ITU Centres of Excellence for Europe

Future Mobile and Wireless Broadband: LTE-A-Pro,

WiFi, Satellites, 5G NR and AI

Module 1:
Mobile broadband: LTE-Advanced/LTE-A-Pro, mobile
Internet and cellular IoT
Table of contents
1.1 Mobile broadband evolution ........................................................................2
1.1.1 Mobile generations: 1G to 5G .............................................................4
1.2 LTE/LTE-Advanced architecture .................................................................7
1.2.1 E-UTRAN ............................................................................................7
1.2.2 LTE radio interface..............................................................................9
1.2.3 LTE-Advanced ..................................................................................11
1.3 4G Evolved Packet Core (EPC) ................................................................15
1.3.1 EPC network nodes ..........................................................................15
1.3.2 Control Plane and User Plane protocols in 4G..................................17
1.4 4.9G: LTE-Advanced-Pro..........................................................................20
1.4.1 LTE-Advanced Pro characteristics ....................................................20
1.4.2 LTE-Advanced-Pro Carrier Aggregation enhancements ...................23
1.5 4.9G QoS and mobile Internet access ......................................................25
1.5.1. QoS for LTE/LTE-Advanced .............................................................25
1.5.2. QoS for LTE-Advanced Pro ..............................................................29
1.5.3 Testing Mobile Internet access .........................................................31
1.5.4 QoS operational strategy for supervision on mobile broadband........33
1.6 Cellular Internet of Things (IoT): LTE-M and NB-IoT.................................36
1.6.1 Cellular IoT standards in LTE-Advanced-Pro....................................38
1.6.2 LTE-M ...............................................................................................39
1.6.3 Narrow-Band IoT (NB-IoT) ................................................................40
1.7 LTE in unlicensed bands...........................................................................44
1.7.1 LWA (LTE-WLAN Aggregation) ........................................................45
1.7.2 LTE Assisted Access (LAA) ..............................................................47
1.8 LTE-A-Pro V2X and mission critical services ............................................50
1.8.1 Vehicular to Everything (V2X) services .............................................50
1.8.2 Mission Critical services....................................................................55
1.9 Business and regulatory aspects of mobile broadband.............................59
1.9.1 Global development of mobile broadband and digital divide .............59
1.9.2 From network expansion to network densification ............................61
1.9.3 Mobile broadband impact on economy through the pandemic..........62
1.9.4 Six golden regulatory rules for take-up of mobile broadband ............63
Abbreviations ......................................................................................................65
References .........................................................................................................67

1
1.1 Mobile broadband evolution

The mobile broadband is now a reality in most of the countries

worldwide. At the end of 2021 we are talking largely about 5G and its
implementations or foreseen deployments, while 4G is still taking the largest
share in the mobile world, including the volume of mobile traffic as well as
number of connections over mobile networks. Considering that 4G will remain
dominant mobile technology worldwide until mid of this decade (2020s), as well
as 4G mobile core is used in current 5G mobile deployments, in this first Module
of this course we cover the development of 4G mobile technology in parallel with
ongoing 5G standardization by 3GPP releases (3GPP stands for 3G Partnership
Project).
Nowadays, mobile networks are largely used for access to open Internet,
which is usually referred to ‘mobile data’. But, how it started?
In fact, the Internet and mobile networks have started separately, each
with different targets at the beginning. Internet has started as a research project
in USA in 1960s and has become globally opened network in 1995, when it
stopped being a project and started to be what it is today. From the beginning of
the Internet, the Internet network grows exponentially in number of IP networks
as well as number of individual users or hosts connected to Internet, [1], [2].
Separation of networking technologies (i.e. protocols such as IP – Internet
Protocol) from applications running on the top has made the Internet flexible and
over the past couple of decades Internet technologies have become the main
approach for networking in the telecom/ICT world. That process of convergence
of telecom world over Internet technologies was largely supported by the ITU in
the past two decades (2000s and 2010s) via standardization of the NGN (Next
Generation Networks) umbrella standards.
What was intention of ITU’s NGN standardization and what was impact on
the present-day telecom world?
Well, the main target of NGN was transfer of legacy telecom services from
20th century (they are telephony and television) onto IP networks, by defining
frameworks and standards (to be globally interoperable) based on synergy with
other Standard Developing Organizations (SDOs) such as IETF (Internet
Engineering Task Force) for all Internet technologies, and 3GPP (3G Partnership
Project) for all mobile technologies being in use today (e.g., 3G, 4G, 5G). For
example, the main signaling protocol for voice (i.e., telephony became SIP
(Session Initiation Protocol) is an IETF standard established in 2002, which
replaced the ITU’s signaling system #7 (SS7), largely being in use in legacy
telecommunication networks at the end of 20th century. Then, SIP (together with
Diameter, as another control protocol from IETF “line”) has become intrinsic part
of IMS (IP Multimedia Subsystem), as a 3GPP standard for both fixed and mobile

2
networks. Also, NGN has provided open interfaces for introduction of different
services (including those that are not invented yet), and also fostered the IoT
(Internet of Things) developments towards global standards, since mid of 2000s
(one may note that IoT is also part of the NGN framework). Another important
note is that NGN (and further, Future Networks recommendations by ITU)
provided the ground for further split of network nodes to network functions, so in
latter mobile standards we largely talk about network functions (which is the main
approach in 5G standards today).
On the other side, mobile technologies started in 1980s with 1G (now it is
called as such), which was fully analogues and without features such as roaming
and other service than mobile telephony. The first breakthrough in the mobile
world was with 2G (second generation) which appeared in early 1990s with GSM
in Europe (standardized by ETSI, now part of the 3GPP mobile standards), IS-95
in Americas, and other standards elsewhere. The GSM (from 3GPP) has become
the dominant digital mobile technology due to several important novelties such as
roaming (using the same end device in different networks or countries, which
provide the same mobile networks, GSM in this case) and data services (such as
SMS and dial-up access to data networks, such as Internet).
So, the mobile technologies started to spread globally in 1990s, in parallel
with the public Internet, however initially not related one to another. Due to its
obvious dominance on the mobile market (e.g., GSM) and data services (e.g.,
Internet network) they have started to converge – that is, mobile network included
IP connectivity (starting with 2.5G, such as GPRS in GSM networks). On the
other side, Internet connectivity has contributed to appearance and growth of
mobile broadband access in 2000s and 2010s. Why is that so?
Well, the reason is that mobile broadband is required for mobile Internet
access on the first place, which is still completely valid statement now in 2021.
However, mobile networks have also changed the look of Internet by
making many services and applications to target mobile users first (which is
completely different than the past – couple of decades ago, when the Internet
was initially created for fixed hosts only).

Figure 1.1 Population coverage by type of mobile networks, 2015-2020, [4].

3
 Figure 1.1 shows that globally almost 85 per cent of the population was
covered by a 4G networks at the end of 2020. In the period between 2015 and
2020, 4G network coverage increased two-fold globally. However, as expected,
annual growth has been slowing down gradually since 2017, and 2020 coverage
is only 1.3 percentage points higher than 2019. So, one may say that 93% of the
world population has access to a mobile-broadband network, less than half a
percentage point higher than a year before.
One may note also note that Figure 1.1 does not show 5G mobile
networks since their contribution to the global mobile traffic was negligible in the
past couple of years (2019 and 2020), although many 5G deployments are
already on the ground today and many end-users (equipped with newer 5G
capable devices) can connect to new generation mobile systems (5G) in places
where they currently exist. However, the predictions show that 4G is still the
strongest mobile generation today (November 2021) and it will continue to be on
the top spot in the mobile world at least in the following several years (until 5G
takes the ‘main’ position).

1.1.1 Mobile generations: 1G to 5G

Mobile technologies evolution goes through with so-called generations

(Table 1.1), so let see a summary of what they include:
 1G (First Generation): It consisted of analogue mobile systems, based
mainly on FDMA (Frequency Division Multiple Access) in the radio access
part, without global roaming, it was used in 1980s.
 2G (Second Generation): This was the first generation of digital mobile
systems, based mainly on TDMA (Time Division Multiple Access) and
FDMA (e.g., GSM is the most known example), Circuit-Switched (CS),
with global roaming, and telephony and SMS as main services - started at
the beginning of 1990s.
 3G (Third Generation): First generation of mobile systems which included
by default Packet Switched (PS) domain (for Internet access, and MMS) in
parallel with CS (for voice and SMS), based on WCDMA – Wideband
Code Division Multiple Access (with TDMA/FDMA) in radio part - started at
the beginning of 2000s.
 4G (Fourth Generation): This is a generation which is all-IP by default in
access and core parts, where access part is based on OFDMA
(Orthogonal Frequency Division Multiple Access) with LTE (Long Term
Evolution) standards from 3GPP being the dominant ones. It started
around 2010.
 5G (Fifth Generation): The last generation of mobile systems, which
increases the data rates of 4G by more than 10 times with new radio
interface called NR (New Radio) and allocation of new spectrum, providing
possibilities for many new emerging services in different verticals. The 5G
decade is 2020s, the ongoing decade.

4
 Table 1.1 Speed and latency evolution of mobile networks, from 2G to 5G, [5].

2G 3G 4G 5G

Approximate
1990 - 2000 2000 - 2010 2010 - 2020 2020 - 2030
deployment date

Average 10s to 100s of Mbit/s to few

10s of Mbit/s 100s of Mbit/s
individual speed Kbit/s 10s of Mbit/s

few 100s of
Expected latency 100s of ms < 100 ms < 10 ms
ms

If we focus on the evolution of mobile generations, it was driven by the

creation and development of new services for mobile devices, and was enabled
by progress of the technology available for mobile systems. There was also an
evolution of the environment in which mobile systems are deployed and
operated, in terms of competition between mobile operators, challenges from
other mobile technologies, and new regulation of spectrum use and market
aspects of mobile systems. Moreover, the continuous "chase" for higher data
rates, higher capacity, higher throughput, lower delay, better spectrum
efficiency and flexibility, high level of QoS provisioning, diversified mobile
speed and greater coverage over cellular (so, many things), resulted in
GPRS/EDGE (2.5/2.75G) evolving to UMTS (3G) to HSPA (3.5G) to HSPA+
(3.75G), then to 4G LTE, LTE-Advanced and LTE-Advanced Pro, finally evolving
to 5G NR, and continuing further.

Figure 1.2 The 3GPP mobile networks evolution, [1].

Figure 1.2 shows the standardization timeline of mobile technologies by

3GPP, including 3G, 4G and first 5G. It is clear that 3GPP has won the mobile

5
technologies battle going towards the 5G, similar to the Internet having won the
packet-switching technologies battle several decades ago.
Their convergence (mobile and Internet) has started with GPRS in 2G era,
continued with UMTS/HSPA (HSPA stands for High Speed Packet Access) in 3G
era, and then to LTE/LTE-Advanced/LTE-Advanced-Pro in 4G era. The newest
generation of mobile networks (5G) is also primarily IP-based, initially targeted to
provide higher capacity, higher bitrates (i.e., speeds), higher quality and
efficiency for mobile Internet access.
And, the access to open Internet via mobile networks with high bitrates,
accompanied with high-level QoS, is referred to as mobile broadband. Of
course, one should note that mobile networks are also used for other
(specialized) telecom services which are not provided over he public Internet.
However, the notion of broadband (and with that, mobile broadband) is
changing over time, with the development of technology, so what is broadband
today is different from what broadband was a decade ago or what will be
broadband in some distant future.

6
1.2 LTE/LTE-Advanced architecture

The 4G architecture from 3GPP is all-IP meaning that all network nodes
are routers/gateways and switches and all links (wired and radio) carry IP
packets. Previous 3GPP standards, such as 2G GSM/GPRS or 3G UMTS/HSPA
were not all-IP, because they had two parts: Packet-Switching (PS) part and
Circuit-Switching (CS) part, where the CS part was used mainly for voice. So, 4G
architecture is fully packet-based (i.e., IP-based), but at the same time it is an
evolution from packet-switching parts in 3G and 2G mobile systems from 3GPP,
therefore it is labeled as Evolved Packet System (EPS). The EPS network
architecture consists of two main parts:
 Evolved Universal Terrestrial Radio Access Network (E-UTRAN), and
 Evolved Packet Core (EPC).

1.2.1 E-UTRAN

E-UTRAN (Evolved Universal Terrestrial Radio Access) consists of 4G

base stations, which are called eNodeB (or shortly eNB), and interfaces
between them and with the core. In E-UTRAN for the first time (compared to
previous mobile generations in respect to 4G), the 3GPP introduces direct
interface between base stations called X2 (Figure 1.3).

Figure 1.3 4G Evolved UTRAN (E-UTRAN).

7
 There are no other network entities in the LTE RAN (LTE Radio Access
Network) called E-UTRAN (“translated” – this means that it is based on evolution
of UTRAN which is the 3G RAN), so eNodeBs are directly connected to LTE
core network nodes (i.e., to EPC – Evolved Packet Core). Direct links between
eNodeBs are crucial for having fast handovers, which was needed for to all-IP
environment as LTE is (including its access and core network parts).
In previous 3G mobile networks from 3GPP there was centralized node in
the UTRAN, called RNC (Radio Network Controller). However, with the evolution
of the 3GPP mobile network architecture, the RNC was removed and the radio
resource management was completely given to the base stations (eNodeB in
4G), as shown in Figure 1.4.

Figure 1.4 3GPP architecture evolution, [1].

All 3GPP mobile networks consist of two main overlay network

architectures, one for user traffic (e.g., voice, web, etc.) and other for control
traffic (e.g., authentication, signaling for voice, etc.). With the convergence of
3GPP networks towards all-IP both user and control traffic have become IP
traffic. Therefore were needed changes in the mobile network architecture with
aim to provide lower delay for delay-sensitive traffic such as voice over IP in 4G
(e.g., VoLTE) and its accompanying signaling when all services (including voice

8
and all others) are transferred over IP end-to-end. However, IP networks have
larger delay budget than legacy telephone networks due to additional delays
from buffering and processing of IP packets in different networks nodes on the
path from source to destination host. That in fact was the main reason for change
of hierarchical network architecture (e.g., with Radio Network Controller – RNC
until 3GPP Release 7) to a flat architecture (which is now used for 4G networks,
and same approach of flat architecture is also used in 5G mobile networks).
Why flat architecture for mobile networks?
Well, because lower number of nodes on the path of IP packets are
required to decrease the delay budget (i.e., total delay due to various causes) to
fit into the end-to-end delay budget for voice when it is carried in a form of VoIP.
First move toward flat architecture was in 3GPP release 7 with the introduction of
so-called direct tunnel, and finally the flat architecture was standardized
completely in Release 8, as shown in Figure 1.4. This evolution is known as
System Architecture Evolution (SAE). Why is called “flat”? Because it is consisted
of only two tiers:
 base stations (e.g., eNodeBs in LTE/LTE-Advanced networks) and
 centralized gateways (nodes in Evolved Packet Core – EPC).

