IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06

______________________________________________________________________________________

From 1G to 5G, What Next?

Opeoluwa Tosin Eluwole, Nsima Udoh, Mike Ojo, Chibuzo Okoro and Akintayo Johnson Akinyoade

wireless technology occurring circa every decade.

Abstract—From the era of electrical telegraphy to the Without any doubt, the evolution from 1G to 5G has been
modern internet protocol (IP)-based networks, the evolution all-encompassing, involving significant changes at an
from the first generation (1G) to the nascent, fifth generation architecture level (core network (CN) and radio access
(5G) networks has been largely gradual, sometimes meteoric.
Today, there is the ‘big data’ buzz, the internet is nearly
network (RAN)) from one generation to the next. These
ubiquitous and everything appears to have ‘smart’ capabilities have covered amongst other things, numerous changes at
given fascinating technologies such as cloud computing, physical, application, transport, session and security layers
machine-to-machine (M2M) and the internet of things (IoT). within the overarching architecture. Even at an end user
Through these eras, developments have witnessed spells of very level, mobile equipment (MEs), mobile units (MUs), mobile
low data rates (about 10 kilobits per second (kbps)) to very high stations (MSs), or user equipment (UEs) have changed
rates in a staggering magnitude of 1 terabit per second (Tbps),
which is hypothetically attainable in 5G technology. This progressively in line with the technological evolutions
survey paper presents a recapitulation of the evolution from 1G attained on the road from 1G to 5G. There have been
to 5G. It further elaborates the key capabilities and periods of bulky UEs using macro subscriber identity
performance targets of 5G; with emphases on intriguing modules (SIMs) and universal SIMs (USIMs) in comparison
characteristics (ultra-low latency, ultra-reliability, ultra- to the modern era of smart, portable devices and tablets
responsiveness, ultra-fast data rate, ultra-connectivity and
using micro and nano SIMs.
ultra-densification), and enabling technologies (software
defined networking (SDN), network functions virtualisation While 2G will arguably remain the breakthrough moment
(NFV), network slicing (NS), mobile edge computing (MEC) in the chronology of cellular wireless networking and
millimetre wave (mmWave), and massive MIMO). The current telecommunications, it may be safe to consider that the
technological trends due to the evolution are also presented, nascent 5G technology could be the game changer for the
with highlights regarding the key issues faced by professionals future, especially for companies who can leverage the
in the field and technology-savvy users in a constantly
capabilities it promises to offer, albeit there might be several
developing world. Some potential areas of research and
development (R&D) are also presented. ups and downs as the technology evolves. There is a
potential of generating revenues in excess of trillions of
Index Terms—5G, IoT, MEC, NFV, NS, R&D, SDN United States Dollars (USDs) across the technology, media
and telecommunications (TMT) sector as a whole given the
feasibility of the ability to seamlessly combine high-speed
I. INTRODUCTION multimedia with mobile portability and IoT capabilities as
5G technology continues to develop. In fact, several giants
S INCE the days of advanced mobile phone service
(AMPS), technological advances in cellular systems and
telecommunications in general have not ceased. Decade after
in the networking and telecommunications world (including
Qualcomm, Cisco, Samsung etc) have undertaken several
decade, we have witnessed periods of purely analog systems studies and provided some forecasts. Some of the study
with no data capabilities (1G), digital circuit-switched results have revealed that there would be an excess of 20
systems optimised for full-duplex communication and million 5G-related jobs by 2035 [1]. By 2020, monthly
superb voice telephony (2G), broadband and multimedia global mobile data traffic would be around 30 Exabytes
systems (3G), all-IP network revolution, and the burgeoning (EB) (in terms of scale) with approximately 75% of these
era of unified IP, massive and seamless end-to-end attributed to mobile video [2].
connectivity and mobility (5G). Notably, a critical If we consider all the happenings from 1G till date, one
examination of the timeline from 1G through 5G depicts a thing is clear; from a technological standpoint, we are
palpable trend of the birth of a new generation of cellular heading towards a truly “networked” society, where
everyone and everything is connected to everyone and
everything respectively, whilst being able to seamlessly
Manuscript received January 20, 2018; revised May 24, 2018. access or share information everywhere and every time.
O. T. Eluwole was with Underwriters Laboratories, Basingstoke, UK
and EY LLP, London, UK. He is now with BDO LLP, Southampton, UK
From the provider and consumer viewpoints, this will
(e-mail: [email protected]). generate numerous technical, commercial and management
N. Udoh was with Underwriters Laboratories, Basingstoke, UK. He is challenges on the long term realisation of the so-called
now with EY LLP, London, UK. (e-mail: [email protected]). networked society [4], [6].
M. Ojo was with the Politecnico di Milano, Milano, Italy. He is now
with the University of Pisa, Pisa, Italy (e-mail: [email protected]). The remainder of the paper is organised as follows:
C. Okoro was with John Lewis Partnership and BAE Systems, UK. He Section II provides a concise, but insightful background on
is now with HSBC, London, UK (e-mail: [email protected]). the evolution from 1G to 4G. Section III presents the nascent
A. J. Akinyoade is with the University of Rochester, New York, USA.
He is now with Visa Inc., USA. (e-mail: [email protected]). 5G technology, describing what it is and how its architecture
will pan out. The section further presents an architectural

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

comparison of 5G and prior wireless technology generations. the cellular networks. Largely, through the use of a more
Section IV presents the main features, capabilities and efficient bandwidth/spectrum allocation by way of multiple
performance targets of 5G, underlining six fascinating access schemes such as frequency division multiple access
characteristics and six enabling technologies. Sections V and (FDMA), time division multiple access (TDMA) or code
VI provide a useful insight into the main current division multiple access (CDMA), 2G communication
technological trends due to the evolution from 1G to 5G, technology was very successful and excellent for voice
highlighting some of the key challenges and concerns applications [10]. The latter could not only be digitally
associated with 5G and these technological trends. Section encrypted, but also able to provide secure short message
VII concludes the paper. service (SMS) and multimedia messaging service (MMS)
services to overcome some of the limitations of 1G. In
II. RECAPPING 1G – 4G addition, 2G was able to provide a semi-global roaming
system to foster connectivity all over the world, a feat that
A. 1G was not achieved by 1G. In practice, the 2G global system
1G represents an analog transmission technology designed for mobile communications (GSM) specification supports
to provide basic voice service. AMPS provided the first cell sizes of up to 35km using macro, micro, pico or femto
commercial cellular phone concept. The radio signals used cells [7]–[9].
are analog in nature with no data capabilities, though digital Figure 2 represents a typical 2G GSM architecture. After
signaling is used to connect the radio towers to the rest of its inception, the 2G era drastically evolved from GSM to
the telephone system by modulating (frequency modulation general packet radio service (GPRS) and enhanced data rates
(FM)) the voice calls to higher frequency of about 150MHz. for GSM evolution (EDGE) (also called pre-3G systems) in
In other words, frequency division multiplexing (FDM) is 1999, partly due to the insatiable nature of the users who
used to divide the bandwidth into specific frequencies that always wanted more in terms of data services, quality of
are assigned to individual calls. The cell size for a typical service (QoS) and throughput speeds [10]–[11].
1G network is about 2-20km [7]–[9]. Nevertheless, a 2G network, in particular GSM, has its fair
Figure 1 shows a typical 1G AMPS cellular architecture. share of drawbacks such as interference issues (including co-
The fact that 1G networks are based on analog channel interference (CCI) or adjacent channel interference
signals/protocol technology (FM) means that one common (ACI)) due to frequency reuse [10], the pulse nature of
problem would be susceptibility to interference, which TDMA and angular decay curve under unfavourable terrain,
reduces call quality. In addition, there is fundamentally lack topographic or electromagnetic conditions, which could
of security, as analog signals do not allow the cause intermittent call dropouts or total failure.
implementation of advanced encryption methods. 1G At a security level, authentication, encryption and
communication technology is mired by limited capacity anonymity form the key aspects of security provision in
(limited number of subscribers), large phone size, poor voice GSM. Although authentication and encryption are provided
quality, battery life and handover reliability (frequent call through A3, A8 (both implemented in the SIM) and A5
dropping) [7]–[9]. algorithmic mechanisms, a number of drawbacks such as
In order to summarise the wide-ranging facts regarding crypto flaws, eavesdropping attack (due to invalid security
the various cellular standards, release versions as well as the assumptions), SIM attack, false or fake base station (BS),
pros and drawbacks of each generation from a technical absence of replay protection and denial of service (DOS) are
and/or commercial standpoint, tables I and II have been put known to be shortcomings of the security arrangements in
together to provide a succinct and relevant summary of each GSM [12]–[14].
generation (i.e. 1G to 5G) in terms of nomenclature, key
features and performance targets such as peak data rate,
available standards, bandwidth, latency etc.

Fig. 1. 1G AMPS architecture; Redrawn version based on [9], [15]. Fig. 2. 2G GSM architecture, Redrawn version based on [7]–[14]

B. 2G C. 3G
2G technology brought digitisation to cellular networking 3G provides dedicated digital networks used to deliver
as it provided the first digital systems as overlays or parallels broadband/multimedia services. Figure 3 shows the 3G
to analog-based systems. 2G was able to provide a universal mobile telecommunications service (UMTS)
significantly improved voice quality and gave birth to the architecture. Driven partly by the advancement in internet
first data service offering (albeit limited) in the evolution of and IP network technology, 3G architecture provides

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

support for an enhanced data rate (throughput speed) and for packet-switched traffic with seamless mobility, QoS and
QoS. Services such as global roaming and enhanced voice minimal latency. This approach allows for the support of all
quality are important feats of the 3G technology. A slight services (data, voice, multimedia) through packet
drawback of the technology is in the area of energy connections [20]. Using only two types of nodes namely the
efficiency, 3G UEs consume significantly more power enhanced node B (eNB or eNodeB) and the mobility
compared to most 2G models and it is less economical to management entity (MME)/system architecture evolution
set-up, operate and maintain a 3G network in contrast to the gateway (SAE GW), a highly streamlined architecture (4G-
prior generation of networks [3], [11], [17]–[18]. RAN) can be defined for LTE as depicted in Figure 4 [19]–
In addition, 3G UMTS is backward compatible with prior [21].
generations of cellular wireless technologies through its In Figure 4, the UE or mobile phone is connected
ability to exist in heterogeneity with the legacy GSM or wirelessly to the eNB or 4G BS. All radio protocols,
AMPS technology. mobility management, header compression, ciphering,
The evolution from UMTS through high speed packet reliable delivery of packets and all packet retransmissions
access (HSPA) and evolved HSPA (HSPA+) further are orchestrated by the eNB, as the radio network controller
provided significantly enhanced end-to-end network (RNC) is incorporated into the latter. On the control side,
performance and eventually led to the development of the eNB incorporates functions such as admission control and
next generation of networks i.e. 4G. radio resource management (RRM). The CN is streamlined
by separating the user and control planes (UP and CP). The
eNBs can communicate with each other using an X2
interface while the eNBs can communicate with the MME in
the control plane and/or SAE in the user plane using an S1
interface. The MME/SAE is called the evolved packet core
(EPC), while for the whole system the term evolved packet
system (EPS) can also be used [19]–[20].
The capability to achieve minimal latency, advancements
in multiple input multiple output (MIMO) techniques
through various radio access technologies (RATs) such as
orthogonal frequency division multiple access (OFDMA)
Fig. 3. 3G UMTS architecture; Redrawn version based on [17] and single carrier frequency division multiple access (SC-
D. 4G FDMA) are part of the key requirements and feats of 4G
4G represents the generation of mobile cellular technology. Numerous innovative concepts such as carrier
communication technology anticipated to productively aggregation, relaying and coordinated multipoint (CoMP)
deliver the demands for broadband data transmission and transmission and/or reception (explained in more detail in
broadcasting, in addition to very high-volume voice users the next section) ultimately designed and implemented to
[3]. provide significantly improved peak data rates, support for
heterogeneous network deployment and spectrum flexibility
amongst other capabilities are wonderful accomplishments
provided by 4G technology [19], [22].
Other major enhancements brought by the 4G era are in
the areas of multicasting, and interference mitigation. 3GPP
Release 12/13 defines several key capabilities and
requirements for green computing (energy efficiency), LTE
for public safety, emergency and location services
communication (machine-type communication (MTC),
M2M and IoT); in addition to multi-broadcasting and
multicasting services i.e. evolved multimedia broadcast
multicast service (eMBMS) [23]–[25].
CCI has been a major bottleneck to achieving higher
capacity in cellular networks. However, the evolution from
1G through 4G has produced various interference
coordination schemes and interference-aware receivers
aimed at mitigating CCI. These have produced promising
performance improvements compared to receivers viewing
CCI as Additive White Gaussian Noise (AWGN). In
particular, 3GPP Release 12 implements a technique, known
as network assisted interference cancellation and
Fig. 4. 4G LTE architecture; Redrawn version based on [19], [20], [26], suppression (NAICS), through considerable improvements
[27] to intra- and inter-cell interference mitigation at the receiver
From an architecture standpoint, 4G long term evolution side. This is achieved by using advanced receivers to
(LTE) network is designed with the aim of providing support increase the degree of awareness about interfering

