and DNS Caching to save
backhaul and transport and enhance QoE
 Augmented Reality (AR) content delivery,
providing local content caching and local object
tracking with a minimized Round Trip Time (RTT)
and optimized throughput, all critical in AR QoE.
An illustration of QoE improvement through MEC is
illustrated in Fig. 16.
Fig. 16. Improved QoE with MEC in close proximity to end users
(Source [65])
Fig. 17. Cellular network evolution to Cloud RAN and MEC (Source
[66])
Note that MEC is related to cloud as illustrated in the
Fig. 17. In fact, Centralized RAN process centralizes
baseband processing capabilities while cloud RAN
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1043

implements baseband processing capabilities in
distributed locations. On the other hand, MEC introduces
new processing capabilities in the BS for new
applications, with a new functions separation and a new
interface between BBU and RRU.
A demonstration for a 5G mobile edge proof of
concept was conducted by Ericsson and Vodafone in their
approach of 5G commercialization. The test considered
remote machine vision in the manufacturing vertical, and
succeeded to showcase a ’5G Smart Network Edge’ with
an enhanced efficiency in machine vision with low
latency, serving to guarantee a detection rate from the
cloud system [67]. In parallel, 5G system specifications
provide new functionalities that serves as enablers for
edge computing, which can be found in [68]. MEC
currently deployed in LTE networks is connected to the
user plane and designed as an add-on to a 4G network in
order to offer services in the edge, but 3GPP designs
MEC in 5G in order to map onto Application Functions
(AF) using services and information offered by other
3GPP network functions, following the configured
policies. Other enabling functionalities are also defined
for flexible MEC deployments [69], [70].
E. Self-organizing Networks (SON)
5G networks are meant to provide a heterogeneous and
dense structure, seen as a “network of networks” with
different types of access technologies to create a seamless
user experience. Use cases like autonomous driving and
tactile internet, in addition to more traditional voice and
data applications, need a fundamental change in the way
networks are managed, with a much more automation and
dynamic predictive resource allocation. SON technology
is seen as an important enabler of these requirements. The
Release 7 of the Small Cell Forum, having among others
3GPP and NGMN as partners, focuses on SON
technology as a 5G enabler, which is defined over
“Hetnets multi-environment-multi-technology, multi-
domain, multi-spectrum, multi-operator and multi-
vendor” which can be deployed for 4G networks
following best practices that smoothen a later transition to
5G [71]. Due to the expected density, the optimization of
the network structure can be complex. SON aim thus to
automate the configuration of network settings, which
provides an easier deployment and operation in addition
to a better performance due to the dynamic way the
process goes. An illustration of SON deployment process
is given in Fig. 18.
Fig. 18. Ideal deployment process with closed loop automation (Source
[71])
The features provided by SON for LTE are [71]:
 Self-configuration: basic backhaul and interface
configuration, automatic inventory and software
update, automatic neighbor relation, handover paths
definition, cell identifiers assignment.
 Self-optimization: load balancing, mobility
robustness, access robustness, frequent handover
and interference mitigation, coverage capacity
optimization, signal strength and quality,
minimization of drive testing, crowd measurement
reports optimization, energy saving, handover
forwarding.
 Self-healing: cell outage detection, cell degradation
detection, cell outage recovery, cell outage
compensation, recovery from cell outage
compensation.
SON should provide in addition more robust solutions
in a 5G dense networks for interference management,
following the deployment scenarios and required QoE.
An example of already issued solutions that enable SON
is the recently published Application Programming
Interface (API) for physical layer and Medium Access
Control (MAC) layer [72].
F. Device to Device Communications (D2D)
Another approach suggested to solve high-density
cellular network is D2D communication. In fact, in
voice-centric systems, parties establishing
communication are supposed to have no spatial proximity
a priori, contrary to some situations for data
communication where several collocated devices would
like to wirelessly share content or interact. In this latter
case, it can be more efficient to adopt a Proximity-based
Device-to-device communications. Moreover, this
communication ability may be used to enhance capacity
and coverage when devices act as transmission relays and
set up multi-hop communication links. D2D plays an
important role in 5G applications. However, the
following issues have to be addressed in 5G, as they
already present challenges in 4G:
 Direct discovery: In 4G systems, the Evolved Packet
Core (EPC)-level Proximity Service discovery
solution is adopted, which relies on the location
services requiring five stages: UE registration,
proximity request, location reporting, proximity
alert and direct discovery. This implies an
increasing in complexity and delay.