1.2.2 LTE radio interface

Regarding the radio access technology, both LTE and LTE-Advanced use
OFDMA (Orthogonal Frequency Division Multiple Access) in the radio
interface in downlink, and Single Carrier FDMA (SC-FDMA) in the uplink (Figure
1.5).

Figure 1.5 LTE radio access technologies, [12].

The generic LTE radio frame for FDD (Frequency Division Duplex) and
TDD (Time Division Duplex) has duration of 10ms and consists of 20 slots with
slot duration of 0.5 ms.
 Two adjacent slots form one sub-frame of length 1ms.
 So-called Resource Block (RB) spans either 12 sub-carriers with a sub-
carrier bandwidth of 15 kHz, or 24 sub-carriers with a sub-carrier
bandwidth of 7.5 kHz (each over a slot duration of 0.5 ms).

9
 LTE uses different sizes of frequency carriers (1.4 MHz, 3 MHz, 5 MHz, 10
MHz, 15 MHz, and 20 MHz). LTE is available in both FDD (Frequency Division
Duplex) and TDD (Time Division Duplex) modes. Many deployments are based
on FDD in paired spectrum. The TDD mode, however, is important in enabling
deployments where paired spectrum is unavailable. They enable LTE to be able
to operate more efficiently with respect to the use of spectrum, and also to
provide the much higher data rates that are being required.

Figure 1.6 MIMO in downlink and uplink in LTE radio interface, [12].

Other novelties in LTE radio interface (considering the 3GPP

standardization) for increasing the spectral efficiency (that is provision of more
bit/s per Hertz) is the introduction of:
 MIMO (Multiple Input Multiple Output) – that is multiple streams over the
same frequency bandwidth between the base station and the mobile
terminal (Figure 1.6). For example, 2x2 MIMO means having two antennas
on each end of the LTE radio interface (base station and mobile terminal),
which gives 2 times higher bitrates over the same frequency bandwidth
than communication with one antenna on each end. In that manner, 4x4
MIMO gives 4 times higher throughput, and so on. MIMO can be used in
one direction only (e.g., DL – downlink) or in both directions (e.g., DL and
UL – Uplink).
 Higher modulation schemes, which results in transmission of more bits
per transmitted so-called symbol over radio interface, which also increases
the bitrate (i.e., throughput). For example, 64QAM modulation has approx.
50% higher throughput than 16QAM modulation on the same frequency
bandwidth (e.g., if 20 MHz frequency carrier with 64 QAM has 300 Mbit/s,
the same LTE carrier of 20 MHz with use of 16 QAM will provide only up to
200 Mbit/s, and with QAM modulation will provide only up to 75 Mbit/s).

LTE and LTE-Advanced have the same radio interface and spectrum
utilization (in bit/s/Hz), with carrier aggregation standardized for LTE-Advanced
which provides the required 4G bitrates in downlink and uplink (according to IMT-
Advanced ITU’s umbrella for 4G).

10
1.2.3 LTE-Advanced

LTE-Advanced is an evolution of the LTE designed to meet or even

exceed the IMT-Advanced requirements set by ITU-R. The built-in backward
compatibility of LTE-Advanced has a direct implication in the way that, for an LTE
terminal a network with LTE-Advanced capabilities should appear as an LTE
network only, while for LTE-Advanced terminals it will provide full capacity of the
LTE-Advanced radio interface. The most important improvements in LTE-
Advanced access can be summarized as follows:
 Better flexibility in spectrum management for wideband deployments
by using carrier aggregation across different frequency bands.
 Flexible and faster network deployment achieved with the help of
heterogeneous networks feature (relay nodes, femto and pico cells,
besides typical macro cells in the radio access network).
 Improved radio network coverage as well as spectral efficiency (at the
cell edges and the average one), which is achieved through robust
interference management.
 Higher peak user bitrates by incorporation of higher order MIMO
(Multiple Input Multiple Output) in downlink and uplink (however, MIMO is
not exclusively used in LTE-Advanced, but it is also used in LTE and
HSPA).

Figure 1.7 Some of the key features in LTE-Advanced, [12].

So, there are several important improvements in the radio interface in

LTE-Advanced, which are aimed in fulfilling the goals for higher capacity, better
cost-efficiency and greater flexibility. There are several other features that
provide enhancements in LTE-Advanced for both, capacity and coverage issues.
They include, but are not limited only to, the following (Figure 1.7):

11
  Advanced MIMO techniques: Extension to up to 8-layer transmission in
downlink, introduction of single-user MIMO up to 4-layer transmission in
uplink, enhancements of multi-user MIMO, improvement of peak data rate
and capacity.
 Heterogeneous network and eICIC (enhanced Inter-Cell Interference
Coordination): Interference coordination for overlaid deployment of cells
with different Tx power, improvement of cell-edge throughput and
coverage.
 Relay: relay supports radio backhaul and creates a separate cell and
appears as LTE eNodeB to LTE UEs, and improvement of coverage and
flexibility of service area extension.
 Coordinated Multi-Point transmission and reception (CoMP): Support
of multi-cell transmission and reception; improvement of cell-edge
throughput and coverage.

For capacity purposes MIMO (Multiple Input Multiple Output) multiplies the
available bitrates over the same spectrum by using multiple transmit and receive
antennas. For example, 4x4 MIMO gives 2 times higher bitrate than 2x2 and four
times higher than SISO (Single Input Single Output). However, MIMO is required
on both ends of the radio link, base station (that is the eNodeB in the case of
LTE/LTE-A) and mobile station, to gain the benefit from it.

LTE-Advanced Carrier aggregation

Carrier aggregation is considered as one of the main features of the

LTE-Advanced (Figure 1.8) for provision of higher bit rates than LTE and for
fulfilling the IMT-Advanced requirements for 4G set by ITU. LTE introduced
flexible spectrum allocations by specification of different spectrum bands for the
radio interface, including FDD (Frequency Division Duplex), i.e., paired bands,
and TDD (Time Division Duplex), i.e., unpaired bands. LTE-Advanced provides
possibility for spectrum allocations of up to 100 MHz. That is accomplished by
carrier aggregation of multiple (separate) component carriers on the physical
layer, with the standard LTE frequency allocations of 1.4 MHz, 3 MHz, 5 MHz, 10
MHz, 15 MHz, and 20 MHz.
Maximum carrier aggregation in LTE-Advanced is 5 bands, each band
with maximum 20 MHz giving overall maximum frequency band allocation of 100
MHz. However, it is not always possible for mobile operators to have large
amount of contiguous spectrum allocation. Therefore, LTE-Advanced provides
possibility for flexible carrier aggregation of up to 5 non-continuous component
carriers, as shown in Figure 1.8. Generally, for carrier aggregation in LTE-
Advanced can be used adjacent or non-adjacent component carriers in the
same frequency band (intra-band aggregation), as well as component carriers
in different frequency bands (inter-band aggregation).
Regarding the protocol layering model, the carrier aggregation is
implemented on OSI layer-2, as shown in Figure 1.8. In particular, it is
implemented above the MAC (Medium Access Control) sublayer, on the RLC

12
(Radio Link Control) sublayer, which is below the network layer (that is IP in
LTE/LTE-Advanced). So, looking from the network layer above (i.e., above the
radio interface, which is consisted of OSI layer 1 and OSI layer 2) the carrier
aggregation provides aggregate bit rates obtained from all carriers, while each
physical layer and it corresponding MAC sublayer are distinct per frequency band
(each LTE carrier has band from minimum 1.4 MHz up to maximum 20 MHz).

Figure 1.8 LTE-Advanced carrier aggregation [3].

Table 1.2 Theoretical maximum LTE/LTE-Advanced bitrates, [2].

Maximum downlink Maximum uplink
bitrate [Mbps] bitrate [Mbps]
LTE, 2x2 MIMO, 64 QAM in
downlink, 16 QAM in uplink, 172.8 57.8
carrier width 20 MHz
LTE, 4x4 MIMO, 64 QAM in
downlink, 16 QAM in uplink, 326.4 86.4
carrier width 20 MHz
LTE-Advanced (5 aggregated
LTE carriers with total of 100 MHz 3000 1500
spectrum)

13
 The aggregate bitrates for LTE and LTE-Advanced are summarized in
Table 1.2. One should note that the given bitrates in the table are only theoretical
maximums and shared among the users on the same frequency carrier(s).
Individual user obtained bitrates depend upon the number of users that are using
the given capacity at the same time and on the capabilities of their mobile
devices (i.e., their User Equipment).
The theoretical maximum bitrates for the UE (e.g., smartphone) are those
that are supported at both ends of LTE/LTE-Advanced radio link, that is – at
eNodeBs in the given mobile network, and at the mobile terminal (i.e., UE). For
example, if mobile terminal has UE 5 (that is up to 300 Mbit/s in downlink) and
eNodeB supports up to 3 Gbit/s in downlink (i.e., 3000 Mbit/s) in the given cell,
then the theoretical maximum bitrate would be the minimum of the two (in this
example, it would be 300 Mbit/s, not 3 Gbit/s).
Why theoretical bitrates are called theoretical?
That is because they are never reached in practice due to many mobile
users sharing the same data link in the given cell in the network, and due to
statistical multiplexing of multiple traffic flows from different users in the network.

14
1.3 4G Evolved Packet Core (EPC)

System Architecture Evolution (SAE) for LTE provides a new approach

for access (E-UTRAN) and core (EPC) network parts, enabling the much higher
data rates that are possible with LTE/LTE-Advanced, and on a long term
resulting in reduced CAPEX and OPEX (for mobile telecom operators) when
compared to previous mobile systems (such as 3G) because all traffic is
transferred over the same all-IP mobile network (no need for parallel operation of
circuit-switched and packet-switched network parts), thereby providing higher
levels of efficiency in the design as well as operation and maintenance of the
mobile networks.

1.3.1 EPC network nodes

EPC (Evolved Packet Core) consists of several types of gateways which
are targeted to carry either control or user data traffic, or both. The main
functionalities in EPC are in fact regrouped from initial SGSN (Serving GPRS
Support Node) and GGSN (Gateway GPRS Support Node) as central and edge
gateways in 2.5G (i.e., GPRS) and 3G (i.e., UMTS/HSPA) mobile networks from
3GPP. Main functions refer to mobility management, QoS support, and AAA
(Authentication, Authorization and Accounting) functions. In that manner, there
are three main gateway nodes in the EPC:
 Mobility Management Entity (MME): It is the main control element in the
EPC, which is responsible for signaling-related mobility management
including user tracking, paging procedures, as well as bearers’ activation
and deactivation. Because mobility is related to roaming between different
mobile networks, MME has functionality for authentication of the user, and
therefore it uses signaling interface towards the users' database HSS
(Home Subscriber Server). Also, the MME provides interfaces for
interconnection with previous 3GPP core networks from 2G and 3G, for
handling the mobility between the heterogeneous mobile networks. It is
control only network node (no user data passes through the MME).
 Serving Gateway (S-GW): It is main gateway for user traffic (i.e., user
plane) within the core network, which establishes bearers for user traffic
with eNodeBs (base stations in LTE/LTE-Advanced) on one side, and with
the P-GW on the other side (towards the global Internet). Since MME is
controlling the mobility, S-GW has a control interface with MME, which is
important for bearer switching at handovers (e.g., a mobile terminal makes
a handover from one eNodeB to another, thus resulting in a change of the
bearer between the S-GW and the eNodeBs).

15
 Figure 1.9 EPC (Evolved Packet Core) network environment, [1].

 Packet Data Network Gateway, i.e. PDN Gateway (P-GW): It provides

connectivity from mobile terminals to external packet data networks. P-GW
also acts as an anchor point between the 3GPP core network and non-
3GPP networks (e.g., WiFi, etc.). Also, it performs typical functionalities for
edge routers of a given core network, such as packet filtering, policies
enforcement, charging support, etc. P-GW and S-GW are connected via
S5/S8 interface, and both nodes are parts of so-called SAE Gateway (SAE
GW).

The architecture of EPC is shown in Figure 1.9. Besides the three main
nodes (MME, S-GW, and P-GW) there are two other important control nodes in
the EPC, given as follows:

16
  Home Subscriber Server (HSS): It is a centralized database in the
mobile network which contains the user-related information, such as
location of the user (i.e., the MME to which it is connected to) and
subscriber profile including available services to the user, allowed PDN
(Packet Data Network) connections or roaming in visited networks, etc.
User authentication is performed via the Authentication Center (AuC),
which is usually a part of the HSS.
 Policy and Charging Rules Function (PCRF): It is a software-based
network node in the EPC which is responsible for Policy and Charging
Control (PCC). The PCRF detects the service flow and determines policy
rules for it in real-time. Also, it enforces charging policy in the mobile
network. In the SAE architecture the PCRF provides PCC information for
bearer setup to the enforcement function located in P-GW, which is named
PCEF (Policy and Charging Enforcement Function). When user
applications communicate directly with the IMS, the Application Function
(AF) in service domain (i.e., service stratum according to the NGN
terminology) requires from PCRF to apply PCC rules for dynamic policy or
charging control. On the other side PCRF is a central node in the general
NGN architecture connecting service stratum and transport stratum.

The main EPC nodes (MME, S-GW, and S-GW) with aim to provide their
functions are interconnected with eNodeBs and between each other via so-called
interfaces (standardized by 3GPP). Since 4G mobile networks are all-IP (i.e.,
EPC and E-UTRAN are all-IP networks) all interfaces are based on Internet
technologies, i.e., standardized IETF protocols (from network layer up to
application layer).