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

transmissions (broadcasts) with potential assistance in the communication of UE assisted mobility [36]. As it implies,
network [23]–[25]. D2D is such that there is direct communication between
It is worth mentioning that other wireless technologies devices and the exchange of UP traffic, essentially avoiding
that evolved around the 2G, 3G and 4G era include going through a network infrastructure [6]. D2D is further
Bluetooth, Wi-Fi, worldwide interoperability for microwave described in section V.
access (WiMAX) and ZigBee amongst others. Since 2G and
B. 5G architecture – Next Generation RAN (NG-RAN),
3GPP family of standards (releases), interoperability with
Non-Standalone (NSA) and Stand Alone (SA) modes
previous generations of mobile, cellular wireless
technologies has been a fundamental principle in design, From an architectural standpoint, 5G network design
development and deployment. This is reflected by the fact (governed by 3GPP Releases 14, 15 and 16, and projected to
that 4G is backward compatible and integrates with various be finalised around June 2018), is expected to reuse the 4G
wireless communications technologies ranging from 2G LTE CN i.e. EPC, being deployed either in an NSA or SA
GSM to Wi-Fi and ZigBee. mode. There is a notable CN in 5G, known as the next
generation core or nextgen core (NGC). In 4G, the LTE eNB
III. WHAT IS 5G AND HOW WILL ITS ARCHITECTURE PAN (4G BS) enables connectivity to the EPC; however, in 5G,
OUT the eLTE eNB will enable connectivity to both the EPC and
the NGC. The NGC establishes connectivity to the new
A. 5G defined radio nextgen node B (NR gNB) (5G BS) via the NG
4G was once tagged the next generation network while 5G interface [37]–[45].
has also recently been referred to as network of the future by From a RAN perspective, the overall system architecture
some scholars in the field [28]. According to [29], in 1G the for 5G i.e. NG-RAN is depicted in Figure 5. The NG-RAN
foundation of mobile telephony was established while in 2G comprises gNBs which provide the NG-RA UP and CP
mobile telephony became available for everyone. Fast-track protocol terminations towards the UE. The gNBs are
to 3G, the foundation of mobile broadband was realised and interconnected with each other and connected to the NGC
the evolution of the latter became the order of the day in 4G. via the Xn interface and the NG interface respectively. The
While it can be argued that 3G and 4G technologies main functions of the gNB include all RRM tasks (radio
connected people and partly things (objects or artefacts) in bearer control (RBC), radio admission control (RAC),
the case of 4G long term evolution (LTE) / LTE-Advanced connection mobility control (CMC), scheduling), IP header
(LTE-A), the developmental efforts to date towards the compression and encryption of user data stream, routing of
realisation of 5G makes it look increasingly indisputable that UP data towards user plane functions (UPFs), scheduling
5G will be able to connect everything, providing a seamless, and transmission of paging messages and system broadcast
coalescing connectivity fabric for at least the next decade information which originate from the access and mobility
and possibly beyond [30]–[35]. In other words, it suffices to management function (AMF), measurement and
assert that the advent of 5G will provide limitless access measurement reporting configuration for mobility and
anywhere, at anytime, for anyone, and for anything [6]. This scheduling [41].
is partly because this anticipated generation of technology, if
successful, would create a unified air interface in
establishing end-to-end connectivity between mundane
things such as smartphones, fridge, freezer, boilers, cars,
wearables, utility meters and many more [30]–[35].
To put it tersely in technical terms, 5G brings a world of
an appreciably enhanced mobile data broadband, ultra-
responsiveness, ultra-reliability, ultra-low latency, ultra-fast
data rate and enormous MTC/M2M or IoT capabilities. In
5G, the essential ingredients of radio resource allocation (a
key component of RRM) including latency, throughput,
reliability, QoS and QoE are expected to be significantly
optimised to entirely new, unprecedented levels.
Atop several key concepts, techniques and schemes such
as CoMP, SC-FDMA, OFDMA, frequency division duplex-
time division duplex carrier aggregation (FDD-TDD CA)
etc, 5G will be epitomised by filter bank multicarrier
(FBMC) [103], non-orthogonal multiple access (NOMA) Fig. 5. 5G NG-RAN; Redrawn version based on [41]

[104], beam division multiple access (BDMA) [105], In a typical 5G NSA mode shown in Figure 6, the
Universal Filtered Multi-Carrier (UFMC) [106] and multi- architecture does not really require the use of the NGC as
RAT amongst others. Some of these will be discussed in the gNB connects directly with the EPC via the UP interface.
subsequent sections of the paper. One other key aspect of 5G In this way, the gNB acts as a secondary serving cell to
is the evolution from cell centricity into device centricity, enhance throughput and capacity, whilst the eNB also
which exploits and harnesses intelligence at the device side connects to the EPC to provide CP functions e.g. paging,
(human or machine) such as via device-to-device (D2D) tracking, session and mobility management. From a CN

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

viewpoint, not depending on the NGC whilst deploying the C. How 1G, 2G, 3G, 4G and 5G architectures compare
gNB, in addition to minimal or no modification required to Using figures 1–7, table IV provides some key
the EPC makes the NSA architecture very attractive to comparisons/contrasts between the main elements of 1G,
service providers and commercial mobile network operators 2G, 3G, 4G and 5G architectures.
(MNOs) [38]–[39], [41].
The 5G SA model on the other hand, shown in Figure 7, is
such that, there is a direct connectivity between the NGC
and the gNB to facilitate the provision of both UP and CP
functions. In this way, the NGC acts as the primary CN for
establishing 4G and 5G access via connectivity with the
eLTE eNB and the gNB respectively. This is conceptually
analogous to the how the EPC establishes 3G and 4G access.
The SA mode which uses the NGC as the common CN for
all access types is the long term goal for 5G deployment as
envisioned by the stakeholders and standardisation bodies
driving 5G development and implementation. One of the
reasons for this as described by [39] is that this allows
network operators to migrate from the EPC to NGC, whilst
also providing the ability to offer high-mobility, low-latency,
access-agnostic, ultra-reliable and “follow-the-user” services
in a flexible manner.
As depicted in Figures 5, 6 and 7, the NGC has several
modules, including (but not limited to) AMF, authentication
server function (AUSF), application function (AF), session
management function (SMF), UPF, policy control function Fig. 6. 5G NSA architecture mode; Redrawn version based on [37]–[45]
(PCF), slice selection function (SSF), unified data
management/user data management (UDM), network
function repository function (NRF) and network exposure
function (NEF) [40]–[41], [49]. The modularised or layered
design of the NGC makes it possible to support various
services in a flexible manner [40]. The main functions of the
highlighted modules are presented in table III.
While there are ongoing research works, trials, debates
and considerations by various institutions and bodies geared
towards standardising and ensuring the success of 5G
technology, [50] provides some of the notable upcoming
events in the run up to the realisation of 5G technology by
2020. These include (but not limited to):
• 2018: Launch of large-scale trials of 5G systems and
technology
• 2018: 5G technology showcasing at PyeongChang 2018
Winter Olympics
• 2019: World Radio Conference (WRC) – International
agreement on radio spectrum for 5G
• 2020: Early commercial deployment of 5G systems for
selected advanced uses Fig. 7. 5G SA architecture mode; Redrawn version based on [37]–[45]
• 2020: Gigabit connectivity based on 5G technology
makes debut at Tokyo 2020 Summer Olympics
• Beyond 2020: Full commercial 5G infrastructure
deployment.
TABLE I
1G TO 5G – TIMELINE, STANDARDS, RELEASES, RATS, ETC

Generation Development Standard/Technology Protocol family/Release Modulation scheme/ Protocol/RAT Forward Error
period MIMO techniques Correction (FEC)
1G 1980s AMPS IS-95 FM FDMA -
NMT TIA-EIA 95 FSK
TACS/JTACS CDMAOne
PSTN
2G 1992 – 1997 GSM CDMAOne GMSK FDMA -
TDMA
CDMA

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

2.5G 1998 GPRS 3GPP Release 97 GMSK TDMA -

CDMA
2.75G 1999 EDGE 3GPP Release 98 8PSK TDMA -
(2.9G) EGPRS CDMA
ECSD
3G 2000 – 2001 UMTS CDMA2000 QPSK TDMA Turbo codes
IMT-2000 CDMA
3GPP Release 99 (R99) WCDMA
3GPP Release 4 EV-DO Rev. A
EV-DO Rev. B
TD-SCDMA
3.5G 2002 – 2007 HSPA: 3GPP Release 5 DL: 16QAM or EV-DO Turbo codes
HSDPA 3GPP Release 6 QPSK (HSDPA)
HSUPA
3.75G 2007 – 2008 HSPA+ 3GPP Release 7 DL: 64QAM or 2x2 EV-DO Concatenated
MIMO stream codes
UL:16QAM
3.9G 2009 – 2010 LTE 3GPP Release 8 DL: 4x4 MIMO DL: OFDMA Concatenated
3GPP Release 9 UL: 64QAM SISO UL: SC-FDMA codes
4G 2011 – 2015 LTE-A 3GPP Release 10 DL: MU-MIMO DL: OFDMA Turbo codes
IMT-Advanced (8x8) UL: SC-FDMA
3GPP Release 11 UL: SU-MIMO (4x4)
TDD
FDD
4.5G (pre- 2015 – 2016 LTE-A Evolution 3GPP Release 12 eMIMO DL: OFDMA Turbo codes
5G) LTE-M 3GPP Release 13 FD-MIMO UL: SC-FDMA
LTE-U comprising eIMT-A Elevation beam
LTE-A Pro (LTE- forming
LAA); LTE-LWA eCoMP
and LTE-MulteFire UL: SU/MU MIMO
DL: 3D-MIMO
FDD-TDD CA
256QAM
5G 2016 – Date NR? 3GPP release 14 FQAM BDMA Low-density
2020* 3GPP release 15 FBMC Multi-RAT parity check codes
3GPP release 16 Massive MIMO NOMA
ITU/IMT-2020 Advanced MIMO
*Anticipated release date

TABLE II
1G TO 5G – KEY PERFORMANCE TARGETS

Generation Data rate/Throughput* Spectrum flexibility/Channel Spectral efficiency Latency

bandwidth
1G 9.6kbps Analog UL: Analog frequency >1000ms
Up to 30KHz (radio channel) channel (BS)
DL: Analog frequency
channel (MS)
2G 64kbps 200kHz - 300–1000ms**
1.25MHz (CDMA)
2.5G 384kbps 200kHz - 600–750ms**
2.75G (2.9G) 2Mbps 200kHz - 600–750ms**
3G 2.4Mbps 5MHz - 100–500ms**
1.25MHz (CDMA) Typical 120ms
3.5G DL: 14.4Mbps 5MHz - 150–400ms**
UL: 5.76Mbps 1.25MHz (EV-DO)
3.75G DL: 28.8–168Mbps 1.4MHz – 20MHz - 100–200ms**
UL: 11.5–22Mbps <100ms
3.9G DL: 100–300Mbps Up to 20MHz (LTE) DL: 15bps/Hz UP: ~10ms (<10ms)
UL: 50–75Mbps UL: 3.75bps/Hz CP: <100ms
40-50ms**
4G, 4.5G (pre-5G) DL: 1–3Gbps Up to 100MHz (LTE-A) DL: ~30bps/Hz CP: <100ms, typically 45ms
UL: 0.5–1.5Gbps UL: ~15bps/Hz (actual)
UP: ~5ms (sub 10ms)
40-50ms**
5G 1Tbps (over 100m) Up to 100GHz DL: 30bps/Hz ≤1ms
DL: ≥20Gbps UL: 15bps/Hz
UL: ≥10Gbps
*Values denote peak or theoretical maximum potentially achievable unless stated otherwise
**
Typical/actual values for deployed networks

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

TABLE III
FUNCTIONS OF 5G NGC MODULES

Module Function

AMF UE-based authentication, authorisation and accounting (AAA)

Access control and mobility management
Orchestration of network slices selection
Controls the SMF and is independent of the access technologies
Non-access stratum (NAS) signalling termination and security
Access-stratum (AS) security control
Inter CN node signalling for mobility between 3GPP access networks
Tracking area list management for UE in idle and active mode
AUSF Data storage for UE authentication
AF Provision of packet flow information to PCF
SMF Network policy-based session management (setting up, maintaining and tearing down)
Allocation of IP addresses to UEs
Selection and control of UPF for data transport
DL data notification
Traffic steering configuration at UPF to route traffic to proper destination
UPF Configuration and location management based on service type
Packet routing and forwarding
Packet inspection of UP traffic
Traffic utilisation reporting
Anchor point for intra-/inter-RAT mobility when applicable
DL packet buffering and data notification triggering
PCF Provision of policy framework for roaming, network slicing, mobility management, session management and QoS support
SSF Selection of (network) slice (instance) to which a service must connect
UDM Storage of subscriber data and profiles of UE
NRF Provision of registration and discovery functionality i.e. assisting node in discovering network services
NEF Exposure and publishing of network data

TABLE IV
COMPARING 1G, 2G, 3G, 4G AND 5G ARCHITECTURES
1G 2G 3G 4G 5G
Element/Network AMPS GSM NR
UMTS LTE
Key element for MTSO MSC RNC EPC NGC
mobility management
Base station BTS BS or BTS NodeB eNB (eNodeB) gNB (gNodeB)
Switching Circuit Circuit Circuit / Packet Packet Packet
Interfaces Um, Abis, A, B, C, D, Iur, Iub, Iu, Cu, and Uu X2-CP, S1-UP, S1- NR – NG-CP, NG-UP, S1-
E, F, G CP, S2 UP, S1-CP, NX, Xn, other
N-interfaces e.g. N2, N3 etc
Connectivity to Via the MTSO Via the MSC Via the GMSC and GGSN Via EPC? Via EPC or NGC as per an
external network NSA/SA mode of operation
(PSTN/internet)
Applications Voice only Voice, data Voice, data, video calling Voice, data, video Voice, data, video calling,
calling, online online gaming, HD TV or
gaming, HD TV or video streaming, UHD
video streaming video, VR/AR/SR, 8K UHD
video streaming

Some of the key features, requirements, performance

IV. KEY FEATURES, CAPABILITIES AND PERFORMANCE targets and enabling technologies for 5G are discussed
TARGETS OF 5G below and further in subsequent sections of the paper.
With commercial deployments set for 2020 and beyond, A. Spectrum flexibility
5G is expected to be able to [30]–[36]: One or the key characteristic features of 5G includes
• Provide minimal or no latency, gigabit, immersive scalable and flexible bandwidths ranging from 1GHz to circa
multimedia experience (i.e. AR and VR) and tactile
100GHz. A typical 5G wireless spectrum will be in the order
internet. The latter is discussed in more detail under
or 5GHz. For 5G to be considered successful, one of the key
5G’s intriguing characteristics.
criteria will be the efficiency in providing massive system
• Meet extremely high capacity and performance
capacity. This apparently calls for network densification.
demands
There is no doubt that 5G networks will be much denser than
• Provide service continuity (high availability) in trains,
sparse and dense areas its immediate predecessor (4G) in order to deliver
• Support connectivity for user devices and IoT/M2M significantly higher cell capacities and per user data rates i.e.
devices in excess of 20 million and one trillion ultra-fast data rate for low-mobility users. In making this
respectively, at or very near, 100% reliability. become a reality, 5G would allow for highly flexible and