 Interference management: as devices act as relays,
there are constraints on distance, scheduling and
emission power in regard to BS and primary cellular
UEs, as shown in Fig. 19.
 Direct communication: In 4G systems, physical
channels of direct communication links reuse
physical uplink shared structures, which reduces the
dynamic range requirements of power amplifiers
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1044

due to high PAPR. 5G systems on the other hand
may use frequency resources dispersed in several
frequency bands. New multi-carrier modulations
and CA techniques have to be used, and D2D
communication link in this case has to be redesigned.
Fig. 19. D2D typical scenario (source [73])
G. Fog Computing
The ultra dense networks paradigm implies a huge
amount of data to transmit. For some applications, data
needs to be centralized and may use cloud computing.
This “cloud model” can be extended to the edge Fog
networking, which is a system-level horizontal
architecture distributing resources and services anywhere
along the continuum from Cloud to Things. Fog and
Cloud fusion can provide a device and service
consolidation in addition to an optimization of life-cycle
management of tenants and virtualized services,
enhancement of data policy management and integration
of applications and data management. A unified
orchestration for Fog and Cloud can enforce service and
security and integrate the entire IoT verticals. Fog
computing nodes are typically located away from the
main cloud data centres, at the edge. Theyare wide-spread
and geographically available in large numbers and can
provide applications with awareness of device
geographical location and device context. In addition,
cloud computing on fog nodes enables low and
predictable latency. A comparison between fog and
conventional cloud is given in [74] and summarized in
Table III.
TABLE III : COMPARISON BETWEEN FOG/EDGE AND CONVENTIONAL
CLOUD COMPUTING (SOURCE:[75])
Fog computing
Cloud computing
Target users
Mobile users
Internet users
Service type
Limited localized
information
Services related to
specific deployment
locations
Global information
Collected from
worldwide
Hardware
Limited storage,
compute power And
wireless interface
Ample and scalable
Storage space and
Compute power
Distance to
Users
In the physical proximity
and Communicate
through single-hop
wireless connection
Faraway from users
And communicate
Through IP
networks
Working
Environment
Outdoor (streets,
parklands, etc.) or indoor
(restaurants, shopping
malls,etc.)
Warehouse-size
building with air
conditioning
systems
Deployment
Centralized or
distributed in regional
areas by local business
(local
telecommunication
vendor, shopping mall
retailer, etc.)
Centralized and
Maintained by
Amazon, Google,
etc.
V. CONCLUSION
5G systems present an evolution of existing systems,
including releases meant for LTE, LTE-A and LTE-A pro
as well as a revolution introducing completely new
techniques along with new use cases targeted for future
networks. The development of 5G new techniques and
standards aims to fulfill a number of requirements much
more demanding than for previous systems. Every day
brings new ideas, achievements, Proof-of-concepts and
demonstrations enabling 5G systems. This paper presents
a picture of research and standardization landscape of 5G
systems development. It also presents some of the most
important key technologies heavily studied for 5G, such
as massive MIMO, C-RAN and SON, which promise to
fulfill 5G requirements announced by International
Mobile Telecommunications (IMT)-2020 and NGMN
Alliance.
REFERENCES
[1] Ericsson Mobility Report, Tech. Rep., June 2019.
[2] Cisco, “Cisco Visual Networking Index: Global Mobile
Data Traffic Forecast Update,” 2017–2022 White Paper,
Tech. Rep., 2018.
[3] NGMN Alliance, “5G White Paper,” Tech. Rep., 2015.
[4] The 3rd Generation Partnership Project (3GPP). [Online].
Available: https://www.3gpp.org/
[5] Nokia Siemens Network, “Lte release 12 and beyond,
white paper,” Tech. Rep., Oct. 2012.
[6] M. Polese, M. Giordani, and M. Zorzi, “3GPP NR: the
standard for 5G cellular networks,” in 5G Italy White
eBook: from Research to Market, 2018.