1.3.2 Control Plane and User Plane protocols in 4G

Although 4G mobile networks by 3GPP are referred to as all-IP, most of

the QoS provision and mobility management is based on different protocol stack
used between different pair of nodes (Figure 1.10). In general, there are two
planes: user plane (used for data originated or terminated from/to mobile
terminals) and control plane (used to carry signaling traffic, such as AAA
signaling, mobility and location management signaling, service-related signaling,
etc.).
For example, eNodeBs (i.e., base stations for LTE/LTE-Advanced) are
connected between each other (neighboring ones, that can make handover of
connections one to another) and additionally have interfaces with central
nodes MME and S-GW. Each of the interfaces is in fact based on standardized
(and sometimes specific) protocols and mechanisms that belong to the Internet
technologies. In that manner, the control traffic on X2 interface (provides
connections between eNodeBs and supports mobility for UE (User Equipment) in
connected mode) is carried over SCTP/IP to provide reliable transmission, where
SCTP (Stream Control Transmission Protocol) is a TCP-based multi-homing

17
and multi-streaming protocol that provides possibility to use parallel streams
through different paths between two end-points of the communication, thus
increasing the reliability needed for signaling traffic.

Figure 1.10 Internet protocol stack in 4G mobile networks:

a) Control plane; b) User plane, [2].

If one considers the interface between UE, i.e., mobile terminals, and
eNodeB, then it may conclude that it is working on layer 2 (with PDCP over RLC
over MAC over Physical layer), and then eNodeB does relay of the IP packets
into GTP (GPRS Tunneling Protocol) tunnels between eNodeB and all other

18
nodes in the mobile network to which it is connected to (other eNodeBs, S-GW,
P-GW, and MME).

Figure 1.11 User Plane and Control Plane protocols in 4G mobile networks, [2].

Then, it is referred to as GTP-U for transport of user plane traffic and

GTP-C for control plane traffic (Figure 1.11).
GTP uses UDP/IP protocol stack for user traffic and SCTP/IP for
control traffic, on all interfaces between each pair of nodes (except the radio
access links). So, the protocol layering in network nodes of 4G mobile networks
is based on tunneling protocols over UDP/IP or SCTP/IP protocol stacks, with
application part on the top of protocol stacks (e.g., X2AP is Application Part on
X2 interface, which connect two adjacent eNodeBs).
The separation of User Plane (UP) and Control Plane (CP) continues
further in 5G mobile network architectures (which are covered further in this
course).

19
1.4 4.9G: LTE-Advanced-Pro

From Release 13 the 3GPP has approved a new LTE marker for the LTE-
Advanced specification in Releases 13 and 14, called LTE-Advanced Pro.
LTE-Advanced Pro does more than just push LTE capabilities closer
towards 5G. It is also considered as an integral part of the 5G mobile network by
provision of many services which are (or will be) essential to the 5G experience.
Figure 1.12 shows the progress of LTE capabilities towards 5G with LTE-
Advanced Pro releases.
The new term “Pro” is intended to mark the point in time where the LTE
platform has been dramatically enhanced to address new markets as well as
adding functionality to improve efficiency.

Figure 1.12 Progress LTE capabilities towards 5G [14].

1.4.1 LTE-Advanced Pro characteristics

There are multiple new capabilities introduced in LTE-Advanced Pro (i.e.,

LTE-A Pro). One key capability that LTE-Advanced Pro is enabling globally is
Gigabit LTE which brings many new immersive experiences, including virtual
reality, or one may say it simpler – it provides higher speeds in mobile network
per user and bigger data caps for mobile Internet users (that is also important to
use new services, such as video with higher resolutions, which are traffic
bandwidth demanding).
One may say that just like LTE sought to provide higher mobile data
transfer rates than legacy 3G, LTE-Advanced aimed to improve on LTE (and also

20
to fulfill the requirement set by the ITU for 4G in IMT-Advanced umbrella), LTE-
Advanced Pro does the same and has given the main preparations for the 5G.
Essentially, one may refer to LTE as 4G, LTE-Advanced as 4.5G, and
LTE-Advanced Pro as 4.9G or pre-5G (Figure 1.13). However, that is more a
jargon (e.g., saying LTE-A Pro is 4.9G).

Figure 1.13 Development of 4G technologies: from 4G to 4.9G, [15].

LTE-Advanced Pro (shortly LTE-A Pro) is built on previous iterations to

further improve the LTE platform as well as address new use cases. The major
advances achieved with the completion of Release 13 include: MTC (Machine
Type Communication) enhancements, public safety features such as D2D
(Device to Device communication), small cell dual connectivity, further carrier
aggregation enhancements (when compared to LTE-Advanced), new
approaches for interworking with WiFi, licensed assisted access (at 5 GHz),
3D/FD-MIMO, indoor positioning, single cell-point to multi-point and work on
latency reduction. Many of these features were started in previous Releases, but
become mature in LTE-A Pro.
In general, the LTE-Advanced Pro incorporate a number of different
technologies, many of which are in fact more advanced forms of features already
present in LTE-Advanced and even LTE, tracing the path to 5G (Figure 1.14).

Figure 1.14 LTE-Advanced Pro essential services towards 5G, [14].

21
Evolving towards massive MIMO

Beamforming and MIMO (Multiple Input Multiple Output) have been

identified as key technologies to address future capacity demand. Prior to LTE-
Advanced, beamforming focused on the azimuth dimension, essentially at ground
level. LTE-Advanced began investigations into elevation beamforming, where
antenna arrays target taller buildings. LTE-Advanced Pro further exploits so-
called Full Dimension MIMO (FD-MIMO), which simultaneously supports both
elevation and azimuth beamforming to significantly boost capacity and coverage.
At the same time, the use of many more antennas will evolve towards Massive
MIMO, which is considered as key enabler for 5G. Regarding the LTE-Advanced
Pro releases, 3GPP Release 13 saw support for 8, 12 and 16-antenna elements,
while Release 14 added support with up to 64 antenna ports at the eNodeB.

Device to Device (D2D) communication

The device-to-device (D2D) technology enables the discovery of

thousands of devices within proximity of around 500m to facilitate proximity
services. Work on proximity services has focused on public safety applications
began in LTE-Advanced projects, and continued into LTE-Advanced Pro.
3GPP Release 13 further aimed to allow operators to multiplex reduced
bandwidth machine-type communications (MTC) devices and regular devices
in existing LTE networks. A specific area of focus in 3GPP Release 14 is D2D
and device-to-network relays for IoT and wearables, considering how remote
MTC devices like wearables connect to other devices, which in turn connect to
the access network.
The use of LTE for emergency services and public safety was an
important aspect of Release 13, culminating in the first set of specifications
covering mission-critical services, notably the mission-critical Push-To-Talk
(MCPTT) protocol. This work continued in Release 14 with definition of mission-
critical video and mission critical data services.
A key project within R14 is Vehicle-to-anything (V2X) communications,
which will ultimately support vehicles connected with infrastructure, people and
networks to enable autonomous (self-driving) cars. V2X includes two
complementary transmission modes:
 Direct communication: Building upon LTE D2D with enhancements for
high speeds, high density, improved synchronization and low latency.
 Network communication: Enabling broadcast of messages from a V2X
server to vehicles and beyond; Vehicles can send messages to server via
unicast.

Many of the features of 5G in fact are being tested in LTE-Advanced Pro

networks around the world creating a strong foundation for 5G to grow out of.
Some may consider that LTE-Advanced Pro is to 5G similar to the first LTE
(Release 8) was to 4G around 2010 (i.e., a decade ago).

22
1.4.2 LTE-Advanced-Pro Carrier Aggregation enhancements

Carrier Aggregation enables faster data speeds by increasing the

bandwidth available to devices by using more than one carrier (channel). It was a
key feature of LTE-Advanced specifications, supporting aggregation across FDD
and TDD (frequency division duplex and time division duplex) spectrum types.
Advanced carrier aggregation in LTE-Advanced Pro increases the number of
different component carriers that can be simultaneously supported from five
(under LTE-Advanced) to 32 (with LTE-A Pro).
One may note that the carrier aggregation in LTE-Advanced-Pro (Figure
1.15 and Figure 1.16) is in fact further improvement in carrier aggregation defined
for LTE-Advanced in 3GPP Releases 10 to 12. In LTE-Advanced-Pro the
theoretical maximum spectrum allocation theoretically goes up to 32 x 20 MHz
= 640 MHz, however how much spectrum will be allocated on the ground
depends on the availability of licensed spectrum by mobile operators as well as
combinations of the licensed and unlicensed spectrum.

Figure 1.15 LTE-Advanced Pro carrier aggregation (up to 32 component

carriers), [1].

Overall, the main technologies which can be “sensed” in practice by the

end users (e.g., using the mobile Internet access services) with LTE-Advanced
Pro are the following:
 Advanced Carrier Aggregation: LTE-A Pro however increases this to 32
carriers, delivering a maximum aggregated bandwidth of 640 MHz. That
makes individual peak rates up to 1 Gbps to become realistic (however, the
really available individual bitrates depends upon the mobile network design
and installed capacity versus number of users with user equipment that
supports the new mobile technologies).
 Increased network capacity and efficiency: LTE-Advanced Pro roughly
doubles the network capacity without the need for any additional spectrum
or base stations, that is, provides more bits per second per Hz. How? Well,
the use of 4×4 MIMO antennas increases capacity by allowing multiple
transmit and receive signals simultaneously, and also higher modulation
schemes, such as 256 QAM modulation scheme (an increase from 64
QAM) which makes it possible to carry more bits of data per symbol,
increasing throughput and making better use of the spectrum.

23
 Figure 1.16 LTE-A Pro carrier aggregation by protocol layers (OSI layers 1 to 3), [1].

Other benefits (that can be felt) delivered by LTE-A Pro over earlier LTE
technologies include longer battery life and a closer alignment with 5G for
improved network future-proofing.

24
1.5 4.9G QoS and mobile Internet access

All mobile networks from 2G to 5G have QoS support and provide service
differentiation (which is more complex in mobile networks due to handovers). The
QoS support in mobile networks standardized by 3GPP (up to 4G) is
accomplished by the definition of so-called bearers.
What is a bearer?
The bearer service is enabler for any data transmission between
defined two end points in the mobile network, including the mobile Radio
Access Network (RAN) and mobile core network.
For example, in 3G UMTS the bearer may refer to control (i.e., signaling)
or data information. For example, there is SMS (Short Message Service) bearer
for SMS, and USSD (Unstructured Supplementary Service Data) for USSD
messaging, both provided via GSM signaling. Then, in 2G and 3G (standards
from 3GPP) there are voice CS (Circuit-Switched) bearer for voice services and
PS (Packet-Switched) bearer for Internet access service. In all cases, the bearer
specifies the configuration on protocol layer 2 (i.e., data-link layer) and layer 1
(i.e., physical layer) of the protocol stack, with aim to have defined QoS attributes
and traffic flow characteristics. The concept of bearer services further continues
in the same manner in 4G standards from 3GPP, LTE/LTE-Advanced/LTE-A-Pro.

1.5.1. QoS for LTE/LTE-Advanced

Quality of Service (QoS) support in 4G/4.5/4.9G mobile broadband

networks is based on definition of QoS classes and QoS parameters and their
mapping on defined bearers. Satisfying the QoS requirements of the various
applications and services entails quantifying these requirements in terms of
parameters that identify target performance levels. Such parameters include
throughput (downlink and uplink), delay, jitter, and packet loss, as the most
important ones.
In that way, the 4G/4.5/4.9G mobile networks by 3GPP have bearers on
different layers, from physical layer radio bearers up to end-to-end service
bearers. The major quantitative parameters:
 Throughput: Characterized through the guaranteed bit rate, maximum bit
rate, and aggregate maximum bit rate
o Guaranteed bit rate (GBR): Allocated fixed network resources that
do not change after bearer establishment or modification. This is
hence a guaranteed service data flow.

25
 o Maximum bit rate (MBR): Limits the bit rate that can be expected
to be provided to GBR bearer, and is enforced by the network
shaper to restrict the traffic to its MBR agreement.
o Aggregate maximum bit rate (AMBR): Used for non-GBR flows,
and has two types, Access Point Name-AMBR (APN-AMBR) and
user equipment-AMBR (UE-AMBR). The APN-AMBR parameter
refers to the maximum bit rate that can be consumed by all non-
GBR bearers and all PDN connections of the APN. The UE-AMBR
parameter, on the other hand, refers to the MBR allowed for all non-
GBR bearer aggregates for the respective UE (enforced in both the
downlink and the uplink).
 Delay: 3GPP initially defined nine categories for delay, with 50 ms being
the tightest and 300 ms the slackest. The latter value is used for delay
tolerant applications. LTE-A Pro adds additional categories in the basket,
with lower delay requirements (for new services).
 Packet loss: Defined as the packet error loss Rate, similar to the packet
delay budget in having nine categories with 10exp(–6) being best and
10exp(–2) being the worst.
 Priority: Specified by the Allocation/Retention Priority (ARP)
parameter, this is used to indicate the priority of both allocation and
retention of the service data flow. The ARP dictates whether a bearer
establishment/modification request can be accepted or rejected in the
event of conflicts in demand for network resources. At times of exceptional
network resources limitations, such as handover, ARP can be used by the
eNodeB to drop a flow with a lower ARP to free up capacity. ARP,
however, has no effect on the network treatment received by the flow once
the flow is successfully established.

Note that GBR and MBR are defined per bearer, while the AMBR
parameters are defined per group of bearers. All throughput parameters have two
components, one for downlink and another for uplink.
Moreover, the traffic from different services in classified in several QoS
classes, in that manner:
 UMTS (as 3G standard by 3GPP) has 4 different QoS classes:
conversational (e.g., for voice over IP), streaming (e.g., video streaming),
interactive (e.g., Web browsing) and background (e.g., email).
 LTE/LTE-Advanced (as 4G standard by 3GPP) mobile networks have 9
different QoS Class Identifiers (QCI), as further evolution from UMTS QoS
framework.

Radio access network is normally the most sensitive to QoS. Why?

Because users are mobile and their location is varying over time so the quality
they experience also varies. So, different QoS techniques are exploited on
different protocol layers in the access:

26
  Physical layer (layer 1) includes channel adaptation, Channel Quality
Indicator (CQI) mapping, power control (the limiting direction is the uplink
due to limited output power of mobile terminals);
 Layer 2 provides link adaptation, link-layer scheduling, and error control
on the link layer (with HARQ);
 Network layer (layer 3) includes QoS management over the access and
core networks, admission control, persistent scheduling in mobile network
nodes.