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

dynamic allocation of TDD (for higher frequency bands throughput to energy efficiency. On the basis of data
i.e.>10GHz in highly dense environments) transmission availability at multipoint, possible types of CoMP include
resources without any restrictions on DL and UL joint processing (JP) CoMP and beamforming-CoMP
configurations. In the same vein, flexibility would need to be (CB/CS). JP-CoMP may take forms such as joint
permitted in a dynamic manner for FDD in orchestrating transmission (JT) or dynamic cell selection (DCS) while
lower frequency bands [30]–[36]. beamforming-CoMP can be coordinated beam switching
Given the astronomical rate at which devices are being (CBS) or coordinated beamforming (CBF) (e.g. elevation
connected to each other in the modern day via the beamforming, azimuth beamforming etc). In terms of
deployment of wireless sensors or actuators for MTC/IoT, coordinated points, classes of CoMP include Intra-eNB and
5G technology will need to be able to handle the network Inter-eNB CoMP [22], [57].
demands to support new models in respect of device and Massive MIMO in conjunction with multiple access
connectivity management. Currently, about a billion of schemes, adaptive coding and modulation such as BDMA,
wireless sensors have been deployed worldwide and at this NOMA, FBMC and frequency shift keying and quadrature
rate [53], virtually everything we can ever imagine will amplitude modulation (FQAM) will help deliver
become connected sooner or later once 5G becomes fully significantly enhanced spectral efficiency (peak, average
operational. user and cell-edge), whilst also seamlessly providing data
rates in the orders of Gbps (or more) anywhere. FBMC
B. MIMO/modulation techniques
technique has been envisioned for 5G technology in order to
Conceptually, MIMO uses multiple receiving and overcome some of the shortcomings of the OFDM-based 4G
transmitting antennas and actually exploits the effects of technology, which includes out-of-band emission and side
multipath as opposed to compensating or eradicating them. lobes in the OFDM waveform. The FBMC technique
The valuable consequences of using MIMO are increased provides a more complex method to drastically diminish side
throughput and greater range of operation [54]. lobes within the spectrum [6], [58].
Beyond the technological advancements achieved by 4G,
5G design and implementation would incorporate several C. Peak spectral efficiency (PSE)
innovative techniques such as massive MIMO, enhanced According to [59], PSE (in bps/Hz), refers to the
CoMP (multi-BS cooperation) and network densification maximum data rate under ideal i.e. error-free conditions
through high density small cells [31], [34]. normalised by channel bandwidth (BW). In 5G networks,
PSE is critical in evaluating enhanced mobile broadband
(eMBB) usage situation. Assuming an antenna
configuration that supports 8 spatial layers (8 multiple
streams) in DL and 4 streams in UL, the minimum
requirements for DL and UL PSEs are 30bps/Hz and
15bps/Hz respectively. In 4G networks, SE requirements are
set between 1 – 3 bps/Hz; this is increased to 10bps/Hz in
5G and this takes the PSE achievable in 5G up to 22 times
that of a 4G LTE/LTE-A technology [30]–[35].
D. Cost, energy consumption and efficiency
Through the support of multiple RATs providing
improved network energy performance, 5G devices are
expected to be available at very low cost, with a battery life
spanning an excess of 10 years without recharging when
used to support MTC/IoT applications–critical machine-type
Fig. 8. CoMP technologies; Redrawn version based on [22], [57] communication (cMTC), ultra-reliable low-latency
Massive MIMO (also called enhanced multi-user (MU) communication (URLLC) and eMBB (discussed shortly).
MIMO, full-dimension (FD) MIMO, large-scale antenna Until recent years, the evolution of cellular mobile network
systems, very large MIMO or hyper MIMO) is a technologies from 1G through 4G has been such that
significantly enhanced form of MIMO technology which emphases have been placed largely on peak data rates. 5G is
uses a collection of antennas, orchestrated to concurrently expected to achieve significantly better energy efficiency
serve multiple tens of UEs using a one-time frequently slot compared to 3G or 4G or any previous generation of cellular
i.e. same time-frequency resource. In this way, the benefits mobile network technology [22], [30]–[35].
of MU-MIMO are expanded to a larger scale [55]–[56]. E. Intriguing characteristics of 5G
Massive MIMO is further described under 5G’s key
enabling technologies (subsection F). 1. Ultra-fast data rate
CoMP technologies (shown in Figure 8) on the other Peak data rate refers the maximum achievable data rate
hand, aid collaborative processing, transmission and (in bps) under ideal conditions i.e. absolutely error-free
reception of user data in a cellular network and can be very conditions on received data bits transferrable to a single MS
vital in achieving capacity, mobility and efficiency under maximum radio resource utilisation (barring those
improvement, ranging from spectral through cell-edge radio resources used for reference signals, guard bands,

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

guard times and physical layer synchronisation). the provision of ultra-high data rate and ultra-connectivity
Mathematically, the user peak data rate is defined as the through an extended support of conventional MBB.
product of the channel BW and the PSE [59]. Four main types of latencies can be incurred by a user
The peak data rate that can be achieved in 5G is initiating a new request in a mobile cellular network namely:
theoretically up to 1Tbps, albeit this is anticipated to be control plane, user plane, core network and internet routing
realised around 2030 [34], [60]. Practically, peak data rates latency [16], [61]. These latencies are often due to
in excess of 10Gbps may be achieved in specific scenarios associated components of delay as mentioned previously. In
such as indoor and dense outdoor environments. A range of designing cellular network architectures, two of these
10-50Gbps can be achieved for low mobility users, with ≥ latencies represent some of the key technical considerations
100Mbps cell-edge data rate guaranteed for 95% of the that need to be made in the overarching network design,
users. In urban and sub-urban environments, rates in excess these are the UP and CP latency.
of 100Mbps is attainable while about 10Mbps is possible for Assuming unloaded service conditions and an active state
IoT applications almost everywhere, including sparsely of MS, UP latency is the additional time taken in delivering
populated rural areas [30]–[35], [59]. a packet (application layer message) as it traverses from a
It is worth noting that the IMT-2020 has set minimum protocol layer 2/3 service data unit (SDU) ingress point to
requirements for peak data rates in a workable 5G network the protocol layer 2/3 egress point of the radio interface in
to be 20Gbps and 10Gbps in the DL and UL respectively either DL or UL. In contrast, CP latency is the transition
[59]. Also, [6] provides a detailed compilation/summary of time from an idle state (optimal battery efficiency) to an
5G-related activities being undertaken by several active state (start of continuous packet transfer). Both
universities, technology groups and other research latencies are critical in assessing eMBB and URLLC usage
institutions across the globe. scenarios [59].
In terms of metrics, the expected end-to-end latency for
2. Ultra-low latency
5G is envisioned to be < 5ms, which is a tenth of the typical
By definition, latency refers to the time (delay), end-to-end latency achievable by 4G. In fact, 5G is
apparently measured in seconds, between the generation anticipated to incur a UP latency of < 1ms over-the-air–this
alongside transmission of data from one device (such as a is nearly imperceptible or real-time, to the extent that it will
sensor) and the error-free reception of the data by another be able handle practically challenging low-latency services
device (such as an actuator) [29]. While throughput tells us and applications such as self-driving cars, augmented reality
how fast data is sent through a network, latency describes (AR), ultra-high definition (UHD) multimedia streaming etc
how long it takes data sent from a particular source to [30]–[35]. To summarise, at a minimum, the following
successfully reach the assigned destination (target). Broadly, latency requirements [59] have been set for 5G:
latency has four associated components namely: • 4ms for eMBB (UP)
transmission delay, queuing delay, propagation delay and • 1ms for URLLC (UP)
processing delay [16]. Extremely low latency is vital in
• 20ms (CP)
order to support 5G’s essential services or major uses cases,
which include a triad of massive machine type 3. Ultra-reliability
communication (mMTC), URLLC and eMBB [30]–[35]. According to [62], reliability can generally be viewed as
The first class of service, mMTC (also called mIoT) is the availability or provisioning of a certain level of service
designed to help achieve the ultra-connectivity (discussed approximately 100% of the time. In the context of cellular
shortly) requirement i.e. wide area coverage for up to several wireless networking, [59] defines reliability as the ability to
thousands of devices per square kilometre of coverage using transmit a specified amount of traffic within a pre-
very cost-effective software and hardware, whilst achieving determined period of time with high probability of success.
highly energy-efficient operations. The letter “m” in mMTC This is akin to the definition provided in [29], which defines
refers to massive number of devices (typically sensors and reliability as being capable of guaranteeing a successful
actuators) which are generally low cost and tend to consume message transmission/delivery within a defined latency
minimal amount of energy for prolonged battery life. budget. Ultra-reliability is vital in assessing the URLLC
Typical mMTC applications include smart agriculture, usage scenario in 5G [59].
logistics, fleet or vehicular management and so on [33]. Irrespective of the possible types of latencies that may be
Smart, MTC and IoT technologies amongst others are encountered in a cellular mobile network and the numerous
discussed further in section IV. factors which could cause impairment to reliability [63] e.g.
URLLC (also called cMTC or cIoT) is essential in order irrepressible interference, equipment failure, signal power
to achieve the ultra-reliability (discussed shortly) instability etc, 5G mobile technology is expected to provide
requirement. The letter “c” in cMTC denotes critical an ultra-high reliability and availability, guaranteeing
applications which require very high availability and successful packet delivery within 1ms or less with a
maximum reliability, incurring minimal or no latency in the probability of nearly 100% (99.9999%).
process. These applications include automated energy
distribution in smart grids, sensor networking and industrial 4. Ultra-connectivity
process control or any type of application where monitoring A wide gamut of new applications, scenarios and use
and control occur in real-time [33]. cases such as wearable devices, smart cities, homes, stores,
The third category of service, eMBB essentially relates to offices or cars, critical infrastructure, telemedicine and

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

industrial processes amongst others will enjoy seamless, Being an essential ingredient for 5G, the design goals and
ubiquitous wireless connectivity with the advent of 5G. To technical requirements of TI apparently include: 1ms round
date, 4G LTE/LTE-A has been evolving in a manner that trip latency (ultrafast reaction times), high availability
identifies the need to provide excellent coverage for mobile (carrier-grade), reliability, robustness, coexistence of
uses; new air interface technologies in 5G networks will be human-to-human (H2H), H2M and M2M, and security. If
designed to be backward compatible and interoperable with these requirements are adequately met, TI will be vital in
4G’s OFDM access technology in order to serve extreme multifarious remote applications such as robot steering,
MBB demands as well as other high-bandwidth and high- monitoring and medical surgery, education and training,
traffic utilisation scenarios [30]–[35]. servicing and decommissioning etc. Other areas where TI
Through the use of advanced D2D and mmWave will prove very useful include smart grid (supplier
technology (explained in the next subsection), 5G will foster synchronisation), telepresence, drones, self-driving cars,
the attainment of extremely low-latency, whilst also public safety communications systems, AR and virtual
providing support for enhanced connection density in excess reality (VR) amongst a host of others [36], [64]–[65].
of 106 concurrent connections per km2. This will be very
6. Ultra-densification
vital especially for big data, MTC/M2M and IoT
applications [30]–[35]. Network densification is a promising cellular technique
5G will also support mobility at speeds up to 500km/h that leverages spatial reuse to enhance coverage and
(envisioned for high-speed trains) and terminal localisation throughput for 5G cellular network. Network densification
has to do with the addition of more BSs and access points
within 1m [30]–[35]. Mobility (measured in km/h) refers to
(APs) and exploiting spatial reuse of the spectrum, thereby
the maximum MS speed at which a defined QoS can be
improving network capacity [107]. It is advocated that ultra-
achieved under a mobility interruption time of 0ms. Four
dense networks (UDNs) will be the main technology enabler
categories of mobility are defined for 5G, including for achieving the 5G requirement of 1000 times increase in
stationary (0km/h), pedestrian (0 – 10km/h), vehicular (10 – mobile network data throughput compared to LTE.
120km/h) and high-speed vehicular (120 – 500km/h) [59]. Network densification significantly reduces transmission
5. Ultra-responsiveness distance and enables proximity communication, which
makes the signal propagation to transit from long to short
Responsiveness refers to a time-based measure of the
range propagation [108]. As network densification has a
ability of a component, system, or an entire functional unit to significant impact on network capacity, it also makes
complete a specified task. In order to bring an entirely new interference more difficult to handle. We introduce in the
dimension to human-to-machine (H2M) communication in next few paragraphs different ways of combatting
5G cellular networking i.e. the ability to transmit touch and interference and further enhancing the system performance
actuation in real-time, the need for an ultra-responsive in UDNs.
connectivity becomes vital [36]. Network capacity can be enhanced via the application of
From fixed internet, through mobile and things internet techniques such as interference management, NOMA and
(IoT), the evolution of the internet has reached a new form mmWave communications. Interference cancelation and
referred to as the ‘tactile internet’ (TI), a term coined by interference coordination are the two predominant
Professor G. Fettweis some few years ago, which is now interference management techniques that have the potential
widely accepted by the community [64]–[65], [68]. to combat interference and improve network capacity [109]–
Aside from fundamentally providing ultra-fast data rate, [111]. Interference cancelation has to deal with the
ultra-low latency, ultra-reliability and ultra-connectivity rebuilding of interfering signals, decoding, whilst also
capabilities through 5G, it is vital that all the afore- removing such from the aliasing signal until the desired
mentioned are accomplished in an ultra-responsiveness signal is retrieved. Interference coordination involves the
manner in order to be able to efficiently deliver haptic use of various techniques such as beamforming, power
control, user scheduling, advanced receiver techniques etc,
experiences remotely. This is because TI centres on the
such that the desired signal and the interfering signals are
remote delivery of a physical haptic experience using an
forced to be spatially orthogonal at the receiver.
ultra-fast, ultra-reliable, and more importantly, ultra-
NOMA serves as a promising method to improve user
responsive network connectivity, so that real-time interactive connectivity and network capacity by fully multiplexing
systems which are able to steer and control both real and available spectrum via non-orthogonal spectrum sharing.
virtual objects can be built [62]. NOMA allows multiple users to share time and frequency
TI enables haptic communications by providing the resources in the same spatial layer via power domain or code
medium for transporting touch and actuation in real-time domain multiplexing. Compared to conventional orthogonal
[36]. In an interview in January 2017 [66], it was asserted multiple access technologies, NOMA can accommodate
that, “Conceivably, a person or a machine could be in one much more users via non-orthogonal resource allocation,
place, yet apply their physical skills in another place through and also allows controllable interferences to realize
the tactile internet. This possibility will enable a burst of overloading at the cost of a tolerable increase of receiver
innovation in so many aspects of our lives that we simply complexity. Therefore, the demands of spectral efficiency
cannot imagine all the applications at this point”. [67] and massive connectivity for 5G can be partially fulfilled by
upheld this by stating that “the tactile internet will be an NOMA [104] [112].
enabler for remote skillset delivery and thereby democratize mmWave communications technologies over 30-100GHz
labour and wealth globally”. would serve as a promising complementary to sub 6GHz