[7] Verizon. First ever 5G NR call made today on a
commercial network with simulated smartphone device.
Press news. 09-10-2018. [Online]. Available at:
https://www.verizon.com/about/news/first-ever-5g-nr-call-
made-today-commercial-network-simulated-smartphone-
device
[8] 5G COMMERCIAL LAUNCH. 5G observatory. [Online].
Available: https://5gobservatory.eu/category/5g-
commercial-launch/
[9] N. Networks, “White paper - 5g use cases and
requirements future works,” Tech. Rep., 2014.
[10] FP7 ICT project: Mobile and Wireless Communications
Enablers for the Twenty-Twenty Information Society 5G.
[Online]. Available: https://www.metis2020.com/
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1045

[11] ZTE Corp, “Driving the convergence of the physical and
digital world - white paper on next generation mobile
technology,” Tech. Rep., 2014.
[12] Samsung Corp., “Samsung announces world’s first 5G
mmWave mobile technology,” Samsung Newsroom, May
13, 2013.
[13] Huawei, “5G: New air interface and radio access
virtualization huawei white paper,” Tech. Rep., Apr. 2015.
[14] 3GPP commits to 5G NR in unlicensed spectrum in its
next release. Press news, 13-12-2018. [Online]. Available:
https://www.qualcomm.com/news/onq/2018/12/13/3gpp-
commits-5g-nr-unlicensed-spectrum-its-next-release
[15] R. Alkhansa, H. Artail, and D. M. Gutierrez-Estevez, “Lte-
wifi carrier aggregation for future 5g systems: A feasibility
study and research challenges,” Procedia Computer
Science, vol. 34, pp. 133 – 140, 2014.
[16] M. J. Alam and M. Ma, “Resource matching in carrier
aggregation enabling 5g networks,” Wireless Personal
Communications, pp. 1–20, 2016.
[17] D. T. Do and D. A. Nguyen, “The maximal sinr selection
mode for 5G millimeter-wave MIMO: Model systems and
analysis,” Indonesian Journal of Electrical Engineering
and Computer Science, vol. 7, no. 1, pp. 150-157, July
2017.
[18] T. Rappaport, S. Sun, R. Mayzus, H. Zhao, et al.,
“Millimeter wave mobile communications for 5g cellular:
It will work!” Access, IEEE, vol. 1, pp. 335–349, 2013.
[19] (2013). nyuwireless. [Online]. Available:
http://nyuwireless.com/
[20] Huawei to bring 73GHz mmWave Mu-MIMO live demo to
Deutsche Telekom. [Online]. Available:
http://www.huawei.com/en/news/2016/2/73GHzmm-
Wave-Mu-MIM-live-demo
[21] A. Bleicher. First Intercontinental 5G Trial Begins at
Winter Olympics. [Online]. Available:
https://spectrum.ieee.org/tech-talk/telecom/wireless/first-
intercontinental-5gtrial-begins-at-winter-olympics
[22] H. Papadopoulos, C. Wang, O. Borsialoglu, X. Hou, and Y.
Kishiyama, “Massive mimo technologies and challenges
towards 5g,” IEIC Trans. Comm., vol. E99-B, no. 3, Mar.
2016.
[23] Y. Kim, H. Ji, H. Lee, J. Lee, B. L. Ng, and J. Zhang,
“Evolution beyond lte-advanced with full dimension
mimo,” in Proc. International Conference on
Communications Workshops, June 2013, pp. 111–115.
[24] Z. Gao, L. Dai, D. Mi, Z. Wang, M. A. Imran, and M. Z.
Shakir, “Mmwave massive mimo based wireless backhaul
for 5g ultra-dense network,” Nov. 2015.
[25] “5g channel model for bands up to100 ghz,” IEEEGC, Dec.
2015.
[26] J. Huang, Y. Liu, C. Wang, J. Sun and H. Xiao, “5G
millimeter wave channel sounders, measurements, and
models: Recent developments and future challenges,”
IEEE Communications Magazine, vol. 57, no. 1, pp. 138-
145, January 2019.