The 3GPP distinguishes between bearers at the different network levels. A

radio bearer, for example, is the over-the-air connection between the UE and the
RAN, while the S1 bearer is that between the eNodeB and the EPC network
entities.
Generally, in 4G/4.5G 3GPP mobile broadband, QoS is provided between
UE and PDN Gateway and is applied to the Evolved Packet System (EPS)
bearer approach via the following three main bearers (Figure 1.17):
 Radio bearer (on the LTE/LTE-Advanced radio interface);
 S1 bearer (on S1 interface, between eNodeB and S-GW);
 S5/S8 bearer (on S5/S8 interface, between S-GW and P-GW).

Figure 1.17 EPS bearers for LTE/LTE-Advanced mobile networks, [2].

An EPS bearer is characterized by the following main parameters:

 QoS Class Identifier (QCI): It is a scalar value which defines a set of QoS
parameters such as delay and packet loss/error rate, and sets the priority
of the IP packets depending of their type and the policy decision of the
mobile operator (the targeted applications are not always the ones that are

27
 used for certain QCI, because each network is autonomous in definition of
its QoS rules).
 Allocation Retention Priority (ARP): This parameter additionally defines
the priority used for the allocation and retention mechanisms for the same
QCI. The ARP was also used in UMTS mobile networks for QoS
differentiation within the same traffic class.
 Guaranteed Bit Rate (GBR): It refers to guaranteed (minimum) bitrate for
certain real-time services provided by mobile operators such as carrier-
grade (i.e., QoS-enabled) voice and live streaming.
 Maximum Bit Rate (MBR): it refers to set a limit on the bitrate which is
expected for the given service.

The GBR parameter is used to differentiate between resource types in

LTE/LTE-Advanced mobile networks. In general, there are two resource types:
 with guaranteed bitrates (i.e., GBR), and
 with non-guaranteed bit rates (i.e., non-GBR).

However, each of the resource types has several QoS Class Identifiers
which are targeted to be used for QoS differentiation in LTE/LTE-Advanced
mobile networks.
Bearers can also be classified into default and dedicated. A default
bearer is initiated and established at startup to carry all traffic toward a
destination. The default bearer is a non-GBR bearer and does not provide bit
rate guarantees.

Table 1.3. QoS Class Identifiers (QCI) for LTE/LTE-Advanced (3GPP Release 10), [2].

Resource Delay Loss

QCI Priority Example application
Type budget rate

1 2 100 ms 10-2 VoIP

2 4 150 ms 10-3 Video call

GBR
3 5 300 ms 10-6 Video streaming

4 3 50 ms 10-3 Real-time gaming

5 1 100 ms 10-6 IMS signaling

Voice, live video, Interactive

6 7 100 ms 10-3
gaming
Non-GBR
7 6 TCP applications (web, email,
-6 p2p file sharing, http video,
8 8 300 ms 10
chat, buffered video streaming,
9 9 etc.)

28
 A dedicated bearer, on the other hand, can be either a GBR or non-GBR
bearer. If dedicated bearer is GBR type, it can specify the bitrate that is
guaranteed, packet delay, and packet loss error rate. Each dedicated bearer is
characterized by a Traffic Flow Template (TFT) detailing the bearer’s QoS
parameters. An uplink TFT is used to map the UE uplink traffic to specific QoS
parameters, with the mapping carried out at both the eNodeB and the UE.
Mapping for the downlink TFT is carried out at the S-GW or P-GW.
Table 1.3 gives an example of a traffic classification based on the QoS
parameters defined in the LTE QoS framework. Each class is identified by a
scalar number, which is the QoS Class Identifier (QCI). In fact, QCI identifies a
group of QoS parameters describing the packet forwarding treatment in terms of
priority, allowable delay, and packet error rate.
The traffic generated by a particular application can be differentiated into
separate Service Data Flows (SDFs) and mapped to a particular bearer, based
on specific parameters provisioned either in the PCRF or defined by the
application layer signaling. SDFs mapped to the same bearer are treated
similarly from the QoS standpoint.

1.5.2. QoS for LTE-Advanced Pro

Initially, LTE/LTE-Advanced had 9 QCI, however with later releases (e.g.,

3GPP Release 14, which is the last release before the appearance of the 5G,
which appeared with 3GPP Release 15) there was needed further granulation
that introduced several new QCIs, having priorities in between some of existing
(initial) QCIs, as shown in Table 1.4.
From all types of traffic, highest priority is given to IMS signaling
although it is served by non-GBR resource type. Signaling has not predefined
bitrates, because it depends on the load (e.g., volume of voice call setup/release
signaling), but on the other side it must have the highest priority because it is
prerequisite for establishing call/sessions (e.g., voice calls over the mobile
network). Further, priority two is given to voice services over LTE (e.g., VoIP
over LTE – VoLTE), however the delay budget for voice is more flexible that for
example required delay for real-time gaming (which is an emerging service) or
V2X (Vehicle to X) messages.
For mission critical services, such as Mission Critical Push To Talk
(MCPTT) voice, and non-mission critical services there are defined QCIs 65
and 66, respectively, which were later added to the initial 9 QCIs. In similar
manner QCI = 75 is also defined with GBR resource types, with very low delay
requirement for the V2X communication. QCI value 2 is targeted to video calls
(and live streaming) which is sensitive to delay (e.g., in video calls the video has
the same delay requirements as voice, due to needed synchronization between
the audio and the accompanied video), while QCI = 4 is targeted for transfer of
buffered video (i.e., Video on Demand – VoD) provided by the mobile operator.

29
 Table 1.4. QCI for LTE-Advanced Pro (3GPP Release 14 – new QCIs are marked with
green color), [1].
Resource Delay Packet
QCI Priority Targeted services
Type budget Loss rate
1 2 100 ms 10-2 Voice over IP (VoIP)

2 4 150 ms 10-3 Video call and live streaming

Real-time gaming, V2X
3 3 50 ms 10-3
messages

GBR Buffered video streaming

4 5 300 ms 10-6
(Guaranteed (not conversational)
Bit-Rate)
Mission Critical Push To Talk
65 0.7 75 ms 10-2
(MCPTT) voice
Non-Mission Critical Push To
66 2 100 ms 10-2
Talk voice
Vehicle-to-X (V2X)
75 2.5 50 ms 10-2
messages
5 1 100 ms 10-6 IMS signaling
Buffered video streaming,
interactive TCP-based
6 6 100 ms 10-3
applications (e.g., Web,
email, file sharing, etc.)
Voice, live video streaming,
7 7 100 ms 10-3
interactive gaming
Buffered video streaming,
Non-GBR interactive TCP-based
8 8 300 ms 10-6
(Non applications (e.g., Web,
Guaranteed email, file sharing, etc.)
Bit-Rate)
9 9 300 ms 10-6
Mission Critical signaling
69 0.5 60 ms 10-6
(e.g., MCPTT signaling)
Mission Critical Data from
buffered video streaming,
70 5.5 200 ms 10-6 interactive TCP-based
applications (e.g., Web,
email, file sharing, etc.)
79 6.5 50 ms 10-2 V2X messages

All services served with GBR are in fact carrier-grade services, because
for Internet access services there is no possibility to guarantee the bitrates due to
network neutrality principle which is applied also for mobile Internet access.
Therefore, for Internet based application and services are typically used QCIs
values 8 and 9, which are served with the lowest priority, targeted for use by

30
TCP-based applications (e.g., Web, email, Web-based services such as social
networking, picture and files sharing, etc.).
There are defined also non-GBR QCIs for voice, video calls, live video,
mission critical and non-mission critical services. There are additionally defined
QCI = 69 for mission critical signaling (which has highest priority of all, noted as
0.5, which is higher priority than IMS signaling), QCI = 70 for mission critical data
services and other TCP-based based services (such as Web-based, email, etc.)
with priority between GBR and non-GBR buffered video streaming, and QCI = 79
for V2X messages (with priority set in the priorities range for non-GBR real-time
and interactive services such as video streaming, voice, gaming, browsing, etc.).
However, with aim to provide the required QoS there mobile network
needs binding of PCC and QoS rules with service data flows. The binding
mechanism includes three steps:
 Session binding: It is performed by PCRF node, and this binding includes
IPv4 or IPv6 network prefix (for the mobile terminal’s IP address) as well
as user equipment identity.
 PCC rule authorization and QoS rule generation (where applicable): It
refers to the selection of the QoS parameters (including QCI, ARP, GBR,
MBR, etc.) for the PCC rules.
 Bearer binding: It is an association of PCC rule and QoS to a given
bearer.

The QoS rules in the mobile network are derived from the PCC rules. The
PCRF is the node that can activate, modify and deactivate QoS rules over given
reference (i.e., network interface) point.

1.5.3 Testing Mobile Internet access

Mobile networks which currently are dominant in the mobile world, based
on 4G standards from 3GPP, are primarily used for mobile broadband
Internet access, with bitrates in range from several tens of Mbit/s up to several
hundreds of Mbit/s.
The real individual bitrates (per mobile user) depend upon the installed
capacity in the mobile network versus average number of connected mobile
users (higher number of simultaneous connections for a given capacity, results
in lower bitrate per connection). However, the bitrates also depend upon
capabilities of the mobile terminals (e.g., smartphones and supported
categories for end devices) and business approaches by telecom operators in a
given country (e.g., available mobile data cap for ‘mobile data’ as a service, e.g.,
is it 20 GB, 100 GB per month, or unlimited mobile data access).
As usual, lower prices of mobile data increase the usage of the mobile
network for data traffic (e.g., may result in congestion in poorly designed and
planned networks) and vice versa. Also, willingness to pay for mobile data (i.e.,
mobile Internet access, which is interchangeably used here with the term –
‘mobile data’) also influences the traffic intensity per mobile user (measured in

31
bits per second). But, one should not forget that online services (also referred to
as OTT – Over The Top service) are highly influential on mobile end users habits
and needs to use mobile Internet access for various purposes (e.g.,
communication with others, work, entertainment, education, business, shopping,
administration, etc.).
The above discussion points to the importance of the QoS for mobile data
services i.e. for mobile Internet access. Considering that 4G is the main mobile
network today for mobile data, testing mobile Internet access has importance
everywhere.
The test methodology aims at delivering an accurate measurement of the
maximum bandwidth available over a given Internet connection. This is
achieved by transferring multiple parallel data streams over separate TCP/IP
connections within a predefined amount of time. The transferred data consist of
randomly generated data with high entropy (e.g., not repetitive data). Also, it
should effectively avoid data compression during the transmission. In order to
increase the probability that the test can be performed even within networks
protected by firewalls and proxy servers, the data should be transferred over
hypertext transfer protocol (HTTP) or hypertext transfer protocol secure (HTTPS)
(i.e., HTTPS refers to HTTP accompanied with transport layer security (TLS) or
secure sockets layer (SSL) connection, for encryption of the HTTP connection
end-to-end). Typical HTTP client is Web browser (e.g., Chrome, Firefox, Opera,
etc.), which typically is used by the end users.

Figure 1.18 Example of broadband data measurements, [33].

The most important performance parameter for the broadband data is the
bitrate (although one should note that end-to-end delay is also very important for

32
high quality user experience). Then, Broadband data measurement campaigns
should be balanced on different monitoring systems to reduce the window test
frame and increase the number of samples; this will enable greater coverage of
the testing map area.
The setup for the scenarios will depend on the service application being
measured and the expected bitrates to be measured. Figure 1.18 shows an
example of a timeline diagram for the evaluation of mobile broadband Internet
access service (i.e., mobile data).
Measurements related to web browsing can be characterized, among
others, by measuring the performance of navigation in terms of page load time
and the overall success rate for page download. Web-browsing tests, with an
average fixed duration, are launched from mobile equipment or a probe
simulating a mobile subscriber in a static or mobile situation to a predetermined
set of recommended websites. Access attempts are randomly distributed
between the different sites tested.
File transfer measures consist of sending (in uplink) and receiving (in
downlink) files of a fixed size for uplink and downlink channels between mobile
devices or a probe simulating a mobile subscriber in a fixed location and a server
having resources dedicated to this measurement. KPIs such as download
session success rate, mean download time or mean upload time can be
measured (with limits placed on the time allowed for completion in either
direction).

1.5.4 QoS operational strategy for supervision on mobile broadband

While fixed broadband is more easily to be dimensioned and planned, due

to fixed access location for the end user premises, there is different situation in
the case of mobile networks. In mobile broadband networks the end users are
mobile and number of connected users connected to a base stations, as well as
the distance between the user device (e.g., smartphone) and base station (e.g.,
4G, 5G) and their velocity changes over time, which makes the QoS provisioning
more complex in mobile environment.
National regulatory frameworks for QoS can vary between countries
based on several factors, like the maturity of the market in terms of competition
or the level of cooperation between stakeholders. There are countries where the
national legislation or the QoS framework requires the involvement of regulators
in QoS supervision.
Regulators may set objectives and targets on mobile QoS for a defined
period. As per regulatory efforts to achieve these objectives, it is expedient for
regulators to assess their existing regulatory actions on mobile QoS and
determine whether each action was carried out in accordance with the principles
of cost-effectiveness, reliability and process efficiency.
In summary, the cardinal principles for any effective operational
management, is the achievement of high quality at reduced cost and
improved delivery time. These principles, and in particular the six viewpoint

33
approach that follows, offer a useful guide on how to implement a QoS
operational strategy. Figure 1.19 depicts a proposal for operational strategy
based on these principles, the objective being to bring regulatory interventions
and processes closer to the user experience. The implementation aspects of the
operational strategy as in the six viewpoints are deducible from ITU QoS
Regulation Manual, [39], which seeks to compile best practice from country case
experiences of QoS monitoring and enforcement.

Figure 1.19 Six-viewpoints of the QoS operational strategy for regulators, [38].