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

technologies in ultra-dense networks [113]. Specifically, enables the simplification of system management by
higher data rates and larger network capacity can readily be providing the ability to abstract and democratise computer
guaranteed under mmWave bands. Subsection F below resources while softwarization via programmability involves
includes addition information on mmWave. the use of software to orchestrate applications and services
F. Key enabling technologies of 5G within the network, whilst also simplifying the scaling and
management of network infrastructure.
In discussing the key enabling technologies for 5G, it is
important to have a quick recapitulation of the cloud concept 1. NFV
as this relates to a fair share of 5G’s key technology NFV leverages central processing unit (CPU)
components. virtualisation and other cloud computing technologies to
A cloud can be referred to as a huge pool of highly establish network functions migration from dedicated
scalable, dynamically reconfigurable, on-demand virtualised hardware to virtual machines (VMs), effectively reducing
computer resources such as storage, processing power, hardware footprint [73]. Implemented through the virtual
input/output devices memory. These resources are usually network function (VNF) (a software function), NFV enables
provided by a harmony of hardware and software platforms, the separation of network functions from hardware
orchestrated in a manner that an optimum resource infrastructure so that they can be managed as a software
utilisation can be achieved at scale. The main characteristics module deployed in a cloud computing infrastructure,
of a cloud or cloud computing system include on-demand effectively providing greater degree of abstraction and
self-service (autonomous provisioning), broadband network increasing overall network flexibility. The VNF software
access (availability for heterogeneity), rapid elasticity undertakes the task of handling specific network functions
(scalability), resource pooling (dynamic resource allocation), that run on the VMs on top of the hardware networking
and measured service level agreements (SLAs) [69]–[71]. infrastructure such as routers, switches or gateways [72],
Nearly every organisation or business nowadays including [74]–[75].
content delivery and social platform giants such as Netflix,
Facebook, Twitter, Instagram etc, has one form of cloud- 2. SDN
based, technology-enabled service or the other including SDN is a networking paradigm which is typically
servers, storage and applications from a cloud service implemented in the CN. SDN establishes a decoupling of
provider delivered to its information technology (IT) control and data planes (CP and DP), enabling direct
landscape via the internet as needed and paid for on a pay- programmability of network control via software-based
as-you-use basis. Microsoft Azure is a popular cloud controllers. This changes the limitations of current network
platform which provides integrated cloud services. Through infrastructures by decoupling the forwarding plane (DP) and
a provisioned cloud infrastructure at a data centre and the network’s control logic (control plane) traditionally
following a public, private, community or hybrid coupled with one another. The control plane is implemented
deployment, cloud services can be delivered through three in a logically centralised controller (or network operating
basic service models, including software-as-a-service (SaaS) system), simplifying policy enforcement and network
e.g. Salesforce.com, platform-as-a-service (PaaS) e.g. configuration and evolution, and moreover reducing
Google App engine and infrastructure-as-a-service (IaaS) hardware footprint. User data and system control separation
e.g. Amazon web services [69]–[71]. in SDN helps to deliver high degree of device-centric
From the manifold research undertaken to date, it is clear optimisation of the active radio links in the network [73],
that the design and implementation of 5G will encompass [75]. SDN can also be extended to the RAN in the form of a
emerging, complementary and supporting technologies such self-organising network (SON) [62].
as SDN, NFV, NS and MEC amongst others, in order to NFV can serve SDN by virtualising the SDN controller to
provide virtualisation and programmability of network, be rendered in the cloud, thus allowing dynamic migration of
services and control functions. A common theme, called the controllers to the optimal locations while SDN can serve
“softwarization” has been coined in recent years, to depict NFV by providing programmable network connectivity
the systemic evolution and introduction of concepts and between VNFs to achieve optimised traffic engineering.
technologies such as SDN, NFV and MEC to the wider, Both technologies are related and paired to each other, but
traditional telecommunication systems landscape [72], [74]. are based upon differing standards with no combined
This is possibly rightly so because these technologies standardised architecture. There are a number of technical
involve tactical use of software to perform myriads of aspects which impact several network elements concerning
functions within the end-to-end network domain across all SDN and NFV. Several key areas need to be addressed for
network layers ranging from the services to the application standardisation in order to accelerate the adoption of
layer. Perhaps, it suffices to envision that MEC, SDN, NFV networks evolving with SDN and NFV. Several
(as will be discussed shortly) and related concepts would standardisation bodies like ETSI, IETF, ONF, 3GPP, and
serve as the platform where core IT and traditional IEEE itself are involved in standardising different technical
telecommunications converge [73]. aspects [5].
From a converged network viewpoint, the key Furthermore, both technologies can be adopted in the
performance targets of 5G such as ultra- low latency, mobile packet core (MPC) network architectures such as the
reliability and connectivity will be achieved through network EPC, which is the most recent MPC network that represents
densification, virtualisation and optimisation. Virtualisation the core of the LTE system. SDN and NFV can be adopted

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

in the main functional entities of the EPC, such as the and machine learning (ML) applications), achieving better
serving gateway (S-GW), packet data network gateway (P- flexibility and performance, whilst also extending battery
GW), the MME, the home subscriber server (HSS), and the life. MEC helps to add intelligent capability to the
policy control and charging rules function (PCRF) [5]. traditional RAN typically used for voice calls and data in
mobile cellular works through the overlay of distributed
3. NS
edge cloud computing onto it, essentially turning it into a
By definition, a network slice can be viewed as an cloud RAN (C-RAN) to provide edge intelligence [28], [36],
autonomous logical network, created by the [75], [77].
interconnectivity of a subset of required building blocks (CP The C-RAN concept involves transferring the baseband
and DP elements) that can be autonomously instantiated and processing units from cell sites to a centralised location to
operated over a physical or virtual infrastructure [76]. In serve a wide area via fronthaul in order to reduce the
other words, with the NS concept, multiple different logical hardware footprint at the cell site, whilst more importantly,
network architectures (i.e. slices) can be defined on top of minimising latency [35], [78]. In conjunction with SDN and
the same physical (IP services) infrastructure so that NFV, MEC is an essential ingredient in achieving extremely
resources may be dedicated wholly to a single slice or shared low latency, very high bandwidth and speed in 5G
among different slices. Practically, the network slice would technology, which aims to achieve seamless
be defined across the entire communication system (i.e. end- interconnectivity of trillions of devices [28].
to-end), including both AN and CN functions alongside their
corresponding nodes and end-systems such as MEs [62]. 5. mmWave
NFV and SDN in conjunction with enhanced analytic A 5G small cell deployment in up to 100GHz band using
tools for RRM will provide autonomous optimisation a 2GHz carrier BW” is considered to represent an mmWave
capabilities to MNOs by following well-defined policy [34]. In 4G, spectrum flexibility (channel BW) is scalable up
controls [35]. With a much greater degree of abstraction to 20MHz; this will be significantly increased to about
through the unification of SDN, NFV and SON, 5G will be 100GHz in 5G technology with the advent of the mmWave
able to offer a NS capability i.e. provision of network-on- technology. MmWave carrier frequencies enable larger BW
demand functions through the delivery of connectivity allocations, which directly translate to higher data transfer
services based on various custom SDN-established functions rates. A resultant effect will be an increase in data capacity
that control availability, geographical coverage area, alongside a reduction in the latency for digital traffic as the
robustness, capacity and security in a dynamic and flexible RF channel BW is increased [80].
manner [62], [75]. One of the key recommendations for 5G deployment
centres on establishing multi-connectivity between LTE-A,
4. MEC
centimetre wave (cmWave) and mmWave, which can
MEC has been identified as a key technology and appreciably improve cell-edge performance (with
architectural model for enabling the transformation to 5G, as deployment inter-site distance of about 100m), whilst
well as providing an environment for innovation and value reducing the required density for small cell deployment. An
creation [77]. MEC ensures compute, storage and network mmWave radio can provide significantly enhanced capacity
resources (e.g. MEC application server) are seamlessly in the region of several Tbps/km2 due to additional carrier
integrated with the BS to orchestrate computationally- BW and multiple sectorized antennas orchestrated for the
intensive and latency-sensitive applications such as AR and mmWave APs. As frequency is inversely proportional to
VR [28]. MEC is about providing IT and cloud computing wavelength (in the “Golden Rule” for waves); having a small
capabilities within the RAN in close contiguity to mobile wavelength in an mmWave frequency system facilitates the
subscribers. With MEC, mobile BSs are transformed into design and orchestration of massive arrays of antennas (e.g.
smart or intelligent service hubs via the exploitation of via CoMP-based MIMO), providing high beamforming
context, contiguity, agility and speed. As a result, a wide gains (via adaptive beamforming, relaying and inter-cell
range of services such as unified mobile communications, interference mitigation) necessary to combat propagation
RAN-aware context optimisation, IoT, AR, video analytics, loss in the mmWave band [31], [34], [80]–[82].
distributed content and domain name system (DNS) caching
etc, can be provided with minimal latency, high rate of data 6. Massive MIMO
processing, streamlined network performance and enhanced Massive MIMO has been recognised as a promising
QoE [36], [77]. technology to meet the demand for higher data capacity for
Through the provision of a highly distributed computing 5G networks by 2020 and beyond. As alluded to earlier in
environment, MEC helps to bring the cloud’s concept the paper, massive MIMO is a communication system where
(dynamic resource management capability) closer to the a BS having multiple hundreds (or possibly thousands) of
edge of the network (RAN), simplifying processing and antenna arrays concurrently serve many tens of user
storage of content (information) and bringing the latter in terminals, each having a single antenna, in the same time-
close propinquity to mobile users. This effectively means frequency resource [55], [127]. The BS with multiple
that users’ growing reliance on mobile devices in antennas sends independent data streams to multiple
terminals in the same time-frequency resource.
undertaking compute- and storage-intensive applications are
Massive MIMO relies on spatial multiplexing, which in
able to be offloaded to happen in the cloud (e.g. the MEC
turn relies on the BS having good enough channel
server platform would host software for real-time analytics
knowledge of both the UL and the DL. On the UL, this is

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

easy to accomplish by having the terminals send pilots, information [85].

based on which the BS estimates the channel responses to The fundamental components of an M2M system
each of the terminals. However, it is more challenging to therefore includes sensors, radio frequency identification
implement massive MIMO on the DL. (RFID), some form of wired and/or wireless communication
The basic advantages offered by massive MIMO are link (e.g. Ethernet, Wi-Fi, cellular link) and an autonomous
multiplexing gain [55], [127], energy efficiency [128], computing software set to assist a networked device
spectral efficiency [128], [129], increased robustness and manipulate and make meaningful decisions of captured
reliability [130], [131], simple linear processing [130], [132] events. Using these components, an M2M system can be
and cost reduction in RF power components [131]. built explicitly to undertake a specific task and/or utilise a
particular device [84]–[86].
V. SOME CURRENT TECHNOLOGICAL TRENDS DUE TO THE
The ETSI M2M architecture describes two main domains
EVOLUTION TO 5G
in a typical M2M service construct: the M2M device domain
A. Smart, M2M and IoT (includes the area network) and the M2M network domain
(includes the core network) which are connected via
Smart gateways. In providing service capabilities, the latter manage
In English grammar, the word smart has numerous the M2M area networks of the M2M devices following an
synonyms such as clever, quick-witted, wise, intelligent etc. established communication with the M2M core network
Technically, smart basically means harmonising or bringing applications. Communication infrastructure, service
together (i.e. seamlessly integrating) a miscellany of functions and application logic are essential to building an
components such as transducers (sensors and/or actuators), M2M solution that is able to provide service capabilities
monitoring systems, automated controls, modelling systems, such as remote entity management and transaction
decision-support applications etc more ‘intelligently’ in a bid management amongst others [85].
to ultimately accomplish massive improvements in At the media access control/physical layer, 2G/3G/LTE
performance, flexibility and scalability. Therefore, the so- cellular communication technology protocols are
called smart apparently derives input from somewhere (a consequential for M2M development and deployment. They
unit, application or system), applies some wits and then form the fundamental building block upon which the
takes actions accordingly [82]–[83]. network, transport and service layers could be built for a
Interestingly, [83] asserts that, “If a machine/artefact does complete M2M system. LTE-A 3GPP Release10
something that we think an intelligent person can do, then specifications includes the MTC architecture, features, and
we consider the machine to smart”. In other words, being standards [85].
smart entails having an inherent capability to carry out The state-of-the-art architecture, recent advances and
sensing, capturing, monitoring, adapting, inferring, relaying, challenges in communication technologies for M2M are
learning, retrieving, anticipating, translating, self-creating, described in [114]. The authors in [114] also describe the
self-organising and self-sustaining tasks, since these actions architectural enhancements and novel techniques expected in
require some sort of intelligence or brainpower, albeit at 5G networks related to M2M. Joint use of carrier
varying degrees. In our world today, we have smart devices aggregation and relay station in OFDMA-based 5G wireless
(e.g. phones and tablets), cars, systems, homes, is proposed in [115]. Latest advances and developments in
environments and even cities. architecture, protocols, standards and security for M2M
Smart homes, smart cities and smart grids increase dense evolution from 4G to 5G wireless are discussed in [116].
and diverse connectivity. Practically, these translate into Moreover, the authors in [117] present an overview of
very high data usage levels, which is extremely daunting in several features introduced in 3GPP Releases 13 and 14 to
address the requirements of M2M communications together
4G LTE cellular systems. 5G is envisioned be able to
with the new air interface characteristics associated with the
provide greater spectrum availability at untapped mmWave
new UE categories. We expect the cognitive radio to emerge
spectrum. In addition, 5G networks are expected to provide
and assist in developing novel cognitive M2M architecture
cumulative capacity for multiple simultaneous users in both for sensing and using the available frequency bands [118].
unlicensed and licensed spectrum [80], providing required
capabilities to address enormous data rates and connectivity IoT
in smart cities, homes and grids [133]. Historically, the IoT concept was first alluded to by
researchers at Carnegie-Mellon University in 1982 having
M2M
successfully connected a vending machine
M2M, sometimes called MTC, is a way of establishing (https://thefutureofsmart.wordpress.com/2014/10/04/the-
direct communication between devices or objects using any first-iot-device-a-coke-machine/) to the Internet. Further
one of, or both, wired or wireless communication channels IoT-related development work on ‘ubiquitous computing’
[84]. It refers to communication among devices (e.g. a was undertaken in the early 1990’s [87] prior to the eventual
sensor, actuator or meter) with minimal or no human official naming and formulation of the term/acronym IoT by
intervention in the process of capturing an event or status Kevin Ashton (Executive Director of Auto-ID Labs at MIT)
(e.g. humidity or temperature related data), which is then whilst making a presentation for Procter & Gamble in 1999
transmitted via a connection-oriented and/or connectionless [88]–[90].
network to an application (software program) which is able The IoT represents a network of connected devices which
to process and translate the captured event into meaningful are able to communicate over the Internet and doing so