[27] P. Manhas and M. K Son, “OFDM system performance
evaluation under different fading channels and channel
coding using matlab simulink,” Indonesian Journal of
Electrical Engineering and Computer Science, vol. 5, no. 2,
pp. 260-266, February 2017.
[28] X. Zhang, M. Jia, L. Chen, J. Ma, and J. Qiu, “Filtered-
ofdm - enabler for flexible waveform in the 5th generation
cellular networks,” in Proc. IEEE Global Communications
Conference, Dec. 2015, pp. 1–6.
[29] A. Farhang, N. Marchetti, F. Figueiredo, and J. P. C. L.
Miranda, “Massive MIMO and waveform design for 5th
generation wireless communication systems,” CoRR, vol.
abs/1501.00305, 2015.
[30] F. Schaich, T. Wild, and Y. Chen, “Waveform contenders
for 5g - suitability for short packet and low latency trans-
missions,” in Proc. 79th Vehicular Technology Conference,
May 2014, pp. 1–5.
[31] N. Michailow, R. Datta, S. Krone, M. Lentmaier, and G.
Fettweis, “Generalized frequency division multiplexing: a
flexible multi-carrier modulation scheme for 5th generation
cellular networks,” in Proc. German Microwave
Conference, Mar. 2012.
[32] “Deliverable d2.4: Proposed solutions for new radio access
(ict-317669-metis/d2.4),” in Mobile and Wireless
Communications Enablers for the Twenty-twenty
Information Society (METIS), 2015.
[33] D. Cho, W. Lee, O. Jo, H. Lee, H. Yu, Y. Oh, and Y. Lee.
(Jul. 1, 2010). Beam division multiple access system and
method for mobile communication system. [Online].
Available: http://www.google.
com.ar/patents/US20100165914
[34] M. Al-Imari, P. Xiao, M. Imran, and R. Tafazolli, “Uplink
non-orthogonal multiple access for 5g wireless networks,”
in Proc. 11th International Symposium on Wireless
Communications Systems, Aug. 2014, pp. 781–785.
[35] S. Timotheou and I. Krikidis, “Fairness for non-orthogonal
multiple access in 5g systems,” CoRR, vol.
abs/1504.02300, 2015.
[36] A. Benjebbour, K. Saito, A. Li, Y. Kishiyama, and T.
Nakamura, “Non-orthogonal multiple access (noma):
Concept, performance evaluation and experimental trials,”
in Wincom, Nov. 2015.
[37] L. Dai, B. Wang, Y. Yuan, et al., “Non-orthogonal
multiple access for 5g: Solutions, challenges, opportunities,
and future research trends,” IEEE Communications
Magazine, vol. 53, no. 9, pp. 74–81, September 2015.
[38] V. Mojtaba, D. Zhiguo, and P. H. Vincent, Multiple Access
Techniques for 5G Wireless Networks and Beyond,
Publisher Springer International Publishing, 2019.
[39] H. Nikopour and H. Baligh, “Sparse code multiple access,”
in Proc. IEEE 24th International Symposium on Personal
Indoor and Mobile Radio Communications, Sept. 2013, pp.
332–336.
[40] M. Aldababsa, M. Toka, S. Gökçeli, G. K. Kurt, and O.
Kucur, “A tutorial on nonorthogonal multiple access for
5G and beyond,” Wireless Communications and Mobile
Computing, vol. 2018, p. 24, 2018.
[41] S. E. Hassani, M. H. Hamon, and P. Pénard, “A
comparison study of binary and non-binary ldpc codes
decoding,” in Proc. 18th International Conference on
Software, Telecommunications and Computer Networks,
Sept. 2010, pp. 355–359.
[42] E. Arikan, “Channel polarization: A method for
constructing capacity-achieving codes for symmetric
binary-input memoryless channels,” IEEE Transactions on
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1046

Information Theory, vol. 55, no. 7, pp. 3051–3073, July
2009.
[43] Fp7 project: Sparse regression codes. [Online]. Available:
https://cordis.europa.eu/project/rcn/186858/factsheet/fr
[44] Huawei achieves 27Gbps using polar codes in 5G trials.