The six viewpoints of the QoS operational strategy for regulators and its
operational aspects include the following:
 QoS field monitoring refers to all measurement methodologies that rely
on information on QoS gathered at the level of the end user of the service,
i.e., at the interface provided by the mobile device. It can be based on
either non-intrusive measurement with real traffic (e.g., crowdsourced data
collection), or on intrusive measurements with test traffic (all measurement
types).
 Network performance monitoring refers to measurement methodologies
that rely on information on QoS gathered inside the network conveying the
service, i.e., network elements (nodes) and related interfaces, this
corresponds to network performance counters. Network performance
monitoring is based on non-intrusive measurement with real traffic (it can
be also based can be based on network performance counters or call
traces).
 Network incident monitoring refers to QoS conditions in respect of
network availability, fault repair time and impact of service outages should
be documented in the license, regulations or guidelines on QoS, where
applicable. The reporting and repair time requirements on network
incidents must be defined and enforced through the application of
appropriate penalties or compensations, depending on the legal
framework.

34
  Network audit provides the regulator with a better understanding of the
capability levels of the Service Provider’s network to translate its QoS
offered or planned into QoS achieved.
 Consumer satisfaction surveys consider the consumer's own
assessment of the QoS delivered by the SP. This is referred to as the QoS
perceived. This subjective measurement method can be done preferably
once a year and the results analyzed along with the QoS measurement
results obtained in the particular geographic area where the survey is
ongoing.
 QoS complaint management refers to operational aspect of the
regulator's activities on QoS assessment. In addition, this viewpoint seeks
to provide a database or an inventory of consumer complaints on QoS.

The adoption by the regulatory authority of an operational strategy to

measure and report the performance of both the communication networks and
the quality of the services provided to end users is useful for determining the
content of a national strategic plan for the development of mobile services in a
country and for the optimal use of the available radio spectrum. The result is the
development of a competitive environment between providers, thus ensuring a
favorable background for investment, while the gain for end users is the
possibility of choice of cost and quality in the network to which they will
decide to become subscribers.

35
1.6 Cellular Internet of Things (IoT): LTE-M and NB-IoT

The Internet of Things (IoT) has started to appear as a paradigm with the
spread of broadband access to the Internet around mid 2000s. In that manner,
the ITU has published its first book-report on IoT back in 2005, [17].
The Internet of Things (IoT) is emerging topics in the past almost two
decades, however from a concept it has grown to wide spread global
deployments using various radio technologies (e.g., LoRa, WiFi, cellular i.e.
mobile radio access technologies). In that manner, today we have billions of
“things” (e.g., meters, trackers, wearables, etc.) connected to the networks
(typically via radio access links) which “feed” databases and servers with data
that is used as input in various types of IoT applications/services.
But, what is the high-level definition of IoT?
According to the ITU, [18], IoT is a global infrastructure for the information society
enabling advanced services by interconnection of different physical and virtual
things, based on ICTs (Information and Communication Technologies). The IoT
defines new dimension to the ICT world, called “any-thing communication”, in
parallel with “any-time communication” and “any-where communication” (Figure
1.20).

Figure 1.20 Three Internet of Things (IoT) dimensions, [1].

The IoT refers to any communication between machines and objects

over Internet. The main challenge for Cellular IoT services is industry
standardization. With the development of 4G and 5G mobile networks,
particular attention is given to IoT segment, because it has spread due to
standardized mobile and wireless networks, and the only prerequisite is to have
end-to-end IP connectivity in many cases. In that manner, with the appearance of
GPRS as IP upgrade to initial non-IP GSM mobile network, started wide usage of

36
GPRS for various IoT devices (e.g., GPRS modems can be found in different
sectors, such as remote control of public utilities or measurement units, fleet
tracking, credit/debit card mobile payment devices, and many more).
Why is that possible, considering that GSM/GPRS (2G technology)
provides lower bitrates (e.g., tens of Kbit/s)? Because most of the IoT devices
(e.g., sensors, meters) do not require high bitrates for transfer of control
information to the IoT device and transfer of the collected data from the IoT
device to a given node (e.g., a server, a database) in the IP network. So,
broadband is not required for the most of the IoT world, but narrowband speeds,
which are in range of hundreds of kbit/s or couple of Mbit/s at the present time,
However, one may note that such narrowband bitrates needed for IoT today were
in fact broadband bitrates a couple of decades ago.
The 4G mobile networks LTE-Advanced and particularly LTE-Advanced-
Pro, which comes from the last 3GPP releases before 5G (LTE-Advanced-Pro is
standardized in 3GPP Releases 13 and 14,as already noted before in this
Module), are focusing in a greater portion to the IoT segment. Since IoT devices
are spread on a wide area, the typical way to connect them is through radio
connections. However, to provide their interoperability there is increasing
demand for standardization and unification of standards (e.g., few standards
supported by most of the devices), especially for long range technologies which
should provide WAN (Wide Area Network) coverage for IoT devices.

Figure 1.21 Long range IoT technologies, [22].

There are different IoT solutions on the market for long range IoT, which
are shown in Figure 1.21. There are 3GPP standards as well as non-3GPP
standards (e.g., LoRa, Sigfox, and others) for long-range IoT communications.
However, considering that all mobile networks today are 3GPP standards, and
they intended to provide worldwide coverage (including 4G and 5G) the focus in
this section is on LTE-A-Pro solutions IoT, such as LTE-M and NB-IoT, which
further continue and integrate also in 5G networks.

37
1.6.1 Cellular IoT standards in LTE-Advanced-Pro

Legacy cellular (i.e., mobile) options for data transfer in LTE networks
consume too much power and too much bandwidth, so they are not appropriate
for non-critical IoT deployments. Moreover, LTE as native mobile broadband
technology does not fit well with applications where only a small amount of data
is transmitted infrequently. In that direction, so-called Cellular IoT approach in
LTE is a respond to the continuous search for better low-power long-range
technologies.
The LTE-Advanced-Pro has focused on mobile Internet of Things (IoT)
by introducing new narrowband technologies for that purpose. Table 1.5
illustrates how the LTE-Advanced-Pro is developed for IoT support, including
further development of well proven (for IoT) GPRS technology. In particular, LTE-
Advanced Pro in Release 13 defines three technologies targeted to IoT markets
to be supported by LTE:
 LTE-M (Long Term Evolution of Machines) provides further LTE
enhancements for Machine Type Communications, which is also
abbreviated as LTE-MTC or eMTC (enhanced Machine Type
Communication)
 NB-IoT (Narrow Band Internet of Things) is defined as new radio added to
the LTE platform optimized for the low end of the market. It continues to
exist in 5G standards.
 EC-GSM-IoT (Extended Coverage GSM for Internet of Things) defines
EGPRS enhancements which make GSM/EDGE markets prepared for IoT
(so, GSM/EDGE continues to live in this decade also for IoTs).

Table 1.5. Mobile and Cellular IoT by 3GPP, [1].

LTE IoT technologies Description Bitrates

LTE/LTE-Advanced enhancement
for Machine Type Communications, 1 Mbit/s on smallest width
LTE-M (LTE for Machine
which is also abbreviated as LTE- LTE frequency carrier of 1.4
Type Communication)
MTC or LTE-M (Long Term MHz
Evolution for Machines)

New radio spectrum added to LTE, 10s to 100s of kbit/s

NB-IoT (Narrow Band
which is optimized for the low-end of (180 kHz narrowband
Internet of Things)
the IoT market frequency carrier)

EC-GSM-IoT (Extended Further enhancements on 10s to 100s of kbit/s

Coverage GSM for GSM/GPRS/EDGE technologies for (200 kHz GSM frequency
Internet of Things) the IoT segment carrier)

The work on new IoT devices is also targeted to longer battery life, lower
costs of the devices and deployments, as well as extended coverage. All of
these requirements are well satisfied with inclusion of IoT specific technologies

38
within the 3GPP mobile standards. However, one may note that LTE-M and
NB-IoT are designed for different types of IoT use cases.

1.6.2 LTE-M

The LTE-M uses only 1.4 MHz (on the other side, LTE carrier is up to 20
MHz, and LTE-Advanced/LTE-Advanced-Pro allows aggregation of multiple such
carriers) to provide up to 1 Mbps of throughput (e.g., possible uses are asset
trackers and wearables). On the other side, high bitrates in mobile networks are
reserved for more powerful devices (e.g., smartphones) which require at the
same time high throughput (e.g., ultra-broadband access speeds) and low
latency.
The main components of LTE-M are a series of low-cost device
categories (e.g. Cat-M1 and Cat-M2) and two coverage enhancement modes
(i.e., modes A and B). LTE-M was originally designed to reduce the device
complexity to make LTE competitive with E-GPRS in the MTC market. In addition
to its low complexity, it does support secure communication, ubiquitous
coverage, and high system capacity. LTE-M’s ability to operate as a full-duplex
system over a larger bandwidth also gives it an additional dimension with its
capability to offer services of lower latency and higher throughput than EC-GSM-
IoT and NB-IoT, qualities which allow LTE-M to support services such as voice
over IP. LTE-M supports both full-duplex frequency division duplex (FDD), half-
duplex FDD, and time division duplex (TDD).
Since LTE-M enables a wide variety of applications that require low data
rates and real-time communications, there are a number of potential use cases
(as shown in Figure 1.22).

Figure 1.22 The LTE-M use cases, [19].

39
1.6.3 Narrow-Band IoT (NB-IoT)

The NB-IoT is developed with simplicity in mind, to be used in cheaper

end-user equipment. It has low throughput, and delay-tolerant applications (e.g.,
sensors as IoT devices). It supports bitrates of several tens of kbps up to several
hundreds of kbit/s by using of 180 kHz of bandwidth (which has size similar to
GMS/GPRS/EDGE frequency carrier width). However, as given in Figure 1.23,
NB-IoT is created to be possible to be deployed within existing LTE bands, by
utilization of so-called spectrum between two ‘standard’ adjacent LTE frequency
carriers. Also, NB-IoT can be used in its own standalone mode, thus providing
easy migration path for the re-farmed GSM/GPRS/EDGE spectrum. So, due to its
low price and affordability NB-IoT is considered to be one of the “accelerators” for
further explosion of the IoT services.

Figure 1.23 Deployment options for NB-IoT, [22].

While EC-GSM-IoT and LTE are based on existing radio access

technologies, NB-IoT introduces to a large extent a new radio access
technology. It can operate over a system bandwidth as low as 200 kHz, in
stand-alone mode, within an LTE carrier, or, within the guard-band of an LTE
carrier (the operation modes for NB-IoT are given in Figure 1.23). It also supports
a minimum channel bandwidth of only 3.75 kHz. This gives an unmatched

40
spectrum flexibility and system capacity (number of connected NB-IoT devices)
which in combination with qualities such as energy efficient operation, ultra-low
device complexity and ubiquities coverage makes NB-IoT a very competitive
technology in the IoT market.
NB-IoT supports massive IoT capacity with only one PRB (Physical
Resource Block) in both uplink and downlink. Also, NB-IoT can support multiple
carrier operation to get more IoT capacity (depending on number of connected
devices). Based on a traffic model with a split of devices of 80%: MAR (Mobile
Autonomous Reporting) periodic and 20%: Network Command is MAR periodic,
NB-IoT with one PRB supports more than 50,000 devices per cell.
Moreover, the NB-IoT significantly improves the power consumption
of user devices, system capacity and spectrum efficiency, especially in deep
coverage. Battery life of more than 10 years can be supported for a wide range of
use cases (see Figure 1.24).

Figure 1.24 Example use cases for Mobile IoT technologies including LTE-M and
NB-IoT, [21].

There are many IoT use cases that can benefit from deeper network
coverage (Figure 1.25), especially for devices deployed in challenging locations
such as utility meters. In many use cases, trading off uplink spectral efficiency
and latency can effectively increase coverage without increasing output power
that will negatively impact the device battery life.

Figure 1.25 Better coverage with LTE-M and NB-IoT enhancements in LTE-Advanced-
Pro, [16].

41
Comparison of LTE-M and NB-IoT

Figure 1.26 gives summary of the key differences between LTE-M and
NB-IoT, while Table 1.6 provides their comparison. Overall, LTE-M is the better
alternative with respect to handling firmware and software updates that are
expected during the lifecycle of the devices. As other LTE interfaces, also LTE-M
supports mobility of the devices better than NB-IoT, and has possibility for
implementation of new cases of existing devices due to possible software
upgrades. On the other side, NB-IoT has further extended coverage compared
to LTE-M (as shown in Figure 1.25), and has lower power consumption (due to
lower transmit power as well as longer transmission cycles, which extended can
go up to 3 hours for NB-IoT, which is several times longer than LTE-M) and also
lower prices per devices due to simpler design. Hence, it may be suitable option
for mass deployments (for massive IoT i.e. massive MTC).

Figure 1.26 Key differences between LTE-M and NB-IoT, [20].

Table 1.6 Comparison of LTE-M and NB-IoT, [22].

LTE-M NB-IoT
Release 13 Release 14 Release 13 Release 14
(Cat M1) (Cat M2) (Cat NB1) (Cat NB2)

Bandwidth 1.4 MHz 5 MHz 180 kHz 180 kHz

Up to several Few 10s of

Downlink bitrate ~ 1 Mbit/s ~100 Kbit/s
Mbit/s Kbit/s
Up to several
Uplink bitrate ~ 1 Mbit/s 10s of Kbit/s ~100 Kbit/s
Mbit/s

Latency ~ 10s or 100s of milliseconds ~ seconds up to 10 sec

Full or Half Full or Half

Duplexing Mode Half Duplex Half Duplex
Duplex Duplex

Antennas 1 (SISO) 1 (SISO) 1 (SISO) 1 (SISO)

Mobility LTE mobility LTE mobility No Limited

Device Transmit
20; 23 dBm 20; 23 dBm 20; 23 dBm 14; 20; 23 dBm
Power

42
 However, one may note that both LTE-M and NB-IoT have significantly
improved indoor coverage, when compared with legacy LTE/LTE-Advanced
mobile networks. Further, LTE-M as a better alternative for moving devices is
more suitable for voice services such as Voice over LTE (VoLTE).
With LTE-M, the IoT devices can react in range of milliseconds if
required, enabling use cases where a fast response is needed which is relevant
for the usability of human-machine interactions. On the other side, NB-IoT is
designed for massive deployments.

Figure 1.27 Expected growth of Massive IoT connections (billion), [27].