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

autonomously without the need for human intervention. IoT opportunities to individuals and organisations across the
is simply about everything being interconnected and world, whilst also providing solutions to myriads of
interrelated. The IoT depicts a world of universally problems which cut across multiple, if not, all industries and
connected objects, things and processes which collectively sectors, ranging from utility to healthcare sector.
creates a digital characteristic of the people and their Applications of M2M technology can be seen in remote
surroundings in real-time, ultimately providing the best monitoring e.g. vending machine and telemetry applications,
living experience. Simply put, IoT is any IT design, build warehouse management, robotics, logistics and supply chain
and implementation having an overarching goal of providing management etc [84], [91]. Also, in the manufacturing,
increased connectivity of people and things. The IoT allows logistic or supply chain sector, having a well-established IoT
the connectivity of smart objects to the internet with a system provides the ability to facilitate document and fleet
distinguishing capability to enable an exchange of data never maintenance as and when needed. This helps to
previously available, whilst bringing information to users in simultaneously achieve reduced costs, increased efficiency
a much more efficient manner [53], [91]. and reliability [92].
With a quintet of technology components, specifically From smart devices such as wearable technology
sensors (RFID), connectivity, infrastructure platform (for products, phones or tablets, to smart cars (adaptive cruise
storing /processing data), analytics platform (for making control systems, self-parking etc), smart homes (lighting,
informed decision of data) and user interface (for presenting controlled heating, door and window solutions such as Nest
data), an IoT product can be designed, built and and Hive) and smart cities (intelligent road networks and
implemented [53], [89]. By using various types of wireless parking bays etc.), the IoT and smart technologies are indeed
technologies such as 3GPP LTE and LTE-A, Wi-Fi, ZigBee turning the world into a ‘global village’. As alluded to earlier
and Bluetooth, seamless connectivity between IoT user-end in the paper, 5G is expected to achieve the ‘real’ networked
devices and other IoT gateways or servers can be established society, where seamless connectivity is attainable
[86], [102]. everywhere and every time for everyone and everything.
As of 2015, a typical smart phone had about a dozen It is almost impossible to defy the good tidings brought to
sensors for orchestrating various IoT functionalities such as humanity by the advent of the internet since the late 1960s;
temperature, inclination, humidity etc. As of today, fewer in fact, we can easily come up with a seemingly endless list
than 1 billion sensors have been deployed globally for of the benefits and logical deliberations in favour of the
establishing IoT applications [53]. However, research internet. Fast forward to 1999, the world wide web was
undertaken by multiple institutions and research invented, and by all accounts, the term IoT soon became a
organisations have predicted that there will be up to 50 formality.
billion connected devices by 2020 and an excess of 30EB of There is a potential capability to integrate sensor data
mobile data traffic every calendar month across the globe. from numerous distributed artefacts through the use of
smart, M2M and IoT technologies, which in turn provide the
Examining Smart, M2M and the IoT in today’s world –
ability to thwart crime and asymmetric warfare. In addition,
benefits, impacts and concerns
an IoT-enabled pervasive positioning technology can help
Although there are obvious discrepancies, some of the trace lost and stolen goods from a superstore. Popular
real-life applications and connectivity nature of smart, M2M demand together with advances in smart and M2M
and IoT technologies appear akin to each other in a manner technologies could drive extensive dispersion of the IoT that
that is sometimes nearly inseparable, and therefore we may could, akin to the contemporary internet, yield invaluable
collectively refer to them as a ‘triplet’ concept. For contributions to the global economic landscape. However, to
example, while a smart technology has an intrinsic capability the extent that commonplace objects turn out to be IT
to automatically adapt its behaviour to fit an environment, it security risks, the IoT appears to have the propensity to
practically uses sensors, databases, wireless access etc to distribute such risks much more rapidly and extensively than
collaboratively sense, adapt and provide information to users the internet has to date [92].
within the surroundings akin to the IoT technology [82]. According to [93]: “Certainly, the internet has moved the
A common theme apparently exists between the IoT and world forward; also, as a result of using the Internet, a lot
M2M systems– minimal or no need for human intervention. has gone wrong, quite a lot is going wrong, and a lot more
In fact, it can be rightly or wrongly argued that M2M forms could go wrong. Therefore, it is crucial to develop a balance
the basis and serves as an enabler for the IoT technology. It on how to manage the deployment and efficient use of the
can also be argued that a product or device with in-built internet so that it can be fit for purpose”. Unsurprisingly, the
M2M capabilities can be considered as being smart. IoT has actually been named as one of the top six disruptive
According to [86], IoT networks are also known as M2M civil technologies out of a bucket of 102 potentially
area networks. There is often, if not always, a constant link disruptive technologies identified at the outset of the study
amongst the triad of technologies either in terms of undertaken in [92]. In the report, it was explicitly stated that:
connectivity, remote access/control, or real-world “By 2025 internet nodes may reside in everyday things–food
applications and, therefore they may be used packages, furniture, paper documents, and more. Today’s
interchangeably. developments point to future opportunities and risks that will
Smart, M2M and IoT technologies have certainly arise when people can remotely control, locate, and monitor
revolutionised the way we live today. On a commercial note, even the most mundane devices and articles”.
they have brought newer, intriguing and fantastic As with every new or emerging technology, challenges,

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

threats and real risks, particularly to smart and IoT devices of today are able to tell us estimated time of arrival
advancements do exist. In a world of connected objects at an to reach our homes, offices or attend a concert. We get
unprecedented scale, there is bound to be concerns and reminders and alerts informing us when it is best to leave for
challenges. From a technical standpoint, interoperability, the airport in order to meet a flight in time, or some
interusability, latency, reliability, energy efficiency (most wearable device telling us how well our hearts have
IoT devices are battery-powered) and synchronisation issues performed in a particular day or over a period of time.
have been found to affect the design and deployment of IoT In simple terms, big data refers to huge and complex sets
systems [53]. Above all, there is a palpable sense of of data whose scale is in the order of quintillion (1018) bytes
jeopardy to data privacy and confidentiality, which is i.e. EBs or higher. The sheer scale or volume of data
understandable because the torrent of data generated via the generated on a daily basis in the modern world has gone
interconnectivity of smart and IoT-enabled devices in the beyond the levels of megabytes (MBs), gigabytes (GBs),
modern society has an apparent tendency to reveal so much terabytes (TBs) and petabytes (PBs) to a few EBs, and as a
about people than is currently being envisaged by the result, it is becoming increasingly difficult for traditional
advocates and developers of these technologies. data management tools and technologies to efficiently
handle such scales of data [94], [96].
B. D2D
More importantly, we use the term big data when a quintet
D2D communication is a promising technique to improve of unique features or characteristics (i.e. volume, velocity,
resource utilisation in 5G cellular networks by offloading the variety, variability and veracity – the five Vs) can be
traffic from backhaul to local direct links. D2D wireless
attributed to a particular dataset. Volume apparently
networks are considered one of the candidates for future 5G
indicates the massive amount of data which need to be
networks. A comprehensive review of D2D communications
seamlessly manipulated or handled to maintain high
is available in [119]. Similarly, a survey of the existing
methodologies related to aspects such as interference availability for storage, processing and retrieval. Velocity
management, network discovery, proximity services, and relates to the speed of data ingress and egress which could
network security in D2D networks is presented in [120]. In complicate the processing and analyses of the data from a
the next few paragraphs, we briefly highlight the major resource management (load balancing) viewpoint. Variety
research works relevant to the context of emerging 5G and variability centre on the range and types of data sources
wireless communications. which could be too varied or great to assimilate, prove,
Major recent research activities in D2D include game transform or track for structural consistency and future
theoretic pricing schemes [121], channel measurement and reuse. Veracity focuses on ensuring ingested, processed,
modelling [122], proximity services (ProSe) [123], and analysed or manipulated data is of great and proven quality
public safety networks [119]. (i.e. sufficiently meets referential integrity, consistency,
An overview of the major challenges in a two-tier cellular validation, reconciliation and provenance tests) given the
network that involves a macrocell tier (i.e. BS-to-device high-volume, high-velocity and high-variety characteristics
communications) and a device tier (i.e. D2D) with some of the data [94]–[97].
proposed pricing schemes for different types of device Enterprises are increasingly investing in big data analytics
relaying are discussed in [121]. Scheduling algorithms for platforms to identify market trends and gaps in their bids to
effectively sharing multimedia content using D2D remain highly competitive amongst their peers. However,
communications are proposed in [124]. In addition, [125] translating real world ubiquitous data into useful, meaningful
describes Network Assisted Routing (NAR) algorithm for formats suitable for orchestrating a prescriptive, predictive,
D2D communication in 5G cellular architectures with the diagnostic or descriptive analytics may be very challenging
goal to extend the coverage of BSs. given the multiple disparate sources of the data, reflected in
More importantly, extremely low latency, high energy fundamental discrepancies in models or structures, which
efficiency and scalability are vital to 5G networks. Thus, it is could range from schema-free to flat, relational, nested,
essential to decrease the control signalling and end-to-end object-oriented, hierarchical, or non-flat data and so on.
latency in network assisted D2D communications [126]. Therefore, with big data there are bound to be data quality
C. Big Data – Definition, Impacts and Challenges issues and challenges. This is particularly so considering the
vastly complex and heterogeneous environments of most
With the propagation of smart, M2M and IoT
enterprises looking to utilise and maximise the power of big
technologies, perhaps, unsurprisingly big data comes on the
data analytics solutions [94]–[97].
scene. There is basically a cornucopia of ‘real-world’
While there are several software platforms, packages,
ubiquitous data about people and what they do, applications,
tools and database technology systems designed and used for
artefacts and things, albeit the data can be characteristically
big data analytics e.g. MS Azure, Cloudera, Apache
viewed as being dirty, noisy, poorly defined, redundant,
Hadoop, Couchbase, MongoDB, RStudio etc, in addition to
unstructured, semi-structured, multidimensional, disparate
various algorithms and techniques suitable for handling the
and volatile, to mention a few [94].
five Vs e.g. statistical machine learning (SML), neural
In today’s world of comprehensive data, there is a
networks (NN), Bayesian, Bandit, principal component
perceived universal picture of the physical world more than
analysis (PCA), support vector machine (SVM), and single
ever previously attainable with the advent of computing and
value decomposition (SVD) amongst others [94]; there is no
internet technologies such as the cloud and IoT.
commercial package till date that is able to provide an all-
Interestingly, our smart, IoT-enabled mobile phones and
encompassing ‘out-of-the-box’ solution to the main