[Online]. Available: https://disruptive.asia/huawei-27-
gbps-polar-code-5g/
[45] Design Aspects of Polar and LDPC codes for NR. [Online].
Available:http://www.3gpp.org/ftp/TSGRAN/WG1RL1/T
SGR187/Docs/R1-1611259.zip
[46] R. G. Maunder, “The 5g channel code contenders,”
AccelerComm White Paper, pp. 1–13, Aug. 2016.
[47] H. Kim, “Coding and modulation techniques for high
spectral efficiency transmission in 5g and satcom,” in Proc.
23rd European Signal Processing Conference, Aug. 2015,
pp. 2746–2750.
[48] A. Gotsis, S. Stefanatos, and A. Alexiou, “Ultradense
networks: The new wireless frontier for enabling 5G
access,” IEEE Vehicular Technology Magazine, vol. 11, no.
2, pp. 71–78, June 2016.
[49] “C-ran the road towards green ran (white paper,version
3.0),” China Mobile, Tech. Rep., dec 2013.
[50] Z. Ding, J. Li, M. Peng, and Y. Yu, Energy-Efficient Joint
Congestion Control and Resource Optimization in
Heterogeneous Cloud Radio Access Networks, CoRR,
abs/1602.05351, 2016.
[51] G. Song, W. Wang, D. Chen, and T. Jiang, “KPI/KQI-
driven coordinated multipoint in 5G: Measurements, field
trials, and technical solutions,” IEEE Wireless
Communications, vol. 25, no. 5, pp. 23-29, October 2018.
[52] L. Lingjia, Z. Jianzhong, Y. Yang, L. Hongxiang, Z.
Jinyun, “Combating interference: MU-MIMO, CoMP, and
HetNet (Invited Paper),” Journal of Communications, 2012.
[53] X. Lu, P. Wang, D. Niyato, D. I. Kim, and Z. Han,
“Wireless networks with RF energy harvesting: A
contemporary survey,” Communications Surveys Tutorials,
vol. 17, no. 2, pp. 757–789, 2015.
[54] A. Salh, L. Audah, N. S. M. Shah, S. A. Hamzah,
“Maximizing energy efficiency for consumption circuit
power in downlink massive MIMO wireless networks,”
International Journal of Electrical and Computer
Engineering, vol. 7, no. 6, pp. 2977-2985, December 2017.
[55] D. Zhang, K. Yu, Z. Zhou, and T. Sato, “Energy efficiency
scheme with cellular partition zooming for massive mimo
systems,” in Proc. IEEE Twelfth International Symposium
on Autonomous Decentralized Systems, March 2015, pp.
266–271.
[56] Ericsson, “5G energy performance, Ericsson white paper,”
Tech. Rep., Apr. 2015.
[57] “View on 5G architecture,” 5G PPP Architecture Working
Group, Jul. 2016.
[58] P. Rost, A. Banchs, I. Berberana, M. Breitbach, et al.,
“Mobile network architecture evolution towards 5G,”
IEEE Communications Magazine, vol. 54, no. 6, June 2016.
[59] System Architecture for the 5G System, (3GPP TS 23.501
version 15.3.0 Release 15) Technical specification, ETSI
TS 123 501 V15.3.0 (2018-09). [Online]. Available:
https://www.etsi.org/deliver/etsi_ts/123500_123599/12350
1/15.03.00_60/ts_123501v150300p.pdf, accessed on 18-
07-2019
[60] 5G Novel Radio Multiservice adaptive network
Architecture (5G NORMA). Deliverable d3.1 : Functional
network architecture and security requirements. Dec. 2015.
[Online]. Available: https://5gnorma.5g-ppp.eu/wp-
content/uploads/2016/11/5g_norma_d3-1.pdf
[61] P. Agyapong, M. Iwamura, D. Staehle, W. Kiess, and A.
Benjebbour, “Design considerations for a 5g network
architecture,” Communications Magazine, vol. 52, no. 11,
pp. 65–75, Nov. 2014.
[62] Z. Chang, Z. Zhou, S. Zhou, T. Ristaniemi, and Z. Niu,
“Towards service-oriented 5G: Virtualizing the networks
for everything-as-a-service,” Apr. 2016.