Overall, in the past 5 years, according to the data from GSA, [25], and
GSMA, [26], there are more than 50 LTE-M networks and more than 100 NB-IoT
networks launched on a global scale. So, near the end of 2021, one may say that
there are more than 150 Mobile IoT (i.e., Cellular IoT, used interchangeably
throughout the text in all Modules) networks worldwide. Considering that Mobile
IoT (which includes LTE-M and NB-IoT) is based on standards (from 3GPP) and
mobile networks that are deployed in all countries worldwide, one may expect
their use in the IoT segment to continue further in this decade (2020s) and to
become dominant in IoT market segment (according to predictions from Ericsson,
shown in Figure 1.27).

43
1.7 LTE in unlicensed bands

The spectrum availability is an issue with LTE but there is significant

unlicensed 5GHz frequency available that is currently used by WiFi and other
technologies.
In that manner, there has been cross-industry engagement in standards
and regulatory bodies with aim to make reality the idea of 3GPP mobile
technologies going to the unlicensed spectrum. That was especially challenging
because two initially independent technologies should be put into the same
spectrum, which include 3GPP mobile technologies on one side (e.g., LTE, 5G
NR), and WiFi technologies which are based on completely different standards
from other SDO – that is IEEE.
The fact is that interworking between WiFi and mobile technologies
started in 2.5G and 3G era (i.e., at the beginning of 2000s, with the introduction
of packet-switching i.e. IP connectivity in mobile networks), however they were
connected mainly via the mobile core network and used different radio access
links and spectrum (mobile working in the licensed spectrum, and WiFi working in
the unlicensed spectrum), Since the middle of last decade 3GPP mobile
technologies started to enter (for the first time) into the unlicensed spectrum with
development of LTE standards.
There are several standards for LTE use in unlicensed bands, with or
without WiFi (which naturally is in the unlicensed bands). Different variants for
LTE are outlined in Figure 1.28.

Figure 1.28 Use of LTE in unlicensed bands, [1].

LTE in Unlicensed spectrum (LTE-U) is based on 3GPP release 12, it

was initially proposed by vendors and backed by several big mobile operators (in

44
2015-2016). It uses 5 GHz unlicensed band, targeted to add capacity to the LTE
by using the unlicensed band initially heavily used by WiFi networks.
Licensed Assisted Access (LAA) is also introduced with the 3GPP
release 13 in 2015, as part of LTE-Advanced-Pro. It provides aggregate
simultaneous use of LTE in licensed and unlicensed spectrum bands in the
downlink direction to provide the most efficient use of both available bands. This
is targeted to give higher bitrate in the range of Gbit/s with single LTE carrier of
20 MHz and unlicensed 5 GHz band, providing better user experience.
Enhanced LAA (eLAA) from 3GPP release 14 allows additionally
licensed and unlicensed spectrum aggregation in the uplink, in addition to the
downlink aggregation by the LAA.
LTE-WLAN Aggregation (LWA) enables aggregation of LTE in licensed
band and WiFi carrier, which is easy for implementation with legacy LTE and
WiFi equipment.
MulteFire is fully operational use of LTE in unlicensed spectrum on 5
GHz, specified by MulteFire Alliance - MFA (not by 3GPP), which is different than
LAA and LWA which aggregate LTE carriers in licensed spectrum with LTE or
WiFi carriers in unlicensed spectrum, respectively. MulteFire adds more capacity
in crowded areas, as well as provides enhanced local broadband access even
without a SIM card (similar to WiFi) and opens possibilities for Internet of Things
(IoT) verticals for LTE value added services.

1.7.1 LWA (LTE-WLAN Aggregation)

LWA offers seamless usage of both LTE and WiFi networks and
substantially increased performance.
For a mobile operator, LWA simplifies WiFi deployment, improves system
utilization and reduces network operation and management costs. LWA can be
deployed in collocated manner, where the eNB and the Wi-Fi AP are integrated
into the same physical device or in non-collocated manner, where the eNB and
the Wi-Fi AP are separate devices connected to an IP network. The latter
deployment option is particularly suitable for the case when WiFi needs to cover
large areas and/or WiFi services are provided by a 3rd party (e.g., university,
company, organization), different than the mobile operator.
As mentioned before, the LWA has been standardized by the 3GPP in
Release-13. Further development of this standard in Release 14 provided
enhanced LWA (eLWA), which adds support also for 60 GHz band (were are
standardized IEEE 802.11ad and IEEE 802.11ay from WiFi segment). With that
eLWA adds additional unlicensed bandwidth in 60 GHz spectrum (around 2 GHz)
and additional features to LWA, regarding the carrier aggregation (WiFi and LTE
carriers), mobility improvements..
From the network perspective, there are two options that provide
flexibility when looking at deploying LWA:
 Collocated: in this scenario WiFi Access Point (AP) or WiFi Access
Controller (AC) is physically integrated with the LTE base station (eNB),

45
  Non-collocated: in this case the WiFi network (i.e. APs and/or ACs) are
connected to the LTE eNB via an external network interface (that is Xw).

Figure 1.29 Architecture for LWA, [36].

LWA design primarily is based on use of LTE Dual Connectivity (DC)

architecture, which is defined in 3GPP Release 12. The feature of LTE DC allows
given UE to connect to multiple base stations simultaneously, with WiFi being (in
LWA case) used instead of LTE Secondary eNB (SeNB).
Figure 1.29 shows the architecture for LWA. In the user plane, LTE and
WiFi are aggregated at the Packet Data Convergence Protocol (PDCP) level, in
similar manner as aggregation of LTE carriers in LTE-Advanced/LTE-Advanced-
Pro. In the downlink, the eNB may schedule PDCP PDUs (Packet Data Units) of
the same bearer to be delivered to the UE either via LTE or WiFi.
In the control plane, eNB (not WiFi Access Point) is the node that is
responsible for LWA service activation, de-activation and the decision as to which
bearers (from LTE) are offloaded to the WiFi. The offload to the WiFi is based on
WiFi measurements reported by the UE to the eNB. The mobility in LWA is
limited due to limitations of mobility in WiFi networks which support only nomadic
mobility on local area network.
Overall, the LWA solution is a balance between LTE technology and WiFi
technology, where the decision on LWA activation (or QoS support) is performed
by the mobile network, however it gives sufficient level of flexibility for the UE by
given limits set by the operator.

46
1.7.2 LTE Assisted Access (LAA)

The 3GPP has started entering unlicensed spectrum bands in 2016 for the
very first time (LTE-U was not 3GPP standard) with the completion of Licensed
Assisted Access (LAA) as part of the LTE standardization in 3GPP Release 13.
Table 1.7 summarizes the evolution of 3GPP technologies regarding the
LAA (based on LTE), where LTE is operated on an unlicensed carrier along with
a carrier in licensed spectrum. The evolution of LAA in follow up 3GPP releases
are noted as enhanced LAA (eLAA) in Release 14 and further enhanced LAA
(feLAA) in Release 15, and NR-U for 5G NR (New Radio) unlicensed access.

Table 1.7 Development of LAA in 4G LTE in 3GPP Releases 13 to 15, [23].

LAA eLAA feLAA NR-U

(4G LTE) (4G LTE) (4G LTE) (5G NR)
3GPP Release Release 13 Release 14 Release 15 Release 16
Unlicensed
Yes Yes Yes Yes
downlink
Unlicensed
No Yes Yes Yes
scheduled uplink
Unlicensed
No No Yes Yes
autonomous uplink
Standalone
unlicensed No No No Yes
operation

Figure 1.30 Examples of Carrier Aggregation for LAA by using LTE carriers in licensed
band (on 1800 MHz) and unlicensed band (on 5 GHz), [15].

47
 The LAA standards include so-called Listen-Before-Talk (LBT)
regulatory, for fair sharing of unlicensed bands with WiFi (which was demanded
since initial LTE-U definition by LTE-U Forum). That is based on Carrier Sensing
Adaptive Transmission (CSAT), to guarantee Wi-Fi transmission time and a
fairly divided channel access time between two networks (LTE and WiFi) in the
unlicensed bands. CSAT defines a duty cycle, and the LTE cell transmits during
an adaptive fraction of the cycle, dictated by a long-term sensed medium activity,
and vacates the channel for the remaining duration. Although LTE-U triggered
debates with the WiFi community, it was ruled out from the standardization
framework, mainly due to inflexibility and long latency caused by duty cycling.
However, it did evolve into the standardized technologies such as License
Assisted Access (LAA).
LAA uses carrier aggregation with at least one Secondary Cell operating
in the unlicensed spectrum (examples are shown in Figure 1.30). In LAA, the
configured set of serving cells for a UE therefore always includes at least one
Secondary Cell operating in the unlicensed spectrum. Unless otherwise
specified, LAA Secondary Cells act as regular Secondary Cells in LTE radio
access network.
LAA eNB and UE apply Listen-Before-Talk (LBT) before performing a
transmission on LAA Secondary Cell. When LBT is applied, the transmitter
listens to/senses the channel to determine whether the channel is free or busy. If
the channel is determined to be free, the transmitter may perform the
transmission; otherwise, it does not perform the transmission. If an LAA eNB
uses channel access signals of other technologies for the purpose of LAA
channel access, it shall continue to meet the LAA maximum energy detection
threshold requirement.

Discussion

What type of cells is ideal for use in unlicensed bands? Well, small cells
are ideal for use in unlicensed spectrum and are typically deployed as hotspots
within macro cell coverage. They are increasingly being used to meet growing
data transfer demands. Aggregation between small and macro cells through dual
connectivity can significantly improve per-user throughput and mobility
robustness, and LTE-Advanced Pro continues the work that has already begun
before in this area.
With LTE-Advanced Pro dynamic uplink and downlink aggregation allows
operators to dynamically adjust to varying traffic needs by changing the
uplink/downlink configuration based on traffic, and/or enabling downlink traffic
offloading.
What can we expect for the use of unlicensed spectrum in the future? In
terms of usage of the unlicensed spectrum, the majority of devices will likely
continue to use WiFi (e.g., in hotspots, in homes and offices), especially when it
is used as an extension (in the last meters) of the fixed broadband access. On
the other side, unlicensed spectrum can be crowded (which results in traffic
congestions) and its capacity neither its overall QoS can be guaranteed (due to

48
that it is free and shared between technologies as well as between end users on
the same location). However, for best-effort Internet traffic it suits well, bur for
time sensitive services and critical services the use of licensed spectrum (in
mobile networks) is required and hence needed.
The availability of 3GPP mobile technologies (4G LTE and 5G NR) in
unlicensed spectrum (coexisting with WLAN and other technologies on fair
sharing policies) is expected to increase the total value that users get from the
limited radio spectrum resources. Considering that the spectrum is scarce, such
approach of fair sharing of the unlicensed bands ‘looks’ like a good one in the
bandwidth ‘hungry’ wireless and mobile world.

49
1.8 LTE-A-Pro V2X and mission critical services

The LTE-Advanced Pro brings enhancements in multiple dimensions and

very important dimension is the development of V2X (Vehicle to Everything)
and mission critical services, which further continue to develop and grow in 5G
networks. In fact, the focus on other professional services (such as V2X and
critical services) is one of the major reasons for having the label “Pro” in the
name of the last 3GPP releases for LTE-Advanced (Releases 13 and 14) before
the appearance of the 5G (which started with 3GPP Release 15).

1.8.1 Vehicular to Everything (V2X) services

LTE-Advanced Pro expands the reach of cellular mobile technology to

new areas such as V2X, which provides a substantial leap in safe driving and
becomes an essential element of autonomous driving.
Vehicle-to-Everything (V2X) LTE-Advanced Pro encompasses all
aspects of the 3GPP work needed to support vehicle-based communications:
enhancements of the air interface, protocols, and impacts on the LTE core
network. More precisely said, V2X is connecting vehicles to other vehicles,
infrastructure and everything else around them to improve safety and enable
better autonomous driving.
In general, V2X covers four specific cases:
 V2V: Vehicle-to-vehicle for the communication between two vehicles. For
example, used by collision avoidance and safety systems.
 V2I: Vehicle-to-infrastructure for communication between vehicle and
infrastructure such as traffic lights, Roadside Units (RSU), etc. For
example, used to understand traffic light conditions, priority, etc.
 V2P: Vehicle-to-pedestrian for communication between vehicles and
people, bicycles, animals, etc. For example, to provide safety alerts to
pedestrians and bicyclists, etc.
 V2N: Vehicle-to-network for communication between vehicle and cloud
services. For example, to provide routing, traffic management, etc.

The first three of these cases are network independent, meaning the
devices communicate directly with each other without involving the cellular
network. The last one is network dependent, meaning the cellular network
facilitates communication.
There are two modes of operation for V2X communication, as shown in
Figure 1.31, which include the following:

50
  V2X communication over PC5 interface: PC5 interface directly connects
UEs (User Equipments) so that over-the-air V2X message from a UE is
directly received by UEs around the transmitter.
 V2X communication over LTE-Uu interface: LTE-Uu interface connects
UEs with eNB (E-UTRAN NodeB) which plays the role of base station in
the LTE networks.

Figure 1.31 Vehicle to anything (V2X) communication in LTE-A-Pro, [29].

These two communication interfaces may be used by a UE independently

for transmission and reception. V2X communication over PC5 is supported using
sidelink when the UE is inside LTE network coverage, i.e. "served by E-UTRAN"
and when the UE is out of network coverage, i.e. "not served by E-UTRAN". For
V2X communication over LTE-Uu, which is supported only when the UE is inside
network coverage, a UE may receive V2X messages via downlink unicast or
broadcast while transmitting V2X messages via uplink.
Figure 1.32 depicts an overall architecture for V2X communication. The
"V2X Control Function", which communicates with UEs via the V3 interface (not
shown on the figure), is the logical function defined for network related actions
required for V2X and performs authorization and provisioning of necessary
parameters for V2X communication to the UE.
When the UE is "served by E-UTRAN", it can send V2X messages over
PC5 interface by using network scheduled operation mode (i.e. centralized
scheduling) and UE autonomous resources selection mode (i.e. distributed
scheduling). When the UE is "not served by E-UTRAN", it can send V2X
messages over PC5 interface only by using UE autonomous resources selection
mode.
For V2X messages over PC5, both IP based and non-IP based are
supported. For IP based V2X messages over PC5, only IPv6 is used. PPPP

51
(ProSe Per-Packet Priority) reflecting priority and latency for V2X message is
applied to schedule the transmission of V2X message over PC5.