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

problems in big data and data analytics (i.e. prescriptive, ultra-fast data rate systems, whilst also providing minimal
predictive, diagnostic or descriptive) in general. These key complexity and highly accurate CSI quantization remains an
problems relate to areas including (but not limited to) ongoing research challenge [31], [81], [98].
classification (categorisation), regression, clustering, Also, due to the growing trend of bringing intelligence in
dimensionality reduction, multi-collinearity, proximity to the edge of cellular networks (MEC, C-RAN,
reproducibility/replicability, hyper-parameter optimisation, SDN, NFV, NS), cutting-edge networking design and
algorithmic computability, pre-/post-processing of covariates approaches will be required in overcoming several
(input variables) and target responses (output variables) challenges [36], [62], [65] such as, providing 1ms round-trip
amongst others [94]–[95]. Extensive experimentations latency and an enhanced haptic perception (multisensory or
would be vital in tackling these challenges which could aid multimodal feedback), establishing the optimal waveform
the development of quantitative and qualitative techniques selection and robust modulation methodologies (physical
and processes for managing big data. layer challenges), designing CP and UP separation and
Owing to an incessant proliferation of data from smart, coordination techniques, building optimal resource
M2M, IoT and cloud technologies, which may be management and task allocations schemes (online or offline
collectively viewed as a form of technology convergence, it scheduling). In resolving some of these challenges, R&D
is highly unlikely that R&D into big data solutions into artificial intelligence (advanced machine learning and
(techniques, methodologies, platforms etc) will cease in the predictive analytics) will be vital as several critical factors
near future. will need to be considered in designing workable 5G-
enabled solutions. Some of the factors include the control
VI. CHALLENGES OF 5G AND RADICALLY CHANGING server and processing mechanisms for sensors and actuators
TECHNOLOGIES IN THE MODERN ERA (smart, MTC, IoT and MEC technologies), real-time
operating system for ultra-low latency capability, and
A. Performance multithreading capability for multiple transmissions over
In order for 5G technology to be successful, [98] parallel communication channels.
highlights the need to move from a traditionally-driven Typically, cloud computing technologies are known to
viewpoints of coverage, reliability, reduced capital attract technical challenges in the areas of data migration and
expenditure (CAPEX) and operational expenditure (OPEX) integration, resource management (due to dynamically
to a more value creation and SLA perception. It further adds changing and varying workloads), data privacy and security
that a fundamental change in design approach from a (data centre location based on deployment model), QoS and
predominant measured key performance indicator (KPI) performance. As cloud technologies and computing
(QoS-centric) to a more perceived KPI (QoE-centric) will be methodologies continue to proliferate and now being
important in achieving 5G’s key goals and performance introduced to the budding 5G wireless technology, therefore
targets. R&D into several innovative techniques for orchestration
Providing concurrent seamless wireless communications, and management (including QoS and QoE) will be
ultra-connectivity, mobility on demand, immersive necessary. These will include designing and implementing
multimedia experience, ultra-responsiveness (TI), MEC, NS, novel methods for establishing optimal load balancing
and NFV capabilities etc – (the intriguing characteristics and required for user experience optimisation and network
enabling technologies of 5G) – will encounter technical capacity given the possibility of multiple users around a
hurdles and performance bottlenecks. There are bound to be boundary zone having a great network coverage, but sub-
challenges in the design, build and implementation for a optimal requested data rate [34]. In addition, to ensure
sustainable anticipated 5G ecosystem (2020) and beyond. efficient QoS and QoE, new techniques and solutions will be
The role of R&D in resolving these challenges cannot be required such as defining resilient, error-free thresholds for
overemphasised. detecting, monitoring and correlating traffic, whilst also
For example, as promising as massive MIMO is (given ensuring the network’s capability to achieve best possible
the capabilities it is able to offer); it attracts several performance (in respect of latency, throughput, spectral
challenges (e.g. scheduling complexity, link adaption efficiency, error-rate, connectivity management etc) without
problems etc), which require substantial experimental efforts compromising security.
to overcome. Although work has been carried out by experts
in the field, some of these challenges are exacerbated as the B. Trust, privacy and security
number of antennas at the BSs increases. This is partly due Security is vital in any technological feat achieved and
to the fact that the accuracy of the channel state information will arguably remain the biggest concern in the TMT sector
(CSI) is a prerequisite for achieving reasonable performance and the wider IT industry in the modern era. In [99], it was
improvements from a massive MIMO system. The CSI has considered that a technology may not be regarded as top-
direct impact on precoding design, coding or modulation class until it amply and capably fulfils all the core
scheme (definition, acquisition, or assignment), signal requirements of modern cryptography, including
modulation and demodulation. These factors can result in confidentiality, authentication, integrity and non-repudiation.
FDD challenges such as pilot and feedback overheads, CSI Security becomes particularly vital in 5G technology given
estimation complexity and quantization. While several the major use cases (mMTC, URLLC and eMBB) the
researchers have undertaken a detailed study of massive technology is envisioned to propagate.
MIMO techniques, applying and adapting the concept in The introduction of cloud computing, virtualisation,

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

softwarization etc to telecommunications or cellular capabilities and performance targets of 5G is to provide

networking will incur new, significant challenges. Therefore, support for low-cost, minimal energy consuming devices (a
R&D in areas such SDN, NFV, MEC, NS etc will be prolonged battery life without having to recharge). Energy
fundamental to 5G’s success. Admittedly, virtualisation will efficiency of the 5G network is an essential ingredient in
enable flexible slicing, mobility management and the ability lowering operational costs which in turn could significantly
to customise radio resources for different applications and increase the reach of wireless connectivity to very remote
UEs [62] while softwarization will help achieve low costs areas [33]. Energy efficiency has been gaining more
(streamlined CAPEX and OPEX), since it will be possible attention lately, [22] proposes a new power consumption
establish ubiquitous connectivity using minimal physical model for LTE/LTE-A wireless network, however further
infrastructure, increasing cost competitiveness between R&D need to be undertaken in this area (such as designing
network operators (including mobile and virtual operators). and creating dynamically controlled, adaptive machine
However, new methodologies will need to be developed to learning-based power management systems [98]) in order to
ensure an assuring security can be provided, whilst also achieve an all-round success of 5G wireless access by 2020
taking into consideration the potential consequential impacts and beyond.
from both socio-economic and techno-economic
perspectives. VII. CONCLUSION
In underlining the need for privacy and security in 5G Having recapped the key highlights from 1G through 4G
networks, [32] talks about the need for the development of technology, this paper has presented some of the salient
innovative trust and service delivery models to handle points regarding 5G technology in terms of its key features,
radically changing and evolving threat landscape, capabilities, performance targets and technology
highlighting several key areas of R&D with regards to 5G components. Key fascinating characteristics and enabling
security. These include identity management, radio network technologies have been underlined. To a reasonable extent,
security, security assurance, energy-efficient security, cloud it may be inferred that 5G is the next big thing in cellular
security, and flexible/scalable security architecture. In and wireless systems and telecommunications in general.
addition to these, the sought solutions would need to take Aside from the inadequately resolved, unaddressed
into consideration all the germane standards and regulations technical challenges, and the limitations of the prior
(e.g. data protection act (DPA), general data protection generation of technologies, we may arguably consider the
regulation (GDPR)) without jeopardising the end-to-end insatiable nature of technology users and service consumers,
availability, performance (e.g. relaying in D2D set-up), often desirous additional capabilities (such as high speed
resilience, reliability; and other 5G essential requirements. data rate systems and extremely low or no latency systems)
C. Flexibility has generally resulted in an astronomical rise in the demand
and consumption of wireless communication radio resources
Typically, the overall network topology in any
in the modern era. As a consequence, technology developers
telecommunication system is often subject to change due to
and service providers have had to innovate a new generation
the heterogeneity of network nodes. If we take into account
of technology atop 3G and 4G.
the addition of a distributed system of server clusters (MEC)
5G technology seeks to achieve a 1ms round-trip latency
working in unison with SDN and NFV to facilitate , then the
for major use cases and critical applications. In addition, it is
design of a practicable and cost-effective 5G architecture
anticipated that ultra-connectivity, ultra-high reliability and
will need to be able to successfully operate in a manner that
extremely high availability etc, reflected in everything
is ‘agnostic’ to the heterogeneity of the network irrespective
and/or everyone interconnected and able to seamlessly share
of the mode of operation (NSA or SA), otherwise the
information harmoniously, efficiently, reliably and securely
envisioned capabilities through the use of MEC, NFV, SDN
irrespective of time or location, whilst also prohibiting
etc may not be realised in a performant and flexible manner.
overall system uncertainty will be achieved. On the
In [76], architecture modularisation and slicing for 5G
presumption that all of these (and many more as discussed in
networks was proposed so that the projected set of use cases
the paper) will be feasibly achieved, would these suffice in
could be achieved with adequate flexibility across a HetNet
fulfilling the demands of the contemporary users or would
landscape. This was centred on the use of six fundamental
there be another generation of technology within the next
building blocks – access function, connectivity management,
decade and subsequently? Perhaps a more appropriate
mobility management, flow management, security and AAA
question ought to be: are we passably prepared on all
management, and context generation and handling function
accounts for the looming 5G technology, including socio-
(CGHF). While the modularisation concept appears
economic, techno-economic, regulatory, and sustainability
promising, significant R&D work would need to be
facets?
undertaken in the design (e.g. device selection and
It is apparent that 5G technology will radically change the
association), instantiation and operation of network slices
cellular, wireless and telecommunications landscape in the
(e.g. concurrent connectivity to multiple slices, context
near future and possibly beyond. The advent of 5G will pave
information sharing etc) in line with the core building
way for the development and deployment of new services
blocks.
and cost-effective telecommunications infrastructure.
D. Energy efficiency However, to the extent that the distinct characteristics and
As alluded to earlier in the paper, one of the key features, enabling technologies of 5G encounter trade-offs during

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

deployment and implementation (e.g. between performance QoE as the “overall acceptability of an application or
and security which is always a significant concern), the long service, as perceived subjectively by the end user”
[47]. In [48], QoE intelligence is considered to be
term viability and sustainability of 5G technology will highly valuable to telecommunication network
remain questionable. operators (including MNOs).
Be it gradual or rapid in propagation, it will surely be a Quality of Service Refers to a collection of characteristics regarding the
(QoS) performance of any connection or network, which
matter of time to discover whether 5G technology will contributes to the level of satisfaction, observed and
ultimately bring about the supposed networked society. Will derived by an end user vis-à-vis the network’s
5G be ultimately able to provide the multifarious technical performance. More technically, QoS means that
which a flow strives to achieve. Peculiar features of a
capabilities and benefits as envisioned or will it create a
flow include latency, jitter, reliability, bandwidth,
situation where the associated risks and technical challenges and data rate. Notably, QoS delivery to a network
markedly outweigh the perceived benefits and capabilities? may be nonguaranteed (best-effort delivery) or
In any case, as we look forward to the impending era of 5G guaranteed [16].
Scheduling Refers to the dynamic allocation of resources to UEs
technology and beyond, R&D of novel techniques and in both DL and UL [41].
enabling technologies in the area of mobile, cellular and Synchronised Coined by researchers at King’s College London UK
wireless communications will certainly not cease, at least, in Reality (SR) (http://www.pressreleasepoint.com/worlds-first-5g-
end-end-network-debuted), SR as the term implies,
the short term.
allows the synchronisation of the real, virtual and
mental worlds far beyond VR and AR in isolation. In
APPENDIX other words, SR creates a technology where VR, AR
(the physical surroundings) and mental stimulations
Definitions of some key terms and some common are synchronised i.e. seamlessly combined
acronyms used (including acronyms that are not defined Throughput Measured in bits per second (bps), it describes the
elsewhere in the paper) are presented in tables V and VI rate at which data (text, voice, video, multimedia etc)
is transported through a network [16].
respectively.
Uplink Opposite to DL and also called reverse link, it refers
TABLE V to the signal transmission in a direction from the MS
DEFINITION OF TERMS to the BS [3], [100]–[101].
Virtual Reality A technology whereby physical (i.e. real) presence is
Term Definition (VR) simulated by computer graphics, allowing the user to
interactively interact with the simulated elements (i.e.
Augmented A technology whereby computer-aided (i.e. digital) virtually). VR creates a wholly artificial environment.
Reality (AR) information based on user context is integrated with Applications of VR can be found in immersive sports
the user’s environment in real-time, such that the broadcasting, medicine, architecture, entertainment,
information is graphically enriched or augmented to arts etc [31], [52]. Both AR and VR help users to
the display. AR finds its applications in healthcare, interact as if they are in the same location. From an
retail, public safety, oil and gas, etc [31], [51]. experience perspective, while VR wholly replaces a
Availability Refers to a system’s endurance against potential user’s audio and visual sensations, AR enriches this
outage scenarios [29]. by providing further information that is germane to
Centimetre wave A type of wave produced by/from a 5G small cell the surroundings [32].
(CmWave) deployed in a 6GHz – 30GHz band using about
500MHz BW [34]. TABLE VI
Cognitive radio “Cognitive radio is an intelligent wireless ACRONYMS
communication system that is aware of its
surrounding environment (i.e., outside world), and Acronym Full form
uses the methodology of understanding-by-building
to learn from the environment and adapt its internal
states to statistical variations in the incoming RF 1G First Generation
stimuli by making corresponding changes in certain 2G Second Generation
operating parameters (e.g., transmit-power, carrier- 3G Third Generation
frequency, and modulation strategy) in real-time, 3GPP Third Generation Partnership Project
with two primary objectives in mind: highly reliable 4G Fourth Generation
communications whenever and wherever needed; 5G Fifth Generation
efficient utilization of the radio spectrum” [134] AUC Authentication Centre
Connection Refers to the total number of devices fulfilling a BSC Base Station Controller
density specific QoS per square kilometre [59]. BSS Base Station Subsystem
Converged Refers to the seamless coexistence of telephony and BTS Base Transceiver Station
network multimedia within a single network. eCoMP Evolved CoMP
Downlink Also called forward link, it refers to the signal ECSD Enhanced Circuit-Switched Data
transmission in a direction from the BS to the MS EGPRS Enhanced GPRS
[3], [100]–[101]. Enhanced International Mobile
eIMT-A
Fronthaul Refers to the connection between a new network Telecommunications Advanced
architecture of centralised baseband controllers and EIR Equipment Identity Register
remote standalone radio heads at cell sites [79]. It is eLTE Evolved LTE
kind of analogous to the backhaul, which connects eMIMO Evolved MIMO
the mobile network to a wired network. ETSI European Telecommunications Standards Institute
EV-DO Evolution Data Only/Optimised
Haptic In tactile internet, haptic information comprises two
GGSN Gateway GPRS Support Node
information separate types of feedbacks – kinesthetic feedback
GMSC Gateway MSC
(e.g. force, torque, position, velocity) and tactile
GMSK Gaussian Minimum Shift Keying
feedback (e.g. surface texture, friction) [62].
HLR Home Location Register
Quality of Influenced by several factors such as content, service,
HSDPA High Speed Downlink Packet Access
Experience (QoE) network, application, device and usage context, QoE
HSUPA High Speed Uplink Packet Access
“refers to the degree of delight or annoyance of the
IEC International Electrotechnical Commission
user of an application or service” [46]. ITU-T defines