[63] J. L. Beylat, “The software network and virtualization
opportunities,” 2014.
[64] E. A. Y. Zaki, “Lte wireless virtualization and spectrum
management,” in Proc. Wireless and MobileNetworking
Conference, Oct. 2010, vol. 2014, pp. 1–25.
[65] ETSI, “Etsi white paper no. 11, mobile edge computing - a
key technology towards 5G,” Sep. 2015.
[66] D. Mavrakis, “Why fronthaul matters - a key foundation
for centralized and cloud ran,” Ovum White Paper, 2015.
[67] Vodafone, ericsson connect german labs in 5g mobile edge
poc. [Online]. Available:
http://www.rcrwireless.com/20160701/europe/germany-
5g-mobile-edge-poc-tag17
[68] 3GPP TS 23.501 V15.1.0. 3rd Generation Partnership
Project; Technical Specification Group Services and
System Aspects; System Architecture for the 5G System:
Stage 2 (Release 15). [Online]. Available:
http://portal.3gpp.org/webapp/meetingCalendar/MeetingDe
tails.asp?m_id=18661, accessed on 18-07-2019
[69] S. Kekki, W. Featherstone, Y. Fang, et al. MEC in 5G
networks. ETSI white paper n 28, First edition – June 2018,
[Online]. Available:
https://www.etsi.org/images/files/ETSIWhitePapers/etsi_w
p28_mec_in_5G_FINAL.pdf
[70] M. Shariat, Ö. Bulakci, A. D. Domenico, et al., “A flexible
network architecture for 5G systems,” Wireless
Communications and Mobile Computing, 2019.
[71] Small Cell FORUM, “Release 7.0: Using son in hetnet
deployments,” Jun 2016.
[72] Small Cell Forum. Small Cell Forum Publishes
Specification to Drive Unified 5G Open RAN. (July 11,
2019). [Online]. Available:
https://www.smallcellforum.org/press-releases/small-cell-
forum-publishes-specification-to-drive-unified-5g-open-
ran/
[73] V. K. Singh, H. Chawla, and V. A. Bohara, “A proof-of-
concept device-to-device communication testbed,” CoRR,
vol. abs/1601.01398, 2016.
[74] T. H. Luan, L. Gao, Z. Li, Y. Xiang, and L. Sun, “Fog
computing: Focusing on mobile users at the edge,” CoRR,
vol. abs/1502.01815, 2015.
[75] W. Xiang, K. Zheng, X. (Sherman) Shen, 5G Mobile
Communications, Springer International Publishing, 2016.
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1047

Sanae El Hassani obtained her master
engineering degree in
telecommunications from Telecom-
Bretagne, France, in 2007. She worked at
Orange Labs (France Telecom R&D)
from 2007 to 2010 as an R&D engineer
and received the PhD degree in Signal
Processing and Telecommunications
from Telecom-Bretagne & Université´ de
Rennes I in 2011. She is currently assistant professor at ENSA
El-Jadida, Chouaib Doukkali University. Her fields of interest
include advanced modulations, error correcting codes and
receiver techniques.
Abdelfatteh Haidine received his PhD
from Dresden University of Technology
in Germany. He worked as consultant and
manager with big companies (KEMA,
ACCENTURE) for the deployment of
smart metering systems and smart gird
applications. Currently he is a Lecturer at
the department of Telecoms, Networking
and Informatics at the National School of
Applied Sciences El Jadida, Morocco. His research interests
include issues related to Machine-to-Machine (M2M)
communications, net- working technologies for smart city and
smart grid applications, as well as application of combinatorial
optimization in network planning and migration and resources
allocation in mobile networks.
Hayat Jebbar obtained her master engineering degree in
telecommunications from Faculty of Sciences and Techniques,
Settat Morocco in 2016. Now she is currently preparing her
PhD degree in 5G technology and especially in physique layer
and iterative receiver in Faculty of Sciences and Techniques
Errachidia and National School of Applied Sciences El-Jadida,
Morocco. Her fields of interest include waveforms candidates
for 5G, Massive MIMO, error correcting and iterative receivers.
Journal
of Communications Vol. 14, No. 11, November 2019
©2019 Journal of Communications 1048