V2X Application Server Data
Network

V2X Control Function
P‐GW
EPC
HSS BM‐SC
MME S‐GW
MBMS‐GW

E‐UTRAN E‐UTRAN E‐UTRAN

UE
UE UE

BM‐SC Broadcast Multicast Service Centre
EPC Evolved Packet Core
E‐UTRAN Evolved Universal Terrestrial Radio Access Network 
UE HSS Home Subscriber Server
MBMS Multimedia Broadcast/Multicast Service
MBMS‐GW MBMS Gateway 
MME Mobile Management Entity
P‐GW Packet Data Network Gateway
S‐GW Serving Gateway
UE User Equipment

Figure 1.32 Mobile architecture for V2X communication, [30].

For V2X communication over LTE-Uu, both IP based and non-IP based
V2X messages are supported. In order to transmit non-IP based V2X messages
over LTE-Uu, the UE encapsulates the V2X messages in IP packets.
For latency improvements for MBMS, localized MBMS can be considered
for localized routing of V2X messages destined to UEs.
V2X applications have certain requirements regarding the QoS (Quality of
Service) and security.
Regarding the QoS, for V2X communication over LTE-Uu interface, the
V2X messages can be delivered via Non-GBR bearer (i.e. an IP transmission
path with no reserved bitrate resources) as well as GBR bearer (i.e. an IP
transmission path with reserved (guaranteed) bitrate resources).
In order to meet the latency requirement for V2X message delivery, the
following standardized QCI (QoS Class Identifier) values can be used:
 QCI=3 (GBR bearer) and QCI=79 (Non-GBR bearer) can be used for the
unicast delivery of V2X messages.
 QCI=75 (GBR bearer) is only used for the delivery of V2X messages over
MBMS bearers.

52
 On the other side, regarding the security for V2X application data, the
security requirements are all satisfied by employing application-layer security
as defined in other SDOs (e.g. IEEE or ETSI ITS), not by 3GPP itself. Also, no
additional privacy features beyond the regular LTE privacy features are
supported for LTE-Uu based V2X communications, while the privacy
requirements will likely depend on regional regulatory requirements and/or
operator policy. For the privacy of PC5 transmissions, the UE changes and
randomizes the source Layer-2 ID, and the source IP address periodically or
when indicated by the V2X application that the application layer identifier has
changed.
Table 1.8 V2X inter-band multi-carrier configurations, [30]

E-UTRA 1.4 3 5 10 15 20 Maximum

Interface
Bands MHz MHz MHz MHz MHz MHz bandwidth [MHz]

1.8 GHz Uu Yes Yes Yes Yes Yes Yes

40
5.9 GHz PC5 Yes Yes

2.6 GHz Uu Yes Yes Yes Yes

40
5.9 GHz PC5 Yes Yes

900 MHz Uu Yes Yes Yes Yes

30
5.9 GHz PC5 Yes Yes

1.9 GHz Uu Yes Yes Yes Yes

40
5.9 GHz PC5 Yes Yes

2.5 GHz Uu Yes Yes Yes Yes

40
5.9 GHz PC5 Yes Yes

Table 1.9 V2X intra-band multi-carrier configurations, [30].

Maximum
V2X multi-carrier Channel BW of Channel BW of
aggregated
Configuration carrier 1 carrier 2
BW [MHz]
5.9 GHz 10 10 20

Regarding the LTE Radio Access Network (i.e., E-UTRAN), LTE-

Advanced Pro extends the RAN operations by support of simultaneous V2X
operations over multiple carriers. In one scenario, a UE can operate LTE-Uu
based V2X in a carrier while operating PC5 based V2X in another carrier. In

53
another scenario, a UE can operate PC5 based V2X simultaneously over two
carriers. Table 1.8 and Table 1.9 list the combinations of V2X carriers supported
for V2X services.
Overall, service requirements for typical V2X application include the
following:
 Message transfer latency no longer than 100 ms with 20 ms maximum
allowed latency in some specific use cases;
 Transfer of message size up to 1200 bytes;
 Support of up to 10 message transfer per second in typical cases while
enabling maximum message transfer frequency of 50 Hz;
 Communication range sufficient to give driver enough response time (e.g.
4 seconds);
 Support of relative vehicle speed up to 500 km/h;
 Support of V2X communication both in and out of network coverage.

Further development for V2X services continues in 5G standards from

3GPP, which includes work in releases after Release-14, such as 3GPP Rel-15
and Rel-16 which provide the main 5G standards of the present day, [31]. The
expected development of V2X for both LTE and 5G is shown in Figure 1.33,
which is based on 5GAA predictions.

Figure 1.33 Timeline for development of V2X services (5GAA view), [32].

The latest work on V2X introduces the requirements related to vehicle

QoS support, which enables a V2X application to be timely notified of expected
or estimated change of QoS. For example, when the communication packet error
is expected to increase or decrease, the V2X application such as platooning
application can increase or decrease inter-vehicle distance. For that purpose
there is required standardized interface towards external V2X application
servers. Using this standardized interface, the V2X application servers can
request specific QoS requirements (and, optionally, related alternative QoS
requirements). It can also be notified when the requested QoS requirements
(and/or the alternative QoS requirements) can or cannot be met.

54
 Notification on QoS Sustainability Analytics to the V2X Application
Server is specified so that the V2X Application Server may request notifications
on QoS Sustainability Analytics for an indicated geographic area and time
interval in order to adjust the application behavior in advance with potential QoS
change.
The focus of V2X in Release 14 is automotive safety, whereas the focus
of 5G NR (which will be covered in next modules in this course), specifically in
Release 16, is more autonomous driving. However, considering that they are
complementary, all investment that industry is making today in Rel-14 based V2X
is protected in the future and used even when upgraded to Rel-16.
Overall, the implementation of V2X services requires support by all parties
including vehicle vendors (e.g., 5GAA – 5G Automotive Association), vehicle
users (e.g., V2X can be installed in the cars regardless of their age or type),
telecom operators (e.g., for V2N services), as well as regulators and
administrations (which should set rules and policies regarding the security and
privacy for V2X services).

1.8.2 Mission Critical services

Mission Critical (MC) communications and MC Services has been a key

priority of 3GPP in LTE-Advanced Pro (3GPP Releases 13 and 14), which
created the foundation for this type of services that further evolves in 5G
(Releases 15 to 18 so far).
The mission critical related functionalities have been phased across them,
each release encompassing a complete set of standards for equipment vendors
and operators to phase their implementations and deployments, based on market
demands.
Prior to Rel-13, 3GPP standardized functionality that was later to serve as
an enabler for MC Services. For example, MC Services benefit from the use of
multicast bearers in LTE due to the standardization of eMBMS (enhanced
Multimedia Broadcast/Multicast) and Group Communication System Enablers
(GCSE). Additionally, Device-to-Device (D2D) Proximity Based Services (ProSe)
was enhanced to support public safety use.
This allows public safety operators to determine whether critical
communication occur on-network using the LTE network infrastructure, or off-
network without the use of the LTE network infrastructure, or both.
MCPTT (Mission Critical Push To Talk) was the first major step in a
series of MC Services and functionalities demanded by the market. In Rel-14,
completed in 2017, 3GPP added additional MC Services and enhancements to
its repertoire of standardized applications, which also include (besides MCPTT)
to other services - MCData (Mission Critical Data) and MCVideo (Mission Critical
Video), as well as general framework for additional MC Services.
Each MC service supports several types of communications amongst the
users (e.g. group call, private call). There are several common functions and
entities (e.g. group, configuration, identity) which are used by the MC services.

55
 For an MC system, the trust domain consists of one or more MC service
functions that are administered by the same or different service providers (e.g.
MC service provider, mobile operator) that have an agreement to share sensitive
information.
Point to multipoint broadcast offered by the LTE MBMS technology is
well suited to group communications, which form a major part of the public safety
related communications. The MC service on-network architecture is shown in
Figure 1.34.

Figure 1.34 MC service on-network architecture showing MBMS, [42].

The MC service server is shown being bundled with the GCS AS (Group
Communication Service Application Server) within the same network entity. It is
illustrated this way for simplicity of the diagram shown in Figure 1.34. MC service
media content is transmitted via LTE bearers, which are communication pipes
with one end in the MC service server and the other end in the MC service UE.
The uplink bearers are always allocated as unicast, but the downlink
bearers can be allocated as unicast or as MBMS bearers, or both.
An MBMS bearer (both network and radio part) is uniquely identified via a
TMGI (Temporary Mobile Group Identity) or via a combination of a TMGI and a
flow identifier. The MC service server is capable, via the MB2 interface, to
request the creation of MBMS bearers and associate a unique TMGI or a
combination of a TMGI and a flow identifier.
The MC service server may determine the MBMS broadcast area based
on the cell identities of the affiliated group members received over GC1. The MC
service server may determine for a user the switching from MBMS bearer to
unicast bearer based on the information reported over GC1.

Bearer management for MC services

The MC service UE use certain APNs in LTE mobile network, which can
be an MC services APN for the SIP-1 reference point (SIP stands for Session
Initiation Protocol, the main signaling protocol in IMS), or HTTP-1 reference point;
or CSC-1 reference point (CSC stands for Call Session Control).

56
 The value of each of these APNs may be the same or may differ, as well
as it may be the same as other non-MC services that have compatible QoS and
PDN (Packet Data Network). Also, the APN shall be made available to the UE
either via UE (pre)configuration or via initial UE configuration on per Home PLMN
(HPLMN) and optionally also a per Visited PLMN (VPLMN) basis.
The PDN connection to the APNs can be of type "IPv4", "IPv6" or
"IPv4v6". If a PDN connection to an APN is of type "IPv4v6" then the MC service
client uses configuration data to determine whether to use IPv4 or IPv6.
When operating in systems that support MBMS functionality, the MC
service can provide downlink MBMS delivery of MC service media. In such
cases, MC service UEs can receive the traffic delivered via MBMS, regardless of
whether or not they have any unicast radio bearers available.
If the PDN connection established during the initial attach by the MC
service UE is to an APN other than the MC services APN, then prior to user
authentication, the MC service UE shall establish another PDN connection to the
MC services APN. PDN connection establishment can also be caused by a SIP
registration request for one or more MC services.
The QCI value of 69 shall be used for the EPS bearer that transports SIP-
1 reference point messaging. The QCI value 8 or better shall be used for the
EPS bearer that transports HTTP-1 reference point messaging.

Business relationships for MC services

The possible business relationships for MC (Mission Critical) services are

shown in Figure 1.35. The main players in the MC services business model are
mobile operators and MC service providers on one side, and MC users and MC
organizations on the other.

Figure 1.35 MC service business relationships, [42].

57
 The MC service user belongs to a single mission critical organization
based on a MC service user agreement between the MC service user and the
mission critical organization. The MC service user can have agreement and MC
service arrangement directly with a single MC service provider.
The mission critical organization and the MC service provider can be part
of the same organization, in which case the business relationship between the
two is internal to a single organization.
Single mission critical organization can have MC service arrangements
with several MC service providers. In this case, a MC service user of a mission
critical organization is always served by only one MC service provider.
However, single MC service provider can have MC service arrangements
with multiple mission critical organizations, as well as with multiple MC service
users. Also, the MC service provider and the home PLMN operator can be part of
the same organization, and in such cases the business relationship between the
two is internal to the given organization (e.g., telecom operator which provides
also mission critical services).
The Home PLMN (HPLMN) operator can have roaming agreements with
multiple Visited PLMN (VPLMN) operators and vice versa (VPLMN operator can
have roaming agreements with multiple HPLMN operators).
Overall, Mission Critical business model in LTE-Advanced-Pro mobile
networks, which further evolves in 5G networks, offers possibilities mobile
operators to provide MC services to various organizations in a given country.

Discussion on Mission Critical services

Mission critical services in this section include push-to-talk (MCPTT),

video communication (MCVideo), messaging and mobile broadband data
services (MCData), whose main capabilities and functions are standardized by
3GPP. So, as already noted, MC service is a generic name for any one of the
three mission critical services, either MCPTT, MCVideo, or MCData. They use
combination of unicast and MBMS communication in 3GPP mobile networks. The
services are targeted to user groups (e.g., in MC organization), however one
should note that managed group communication requires efficient administration
of who may connect to whom, in what context, and with what priority.
Mission Critical applications enable mission critical organizations, such as
public safety, utilities, ports, mines, airports, and other verticals to evolve
their communication capabilities from today's Land Mobile Radio (LMR) based
push-to-talk to advanced mobile broadband services and their MC services. That
way there is no need for building parallel infrastructures for MC services,
because they can be provided via the 4G or 5G mobile networks by using the
given standards and possibilities for business relationship among MC players.
Looking into the future, mission-critical service innovations will be largely
driven by applications and devices addressing existing and new use cases.
Taking such innovation to a mass market requires benefits of scale, which in turn
requires standardization.

58
1.9 Business and regulatory aspects of mobile broadband

For much of the world, return to pre-pandemic normalcy has yet to

materialize. Various measures to support governments and society to function as
normal have been explored, and many have turned to Internet connectivity to
make these measures effective. As the world plans a transition out of the
pandemic, Internet connectivity continues to play a pivotal role for countries,
communities and individuals. For example, remote work remains the norm where
possible.
In terms of connectivity infrastructure, governments at national and sub-
national levels exhibit renewed and urgent interest in addressing disparities of
access, adoption and affordability. Network operators and service providers
are investing heavily not only to expand networks but also to increase capacity
and respond to the acceleration in broadband demand. And regulators, not just
in telecommunications but in other sectors, increasingly recognize the importance
of digital technology across all facets of economic activity.

1.9.1 Global development of mobile broadband and digital divide

The massive spike in the use and importance of Internet connectivity,

triggered by the COVID-19 crisis, stands in sharp contrast to the slowing growth
in Internet users globally, and the persistent, pernicious digital divide. While
Internet connectivity has proved to be a crucial tool in responding to the impact of
the COVID-19 pandemic and in coping with massive economic, social shocks,
countries that were most vulnerable (because of more limited resource bases)
are also those with the poorest Internet adoption levels. In 2019, the year before
the start of the crisis, close to 87 per cent of individuals in developed countries
were online using the Internet, while in stark contrast only 19 per cent of
individuals in the Least Developed Countries (LDCs) were online.
Progress is being made in closing gaps and expanding the impact of
broadband for all people on the planet.
While some form of network coverage reaches nearly every single human
on the planet either via terrestrial or satellite networks, further effort is needed to
densify network infrastructure and close the usage gap by increasing capacity
and improving affordability.
To respond to massive demand for broadband connectivity in 2020 and
the shift to online activity as a result of the COVID-19 crisis, extraordinary
measures have been taken by governments, ministries, regulators, service
providers, and many other participants in the global connectivity ecosystem.