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

IMT International Mobile Telecommunications [17] H. Holma and A. Toskala, “WCDMA for UMTS, Radio Access for
ISDN Integrated Services Digital Network Third Generation Mobile Communications,” Third Edition, John
ITU International Telecommunication Union Wiley & Sons Ltd, 2004.
ITU-R ITU Radio communication standardisation sector [18] T. Halonen, J. Romero and J. Melero, “GSM, GPRS and EDGE
ITU-T ITU Telecommunication standardisation sector Performance: Evolution Towards 3G/UMTS,” 2nd Edition, John
JTACS Japanese Total Access Communication System Wiley & Sons Ltd, 2003.
LTE-A LTE Advanced [19] M. Rumney (Editor-in-Chief), “LTE and the Evolution to 4G
LTE-LAA LTE-A Licence-Assisted Access Wireless – Design and Measurement Challenges,” Agilent
LTE-LWA LTE-A Wi-Fi Link Aggregation Technologies Inc., 2009.
LTE-M LTE for Machines (MTC, M2M, IoT) [20] F. Khan, “LTE for 4G Mobile Broadband - Air Interface
LTE-U LTE-A in Unlicensed Spectrum Technologies and Performance,” Cambridge University Press, 2009.
MSC Mobile (Services) Switching Centre [21] H. Holma and A. Toskala, “LTE for UMTS, OFDMA and SC-FDMA
MSS Mobile Station Subsystem Based Radio Access,” John Wiley & Sons, 2010.
MTSO Mobile Telephone Switching Office [22] O. T. Eluwole and M. Lohi, “Coordinated Multipoint Power
NMT Nordic Mobile Telephony Consumption Modeling for Energy Efficiency Assessment in
NSS Network Switching Subsystem LTE/LTE-advanced Cellular Networks,” IEEE 19th International
NX New Flexible air interface for 5G Conference on Telecommunications (ICT), Jounieh, Lebanon, April,
OMC Operations and Maintenance Centre 2012.
PLMN Public Land Mobile Network [23] Nokia Networks White Paper, “LTE Release 12 and Beyond,” 2014.
PSK Phase Shift Keying [Online]. Available: https://resources.ext.nokia.com/asset/200174.
PSTN Public Switched Telephone Network [24] 4G Americas, “Understanding 3GPP Release 12 Standards for
QPSK Quadrature Phase Shift Keying HSPA+ and LTE-Advanced Enhancements”, February 2015.
RNS Radio Network Subsystem [Online]. Available:
SGSN Serving GPRS Support Node http://www.5gamericas.org/files/6614/2359/0457/4G_Americas_-
SISO Single Input Single Output _3GPP_Release_12_Executive_Summary_-_February_2015.pdf
SU-MIMO Single-User MIMO [25] T. Nakamura, “Proposal for Candidate Radio Interface Technologies
TACS Total Access Communication System for IMT-Advanced Based on LTE Release 10 and Beyond (LTE-
TD-SCDMA Time Division Synchronous CDMA Advanced), 3GPP ITU-R WP 5D 3rd Workshop on IMT-Advanced,
Telecommunications Industries Association - 15 October 2009.
TIA-EIA
Electronic Industries Association [26] 3GPP TR 36.913, V8.0.0, Technical Specification Group RAN,
UTRAN UMTS Terrestrial RAN “Requirements for Further Advancements for Evolved Universal
VLR Visitor (Visited) Location Register Terrestrial Radio Access (E-UTRA) (LTE-Advanced) (Release 8),”
June 2008.
[27] 3GPP TR 36.913, V9.0.0, Technical Specification Group RAN,
REFERENCES “Requirements for Further Advancements for Evolved Universal
[1] K. Campbell, J. Diffley, B. Flanagan, B. Morelli, B. O’Neil and F. Terrestrial Radio Access (E-UTRA) (LTE-Advanced) (Release 9),”
Sideco, “The 5G economy: How 5G technology will contribute to the December 2009.
global economy,” The IHS 5G economic impact study, IHS [28] L. Gupta, R. Jain, and H. A. Chan, “Mobile Edge Computing – An
Economics and IHS Technology, January 2017. Important Ingredient of 5G Networks,” IEEE Software Defined
[2] Cisco, “Cisco Visual Networking Index: Global Mobile Data Traffic Networks Technical Community, Newsletter, March 2016.
Forecast Update, 2015–2020,” White Paper, February, 2016. [29] O. N. C. Yilmaz, “Ultra-Reliable and Low-Latency (URLLC) 5G
[3] D. T. C. Wong, P. Kong, Y. Liang, K.C. Chua, and J.W. Mark, Communication,” European Conference on Networks and
“Wireless Broadband Networks,” John Wiley & Sons, April 2009. Communications (EuCNC), June 2016.
[4] M. Shafi, et al. "5G: A Tutorial Overview of Standards, Trials, [30] Ericsson, “5G - key Component of the Networked Society (RWS-
Challenges, Deployment, and Practice," IEEE Journal on Selected 150009),” 3GPP RAN Workshop on 5G, Phoenix, AZ, USA, 17 – 18
Areas in Communications Vol. 35, No. 6, pp. 1201-1221, 2017. September 2015.
[5] V. Nguyen, A. Brunstrom, K. Grinnemo, and J. Taheri, "SDN/NFV- [31] Samsung Electronics White Paper, “5G Vision,” DMC R&D Center,
based Mobile Packet Core Network Architectures: A Survey," IEEE August 2015. [Online]. Available:
Communications Surveys & Tutorials, Vol. 19, No. 3, pp. 1567- http://www.samsung.com/global/business-
1602, Third Quarter, 2017. images/insights/2015/Samsung-5G-Vision-2.pdf
[6] A. Gupta and R. K. Jha, “A Survey of 5G Network: Architecture and [32] Ericsson White Paper, “5G systems – Enabling the Transformation of
Emerging Technologies,” IEEE Access Special Section on Recent Industry and Society,” UEN 284 23-3251 rev B, January 2017.
Advances in Software Defined Networking for 5G Networks, Volume [Online]. Available:
3, pp. 1206 – 1232, 2015. https://www.ericsson.com/assets/local/publications/white-papers/wp-
[7] J. H. Schiller, “Mobile Communications,” 2nd Edition, Pearson 5g-systems.pdf
Education Limited, 2003. [33] Ericsson White Paper, “5G Radio Access – Capabilities and
[8] W. C. Y. Lee, “Wireless and Cellular Telecommunications,” 3rd Technologies,” Uen 284 23-3204 Rev C, April 2016. [Online].
Edition, McGraw-Hill Companies, Inc., 2006. Available:https://www.ericsson.com/assets/local/publications/white-
[9] I. Stojmenovic, “Handbook of Wireless Networks and Mobile papers/wp-5g.pdf
Computing,” John Wiley & Sons Inc., 2002. [34] Nokia Networks, “Ten key rules of 5G deployment, Enabling 1
[10] J. Eberspächer, H-J Vögel, C. Bettstetter and C. Hartmann, “GSM – Tbit/s/km2 in 2030,” White Paper, May 2015. [Online]. Available:
Architecture, Protocols and Services,” 3rd Edition, John Wiley & https://networks.nokia.com/file/39891/ten-key-rules-of-5g-
Sons Ltd, 2009. deployment
[11] J. Bannister, P. Mather and S. Coope, “Convergence Technologies for [35] P. Sorrells, “5G: Advancing the world of connectivity,” CommScope
3G Networks - IP, UMTS, EGPRS and ATM,” John Wiley & Sons White Paper, September 2016. [Online]. Available:
Ltd, 2004. http://www.commscope.com/5g/wp-advancing-the-world-of-
[12] P. Chandra, “Bulletproof Wireless Security - GSM, UMTS, 802.11 connectivity/
and Ad Hoc Security,” Elsevier Inc., 2005. [36] M. Maier, M. Chowdhury, B. P. Rimal, and D. P. Van, “The Tactile
[13] M. Stamp, “Information Security – Principles and Practice,” John Internet: Vision, Recent Progress, and Open Challenges,” IEEE
Wiley & Sons, Inc., 2006. Communications, Volume 54, Issue 5, pp. 138 – 145, May 2016.
[14] T. S. Rappaport, “Wireless Communications - Principles and [37] N. Kundargi, “5G Standardization: 3GPP Status Update and
Practice,” 2nd Edition, Prentice Hall Inc., 2002. Overview,” National Instruments. [Online]. Available: ni.com/5g
[15] L. Goleniewski and K. W. Jarrett, “Telecommunications Essentials, [38] G. Brown, “Designing 5G-Ready Mobile Core Networks,” Heavy
Second Edition: The Complete Global Source,” 2nd Edition, Pearson Reading White Paper, September 2016. [Online]. Available:
Education Inc., October 2006. https://www.affirmednetworks.com/wp-
[16] B. A. Forouzan, “Data Communications and Networking,” Fourth content/themes/renden/pdf/5G_Whitepaper_Heavy_Reading.pdf
Edition, McGraw-Hill, 2007. [39] G. Brown, “Service-Oriented 5G Core Networks,” Heavy Reading
White Paper. [Online]. Available:

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

http://carrier.huawei.com/~/media/CNBG/Downloads/track/HeavyRe [62] A. Aijaz, M. Dohler, A. H. Aghvami, V. Friderikos, and M. Frodigh,

adingWhitepaperServiceOriented5GCoreNetworks.pdf “Realizing the Tactile Internet: Haptic Communications over Next
[40] J. Kima, D. Kimb, and S. Choi, “3GPP SA2 architecture and Generation 5G Cellular Networks,” IEEE Wireless Communications,
functions for 5G mobile communication system,” The Korean Volume 24, Issue 2, pp. 82 – 89, April 2017.
Institute of Communications Information Sciences, Publishing [63] P. Popovski, “Ultra-Reliable Communication in 5G Wireless
Services by Elsevier B.V, ICT Express, Volume 3, Issue 1, pp. 1–8, Systems,” IEEE 1st International Conference on 5G for Ubiquitous
March 2017. Connectivity (5GU), pp. 146 – 151, November2014.
[41] 3GPP TR 38.804 V14.0.0, “Technical Specification Group Radio [64] M. Dohler, “The Tactile Internet, IoT, 5G and Cloud on Steroids,”
Access Network; Study on New Radio Access Technology; Radio IET 5G Radio Technology Seminar, Exploring Technical Challenges
Interface Protocol Aspects (Release 14),” March 2017. in the Emerging 5G Ecosystem, London, UK, March 2015.
[42] 3GPP TR 38.801 V14.0.0, “Technical Specification Group Radio [65] ITU-T Technology Watch Report, “The Tactile Internet,” ITU,
Access Network; Study on new radio access technology: Radio access August 2014. [Online]. Available: https://www.itu.int/dms_pub/itu-
architecture and interfaces (Release 14),” March 2017. t/oth/23/01/T23010000230001PDFE.pdf
[43] 3GPP TS 23.501 V1.4.0, “Technical Specification Group Services [66] IEEE Talks 5G, “Three Questions for Gerhard Fettweis,” January
and System Aspects; System Architecture for the 5G System; Stage 2 2017. [Online]. Available: https://5g.ieee.org/publications/ieee-talks-
(Release 15),” September 2017. 5g/50-ieee-talks-5g-three-questions-for-gerhard-fettweis
[44] 3GPP TR 23.799 V14.0.0, “Technical Specification Group Services [67] M. Dohler, “5G Tactile Internet – Lab Experimentation @ King’s,”
and System Aspects; Study on Architecture for Next Generation UK 5G Testbed Meeting, London, UK, 5 July 2016. [Online].
System (Release 14),” December 2016. Available: https://www.commnet.ac.uk/wp-
[45] 3GPP TR 22.891 V14.2.0, “Technical Specification Group Services content/uploads/2016/08/05Jul16-KCL-Mischa-Dohler.pdf
and System Aspects; Feasibility Study on New Services and Markets [68] G. P. Fettweis, “The Tactile Internet: Applications and Challenges,”
Technology Enablers (SMARTER); Stage 1 (Release 14),” September IEEE Vehicular Technology Magazine, Volume 9, Issue 1, pp. 64 –
2016. 70, March 2014.
[46] Qualinet, “White Paper on Definitions of Quality of Experience,” [69] A. B. Grant and O. T. Eluwole, “Cloud Resource Management –
Output from the fifth Qualinet meeting, Version 1.2, Novi Sad, Virtual Machines Competing for Limited Resources,” IEEE 55th
March 12, 2013. International Symposium ELMAR, Zadar, Croatia, pp. 269–274,
[47] ITU-T Rec. P.10, “Vocabulary for performance and quality of September, 2013.
service, Amendment 2: New definitions for inclusion in [70] M. H. Ghahramani, M. C. Zhou, and C. T. Hon, “Toward Cloud
Recommendation ITU-T P.10/G.100,” Int. Telecomm. Union, Computing QoS Architecture: Analysis of Cloud Systems and Cloud
Geneva, 2008. Services,” IEEE/CAA Journal of Automatica Sinica, vol. 4, no. 1, pp.
[48] E. Liotou, N. Passas and L. Merakos, “The Emergence of Experience 6–18, January 2017.
Packages in the 5G Era,” IEEE 5G Tech Focus, Volume 1, Number 3, [71] P. Mell and T. Grance, “The NIST Definition of Cloud Computing,”
September 2017. National Institute of Standards and Technology, Information
[49] T. Shimojo, A. Umesh, D. Fujishima and A. Minokuchi, “Future Technology Laboratory, Version 15, Oct. 7, 2009.
Core Network for the 5G Era,” NTT DOCOMO Technical Journal, [72] A. Manzalini, C. Buyukkoc, P. Chemouil, S. Kuklinski, F. Callegati,
vol. 17, no. 4, pp. 50 – 59, 2016. [Online]. Available: A. Galis, M. Odini, C. Lin I, J. Huang, M. Bursell, N. Crespi, E.
https://www.nttdocomo.co.jp/english/binary/pdf/corporate/technology Healy, and S. Sharrock, “Towards 5G Software-Defined Ecosystems:
/rd/technical_journal/bn/vol17_4/vol17_4_006en.pdf Technical Challenges, Business Sustainability and Policy Issues,”
[50] R. Kirichek, “5G and Tactile Internet. Network and Services IEEE SDN Initiative White Paper, IEEE Future Directions, July 2016.
Decentralization,” ITU Workshop on Innovative ICT Technologies, [73] K. Pretz, “The “Softwarization” of Telecommunications Systems,”
Tashkent, Uzbekistan, 21-22 September 2016. The Institute, The IEEE news source, 26 July 2016.
[51] TechTarget, “augmented reality (AR)”. [Online]. Available: [74] A. Manzalini and Noel Crespi, “SDN and NFV for Network Cloud
http://whatis.techtarget.com/definition/augmented-reality-AR Computing: A Universal Operating System for SD Infrastructures,”
[52] Virtual Reality Society (VRS), “What is Virtual Reality”. [Online]. IEEE Fourth Symposium on Network Cloud Computing and
Available: https://www.vrs.org.uk/virtual-reality/what-is-virtual- Applications (NCCA), 11 – 12 June 2015.
reality.html [75] L. Gorrati, “The 5G Infrastructure Public private partnership (5G-
[53] M. Dohler and K. O’Brien, “The Internet of Things,” online course, PPP) Project SESAME: Bringing Virtualization, Control and
FutureLearn, 2016. Intelligence to the Network Edge”. [Online]. Available:
[54] D. D. Coleman and D. A. Westcott, “Certified Wireless Network http://www.charisma5g.eu/wp-content/uploads/2016/07/5G-PPP-
Administrator, CWNA Study Guide,” Wiley Inc., 2006. Project-SESAME.Bringing-Virtualization-Control-and-Intelligence-
[55] E. G. Larsson, F. Tufvesson, O. Edfors, and T. L. Marzetta, “Massive to-the-Network-Edge.pdf
MIMO for Next Generation Wireless systems,” IEEE [76] X. An, R. Trivisonno, H. Einsiedler, D. V. Hugo, K. Haensge, X.
Communications Magazine, Volume 52, Issue 2, pp. 186–195, Huang, Q. Shen, D. Corujo, K. Mahmood, D. Trossen, M. Liebsch, F.
February 2014. Leitao, C. Phan and F. Klamm, “End-to-End Architecture
[56] L. Lu, G. Y. Li, A. L. Swindlehurst, A. Ashikhmin, and R. Zhang, Modularisation and Slicing for Next Generation Networks”. [Online].
“An Overview of Massive MIMO: Benefits and Challenges,” IEEE Available: https://arxiv.org/ftp/arxiv/papers/1611/1611.00566.pdf
Journal of Selected Topics in Signal Processing, Volume 8, Issue 5, [77] N. Sprecher, “Mobile Edge Computing, a key technology towards
pp. 742 - 758, October 2014. 5G,” 5G World, London, UK, 2016. [Online]. Available:
[57] M. Sawahashi, Y. Kishiyama, A. Morimoto, D. Nishikawa, M. http://www.etsi.org/component/rsfiles/download-
Tanno, “Coordinated Multipoint Transmission/Reception Techniques file/files?path=MEC+Presentations+at+Industry+Events%5C201606
for LTE-Advanced [Coordinated and Distributed MIMO],” IEEE +-+5G+World+-+ETSI+Mobile+Edge+Computing.pdf
Wireless Communications, Volume 17, Issue 3, pp. 26–34, June [78] iGillottResearch (iGR) Inc., “Moving to Cloud RAN and the Potential
2010. of Edge Data Centers,” White Paper, Version 1.0, October 2016.
[58] Anritsu, “5G Standardization Status in 3GPP,” No. 91, March 2016. [Online]. Available:
[Online]. Available: http://dl.cdn-anritsu.com/ja-jp/test- www.commscope.com/Docs/iGR_cRAN_White_Paper.pdf
measurement/reffiles/About-Anritsu/R_D/Technical/91/91-02-5g- [79] D. Jones (Mobile Editor), “What is [Bleep] Fronthaul?”. [Online].
2.pdf Available: http://www.lightreading.com/what-the--bleep--is-
[59] ITU-R SG05 Contribution 40, “Draft new Report ITU-R M.[IMT- fronthaul/a/d-id/707868
2020.TECH PERF REQ] - Minimum requirements related to [80] T. S. Rappaport, S. Sun, R. Mayzus, H. Zhao, Y. Azar, K. Wang, G.
technical performance for IMT-2020 radio interface(s),” February N. Wong, J. K. Schulz, M. Samimi, and F. Gutierrez, “Millimeter
2017. [Online]. Available: https://www.itu.int/md/R15-SG05-C- Wave Mobile Communications for 5G Cellular: It Will Work!,” IEEE
0040/en Access, Volume 1, pp. 335 – 349, May 2013.
[60] BBC Technology News, “5G researchers manage record connection [81] A. Bader and M. Alouini, “5G "On The Fly",” IEEE 5G Tech Focus,
speed,” 25 February 2015. [Online]. Available: Volume 1, Number 3, September 2017.
http://www.bbc.com/news/technology-31622297 [82] IGI Global, “What is Smart Technology,” Online dictionary. [Online].
[61] I. Grigorik, “High Performance Browser Networking,” O'Reilly Available: https://www.igi-global.com/dictionary/smart-
Media, Inc., 2013. technology/38186

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

[83] W. Derzko, “Smart Technologies,” World Future Society, 29 July Processing (ISCCSP), 2014 6th International Symposium on. IEEE,
2006. [Online]. Available: 2014.
http://archives.ocediscovery.com/2007/presentations/Session3Walter [107] N. Bhushan, et al. "Network densification: the dominant theme for
DrezkoFINAL.pdf wireless evolution into 5G." IEEE Communications Magazine 52.2
[84] TechTarget, “machine-to-machine (M2M)”. [Online]. Available: (2014): 82-89.
http://internetofthingsagenda.techtarget.com/definition/machine-to- [108] J. Liu, et al. "Network densification in 5G: From the short-range
machine-M2M communications perspective." IEEE Communications
[85] C. Wang, “M2M Service Architecture: Delivering M2M Services Magazine 55.12 (2017): 96-102.
Over Heterogeneous Networks,” IEEE Communications Quality & [109] L. Yang, and Z. Wei. Interference Coordination for 5G Cellular
Reliability 2012 International Workshop, May 17, 2012. Networks. Springer, 2015.
[86] Z. Abbas and W. Yoon, “A Survey on Energy Conserving [110] W. Nam, et al. "Advanced interference management for 5G cellular
Mechanisms for the Internet of Things: Wireless Networking networks." IEEE Communications Magazine52.5 (2014): 52-60.
Aspects,” Sensors Journal, no. 15, pp. 24818 – 24847, September [111] M. O. Ojo and O. O. Aramide. "Various interference models for
2015. multicellular scenarios: A comparative study." Digital Information
[87] M. Weiser, “The Computer for the 21st Century,” Scientific and Communication Technology and its Applications (DICTAP),
American Journal, September 1991. 2015 Fifth International Conference on. IEEE, 2015.
[88] Postscapes, “Internet of Things (IoT) History”. [Online]. [112] Z. Ding, et al. "On the performance of non-orthogonal multiple
Available:https://www.postscapes.com/internet-of-things-history/ access in 5G systems with randomly deployed users." IEEE Signal
[89] K. D. Foote, “A Brief History of the Internet of Things,” Processing Letters 21.12 (2014): 1501-1505.
DATAVERSITY Education LLC, 16 August 2016. [Online]. [113] R. Baldemair, et al. "Ultra-dense networks in millimeter-wave
Available: http://www.dataversity.net/brief-history-internet-things/ frequencies." IEEE Communications Magazine 53.1 (2015): 202-
[90] K. Ashton, “That 'Internet of Things' Thing,” RFID Journal. [Online]. 208.
Available: http://www.rfidjournal.com/articles/view?4986 [114] Y. Mehmood, et al. "M2M communications in 5G: state-of-the-art
[91] TechTarget, “IoT analytics guide: Understanding Internet of Things architecture, recent advances, and research challenges." IEEE
data”. [Online]. Available: Communications Magazine 55.9 (2017): 194-201.
http://searchbusinessanalytics.techtarget.com/essentialguide/IoT- [115] E. Yaacoub, and D. Zaher. "On using relays with carrier aggregation
analytics-guide-Understanding-Internet-of-Things-data for planning 5G networks supporting M2M traffic." Wireless and
[92] SRI Consulting Business Intelligence (SRIC BI), Conference Report Mobile Computing, Networking and Communications (WiMob),
CR 2008-07, “Disruptive civil technologies: Six technologies with 2014 IEEE 10th International Conference on. IEEE, 2014.
potential impacts on US interests out to 2025,” April 2008. [Online]. [116] R. Ratasuk, et al. "Recent advancements in M2M communications in
Available: https://fas.org/irp/nic/disruptive.pdf 4G networks and evolution towards 5G." Intelligence in Next
[93] O. T. Eluwole, N. S. Udoh and O. D. Ojo, “The Impact of Internet on Generation Networks (ICIN), 2015 18th International Conference on.
African Education and Culture,” International Journal of Business, IEEE, 2015.
Humanities and Technology (IJBHT), vol. 4, no. 3, pp. 69-77, May [117] S. Guirguis, et al. "M2M communications: Enablement in 4G LTE,
2014. deployment considerations and evolution path to 5G." Wireless
[94] K. Mengersen (Lead Educator), “Big Data: Statistical Inference and Networks and Mobile Communications (WINCOM), 2017
Machine Learning”, online course, FutureLearn, 2016. International Conference on. IEEE, 2017.
[95] L. Xu, C. Jiang, J. Wang, J. Yuan, and Y. Ren, “Information Security [118] Y. Zhang, et al. "Cognitive machine-to-machine communications:
in Big Data: Privacy and Data Mining”, IEEE Access, Volume 2, pp. visions and potentials for the smart grid." IEEE network 26.3 (2012).
1149 – 1176, October 2014. [119] A. Asadi, Q. Wang and V. Mancuso. "A survey on device-to-device
[96] Tableau Software, “Top Ten Big Data Trends for 2017”, White communication in cellular networks." IEEE Communications Surveys
Paper, 2017. [Online]. Available: & Tutorials 16.4 (2014): 1801-1819.
https://www.tableau.com/learn/whitepapers [120] R. I. Ansari, et al. "5G D2D Networks: Techniques, Challenges, and
[97] V. Krishnan, “Solving the Internet of Things’ Last-Mile Problem”, Future Prospects." IEEE Systems Journal (2017).
Tableau Software White Paper, 2017. [Online]. Available: [121] M. N. Tehrani, M. Uysal, and H. Yanikomeroglu. "Device-to-device
https://www.tableau.com/learn/whitepapers communication in 5G cellular networks: challenges, solutions, and
[98] The IEEE 5G Initiative, “IEEE 5G and Beyond Technology Roadmap future directions." IEEE Communications Magazine 52.5 (2014): 86-
White Paper”, 2017. [Online]. Available: 92.
https://5g.ieee.org/images/files/pdf/ieee-5g-roadmap-white-paper.pdf [122] X. Cheng, et al. "Device-to-device channel measurements and
[99] O. T. Eluwole, M. Lohi and N. S. Udoh, “Smart Card Technology: models: a survey." IET communications 9.3 (2015): 312-325.
Contactless Payment System and Near Field Communication,” SDPS [123] X. Lin, et al. "An overview of 3GPP device-to-device proximity
19th International Conference on Transformative Research in Science services." IEEE Communications Magazine 52.4 (2014): 40-48.
and Engineering, Business and Social Innovation, Sarawak, Malaysia, [124] U. Abbasi and H. Elbiaze. "Multimedia streaming using D2D in 5G
pp. 19-26, June 2014. ultra dense networks." Consumer Communications & Networking
[100] N. Patwari, “Wireless Communication Systems,” Department of Conference (CCNC), 2018 15th IEEE Annual. IEEE, 2018.
Electrical and Computer Engineering, University of Utah, Fall 2011. [125] A. V. Bastos, C. M. Silva and D. C. da Silva Jr. "Assisted Routing
[Online]. Available: Algorithm for D2D Communication in 5G Wireless Networks."
http://www.ece.utah.edu/~npatwari/pubs/lectureAll_ece5325_6325_f [126] O. N. C. Yilmaz et al., “Smart mobility management for D2D
11.pdf communications in 5G networks,” in Proc. IEEE Wireless Commun.
[101] A. Manikas, “Communication theory,” Department of Electrical and Netw. Conf. Workshops (WCNCW), 2014, pp. 219–223.
Electronic Engineering, Imperial College of Science, Technology and [127] Z. Gao, et al. "MmWave massive-MIMO-based wireless backhaul for
Medicine, November 2001. [Online]. Available: the 5G ultra-dense network." IEEE Wireless Communications 22.5
http://skynet.ee.ic.ac.uk/notes2007/CS%20Part%203(1)%202001%2 (2015): 13-21.
0-%20Mobile%20Communication%20Systems%20(GSM).pdf [128] H. Q. Ngo, E. Larsson, and T. Marzetta, “Energy and spectral
[102] M. Ojo, D. Adami and S. Giordano, "A SDN-IoT Architecture with efficiency of very large multiuser MIMO systems,” IEEE Trans.
NFV Implementation," 2016 IEEE Globecom Workshops (GC Commun., vol. 61, no. 4, pp. 1436–1449, Apr. 2013
Wkshps), Washington, DC, USA, 2016. [129] E. Bjornson, E. G. Larsson, and M. Debbah, “Optimizing multicell
[103] M. Bellanger, et al. "FBMC physical layer: a primer." PHYDYAS, massive MIMO for spectral efficiency: How many users should be
January 25.4 (2010): 7-10. scheduled?” in Proc. IEEE Global Conf. Signal Inf. Process.
[104] L. Dai, et al. "Non-orthogonal multiple access for 5G: solutions, (GlobalSIP), 2014, pp. 612–616.
challenges, opportunities, and future research trends." IEEE [130] T. L. Marzetta, “Noncooperative cellular wireless with unlimited
Communications Magazine 53.9 (2015): 74-81. numbers of base station antennas,” IEEE Trans. Wireless Commun.,
[105] C. Sun, et al. "Beam division multiple access transmission for vol. 9, no. 11, pp. 3590–3600, Nov. 2010.
massive MIMO communications." IEEE Transactions on [131] F. Rusek et al., “Scaling up MIMO: Opportunities and challenges
Communications 63.6 (2015): 2170-2184. with very large arrays,” IEEE Signal Process. Mag., vol. 30, no. 1,
[106] F. Schaich and W. Thorsten "Waveform contenders for 5G—OFDM pp. 40–60, Jan. 2013
vs. FBMC vs. UFMC." Communications, Control and Signal

(Advance online publication: 28 August 2018)

 IAENG International Journal of Computer Science, 45:3, IJCS_45_3_06
______________________________________________________________________________________

[132] J. Hoydis et al., “Massive MIMO in the UL/DL of cellular networks:

How many antennas do we need?” IEEE J. Sel. Areas Commun., vol.
31, no. 2, pp. 160–171, Feb. 2013.
[133] K. E. Skouby and P. Lynggaard. "Smart home and smart city
solutions enabled by 5G, IoT, AAI and CoT services." Contemporary
Computing and Informatics (IC3I), 2014 International Conference
on. IEEE, 2014.
[134] S. Haykin "Cognitive radio: brain-empowered wireless
communications." IEEE Journal on Selected Areas in
Communications 23.2 (2005): 201-220.

(Advance online publication: 28 August 2018)