59
 The successful expansion of Internet connectivity over the past three
decades – led by the private sector and guided by policy and regulatory
frameworks that encourage commercial developments – needs to continue and
be accelerated. However, of the 3.7 billion people who remain unconnected,
85% are covered by a mobile broadband network.

Figure 1.36 Mobile broadband subscriptions per 100 inhabitants, [4].

The total number of mobile-cellular telephone subscriptions declined for

the first time in history in the period 2019-2020 (Figure 1.36). However, further
research is needed to understand whether this is caused by the disruptions
related to the COVID-19 pandemic, or whether this can be explained by other
socio-economic forces. For example, in the middle of 2020 there were an
estimated 105 mobile-cellular subscriptions per 100 inhabitants, down from 108
in 2019. This decline was driven by developing countries, where the number of
subscriptions went down from 103 in 2019 to 99 in the middle of 2020. On the
other side, in developed countries the trend was still upwards.
The number of active mobile-broadband subscriptions stood at 75 per
100 inhabitants in 2020. After substantial growth in previous years, this was only
1.1 per cent higher than in 2019.
But, one should note that the gaps between the regions are largest for
fixed-broadband subscriptions than for mobile-broadband subscriptions. So,
mobile broadband remains the main driver for the broadband access on a
global scale. However, the mobile broadband access needs fixed fiber
infrastructure which should support many Tbit/s traffic from/to end users in a
given country or region.

60
The Digital Divide

While overall and individual data consumption increased during the

pandemic, it remains to be seen what the net effect on total users will be; while
the crisis has brought many new users online, others have had to limit or stop
their access because of the economic impact to income and affordability that
most affects populations that use the Internet the least. As the world shifted to
online engagement in the second quarter of 2020, Internet traffic growth
significantly outpaced pre-COVID-19 era forecasts: average traffic in 2020 grew
by 48 per cent while global peak traffic grew by 47 per cent, compared to
forecasts of average annual growth between 2016 and 2020 of 30 per cent.
Globally, last year (2020) consumers’ use of fixed broadband increased by an
average of two and a half hours per day, and on mobile by one hour.
In the US and Europe in particular, traffic on broadband networks
increased 51 per cent because of the COVID-19 pandemic, and average per-
subscriber (or household) usage increased from 344 GB per subscriber in Q4 of
2019 to 482.6 GB per month in Q4 of 2020, an increase of 40 per cent.
In developing economies, however, the absence of broadband remains a
key issue in many countries where investment in broadband infrastructure
remains uneconomic under current financing models. In addition, where
broadband coverage exists, both broadband pricing and the cost of 4G mobile
devices before the COVID-19 were already at prohibitive levels for significant
segments of lower income populations.
The high cost of 4G devices remains another major barrier to
connectivity in lower-income countries. Despite the cost of an entry level Internet-
enabled device falling over years, nearly 2.5 billion people live in countries where
the cost of the cheapest available smartphone is a quarter or more of the
average monthly income.

1.9.2 From network expansion to network densification

Global networks (terrestrial, space and undersea) across various

technologies and dimensions combine to reach and cover every part of the world;
the challenge now is ensuring sufficient capacity, competition, and affordability
while continuing to attract sustainable investments into networks, services,
technologies and capacities. According to mobile network carriers, only 7% of the
world’s total population resides in geographic areas where they cannot provide
mobile Internet connectivity (at least with 3G data service). These 570 million
people outside of at least 3G cellular coverage may still have some connectivity
via 2G for voice and basic text message functionality, though 2G and 3G services
are now starting to be shut down in order to re-farm spectrum for 5G, and in
some cases for 4G in emerging markets.
These transitions will have to be managed carefully as significant numbers
of mobile users around the world, in developing as well as developed countries,
continue to use 2G and 3G devices for a wide range of reasons, including

61
affordability, limited digital skills and familiarity, as well as limits to service
options.
Technology neutral spectrum regulations are crucial to mobile
operators to smoothly transition users to new generation networks, increasing
network capacity but avoiding shutting down old networks that may affect those
who cannot afford new devices.
While 85 per cent of the world’s population is already covered by 4G
networks, nearly half of them are still offline, in part because of the relatively high
price of Internet access. Building broadband connectivity on top of existing
mobile broadband infrastructure is a fast and cost-efficient option to bridge the
digital divide.
A common challenge for many individuals, particularly on lower speed
connections in developing countries or via mobile devices, is the difficulty in
working or studying remotely. Even in more advanced markets, or even for
subscribers with high-capacity download connections, significant asymmetries
in network design cause subscribers with poor upload capacity to also face
difficulties in engaging with cloud-based platforms for different services. The
significant shift to remote-based activities has shined a spotlight on the
importance of upload capacity, not just download, particularly as this segment
of network infrastructure has shown some of the largest growth in the past year,
increasing in some markets, for example, by 63 per cent. This recognition of the
importance of upload and other QoS issues highlighted by the pandemic is
leading to new consumer protection legislation, calls for symmetrical
broadband consumer packages, and a focus on minimum advertised speeds.

1.9.3 Mobile broadband impact on economy through the pandemic

In the context of an increase in mobile broadband penetration, a

generalized decline in service prices, and a decrease in GDP per capita, the
2010–2020 mobile broadband model used in the 2021 study shows that mobile
infrastructure continues to have a stable impact on the world economy.
Contrary to fixed broadband, mobile broadband service pricing is not
statistically significant in driving adoption. The structural model was also run with
2010–2020 data to test whether the economic impact of mobile broadband is
affected by a “returns to scale” or a “diminishing returns” effect, [48]. The analysis
of impact of broadband (including fixed and mobile) is shown in Figure 1.37, by
splitting the countries in through groups: 1) countries with GDP per capita higher
than USD 22 000, 2) countries with GDP per capita between USD 12 000 and
USD 22 000, and 3) countries with GDP per capita lower than USD 12 000.
In summary, the following conclusions on economic impact of broadband
can be drawn:
 The changes in the broadband ecosystem in many countries due to the
pandemic (e.g., capital spending, increase in broadband penetration,
decline in pricing, especially in mobile, etc.) have not substantially altered
the impact of broadband on the economy.

62
  The contribution of fixed broadband continues to be greater in developed
countries than in developing countries, reflecting the “returns to scale”
effect.
 The economic dividend of mobile broadband continues to be greater in
countries with lower levels of economic development, diminishing in
countries and regions with higher penetration and development.
 There is higher economic contribution of mobile broadband in low-
income developed countries and of fixed broadband in middle-income
developed countries.

Figure 1.37 2020 versus 2021 studies – GDP growth impact of a 10 per cent
increase in broadband penetration (in per cent), [48].

1.9.4 Six golden regulatory rules for take-up of mobile broadband

Due to the ubiquity of digital infrastructure across business,

government and at the individual level, approaches to policy on broadband are
increasingly incorporating other sectoral issues, and vice-versa. The increasing
reliance on digital infrastructure, services, and applications in nearly every
sector and facet of economies and societies is made clear by the overlaps and
engagements in policy issues between sectors and the subsequent government
departments and ministries responsible for the sets of issues in those domains.
Organizations around the world are developing their own internal digital
transformation strategies as well as incorporating digital infrastructure advisory
into their project work.

63
 In that manner, as telecommunications regulators move towards a
notional next generation (or ’G5‘ as it is being called) of sector regulation,
transitioning from “integrated regulation, led by socio-economic policy” to
“collaborative regulation with metrics-backed decision making” (Figure 1.38
shows five generations of regulations), examples of collaboration between ICT
regulators and non-ICT regulators are emerging.

Figure 1.38 Generations of regulation: G1 to G5, [44].

The case of mobile broadband demonstrates the crucial role played by

good regulatory frameworks.
The ITU analysis demonstrates that just six regulations, or so-called ‘six
golden rules’, will accelerate the take-up of mobile broadband, removing
roadblocks and incentivizing market players. The rules range includes mandating
the sharing of infrastructure, enabling number portability and gearing up markets
for full competition. The Golden Six may be noted as a regulatory recipe for
successful mobile broadband adoption, they include the following:
1. General authorization regime;
2. Infrastructure sharing mandated;
3. Full competition in broadband access;
4. Full competition in international gateways;
5. Legal concept of dominance or SMP;
6. Foreign participation/ownership in Internet Service Providers (ISPs).

The mobile broadband technologies have put additional pressure on

regulatory frameworks in the ICT sector and have pushed the gold standard for
regulation beyond and above past requirements. As a result, G3 regulation is
maintaining world average penetration while G4 regulation is now the norm for
vibrant mobile broadband markets. Average mobile broadband penetration in G4
peers is around 20 percentage points higher than countries in lower generations.
The new G5 however, outperforms all others substantially, reaching a level of
near universal penetration. The take-up of mobile broadband is testament to
the power of regulation and how movement up the regulatory ladder leverages
new technologies to meet market demand in each country on the global scale.

64
Abbreviations
4G Fourth Generation mobile systems
5G Fifth Generation mobile systems
AC Access Controller
AP Access Point
APN Access Point Name
BM-SC Broadcast Multicast Service Centre
CA Carrier Aggregation
CSC Call Session Control
CSCF Call Session Control Function
EC-GSM-IoT Extended Coverage GSM for Internet of Things
eLAA enhanced LAA
EPC Evolved Packet Core
EPS Evolved Packet System
FMC Fixed Mobile Convergence
GCS AS Group Communication Service Application Server
HeNB eNodeB
HNB Home Node B
HSPA High Speed Packet Access
HSS Home Subscriber Server
HTTP Hypertext Transfer Protocol
ICANN Internet Corporation for Assigned Names and Numbers
ICT Information and Communication Technology
IETF Internet Engineering Task Force
IP Internet Protocol
IPv4 IP version 4
IPv6 IP version 6
ITU International Telecommunication Union
LAA Licensed Assisted Access
LAN Local Area Network
LoRa Long Range Wide Area
LPWAN Low Power Wide Area Networks
LTE Long Term Evolution
LTE-M Long Term Evolution for Machines
LTE-U LTE in Unlicensed band
LWA LTE-WLAN Aggregation
MBMS Multimedia Broadcast and Multicast Service
MBSFN Multimedia Broadcast multicast service Single Frequency Network
MC Mission Critical
MCPTT AS MCPTT Application Server
NB-IoT Narrow Band Internet of Things
NR New Radio
NR-U New Radio Unlicensed
PCRF Policy and Charging Rules Function

65
PDN Packet Data Network
QCI QoS Class Identifier
QoS Quality of Service
PLMN Public Land Mobile Network
RAN Radio Access Network
RATG Radio Access Technology Group
RFC Request for Comment
SAE System Architecture Evolution
TCP Transmission Control Protocol
TMGI Temporary Mobile Group Identity
UE User Equipment
UNI User Network Interface
URI Uniform Resource Identifier
V2I Vehicle-to-Infrastructure
V2N Vehicle-to-Network
V2P Vehicle-to-Pedestrian
V2V Vehicle-to-Vehicle
VoIP Voice over IP
VoD Video on Demand
WiFi Wireless Fidelity
WLAN Wireless Local Area Network

66
References

[1] Toni Janevski, "QoS for Fixed and Mobile Ultra-Broadband", John Wiley & Sons,
USA, April 2019.
[2] Toni Janevski, "Internet Technologies for Fixed and Mobile Networks", Artech House,
USA, November 2015.
[3] Toni Janevski, "NGN Architectures, Protocols and Services", John Wiley & Sons, UK,
April 2014.
[4] ITU, “Measuring Digital Development - Facts and Figures”, 2020.
[5] ITU, “Setting the Scene for 5G: Opportunities & Challenges”, 2018.
[6] RFC 791, "Internet Protocol", IETF, September 1981.
[7] ITU Rec. Y.3101, “Requirements of the IMT-2020 network”, January 2018.
[8] ITU Rec. Y.3130, “Requirements of IMT-2020 fixed mobile, January 2018.
[9] ITU-R Report M.2134, “Requirements related to technical performance for IMT-
Advanced radio interface(s)”, 2008.
[10] Report ITU-R M.2410-0, “Minimum requirements related to technical performance
for IMT-2020 radio interface(s)”, November 2017.
[11] Telit, “The Role of 4G LTE-Advanced Pro in the Evolution Toward 5G”,
https://www.telit.com/, last accessed in October 2021.
[12] 3G Partnership Project (3GPP), www.3gpp.org, last accessed in November 2021.
[13] 3GPP TS 36.306, “Evolved Universal Terrestrial Radio Access (E-UTRA); User
Equipment (UE) radio access capabilities (Release 14)”, March 2018.
[14] Qualcomm, “Accelerating the mobile ecosystem expansion in the 5G Era with LTE
Advanced Pro” May 2018.
[15] Nokia, " LTE Path from 4G to 5G - How to use unlicensed spectrum", 2017.
[16] Qualcomm, “Leading the LTE IoT evolution to connect the massive Internet of
Things”, 2016.
[17] ITU Internet Reports, “The Internet of Things”, 2005.
[18] ITU-T Rec. Y.2069, "Terms and definitions for the Internet of things", July 2012.
[19] Avsystem, “What is LTE-M?”, https://www.avsystem.com/, last accessed in
October 2021.
[20] Telenor Connexion, “LTE-M vs NB-IoT – a guide exploring the differences between
LTE-M and NB-IoT”, https://www.telenorconnexion.com/, last accessed in October 2021.
[21] AT&T, “What you need to know about IoT wide area networks”, ITU event, last
accessed in October 2021.
[22] S. Tabbane, “IoT Standards – 3GPP Standards”, ITU event, September 2018.
[23] Ericsson, “3GPP technologies in unlicensed spectrum: A contributor to the common
good”, September 2020.

67
[24] MFA, https://www.mfa-tech.org/, last accessed in October 2021.
[25] GSA, “Global NB-IoT/LTE-M networks”, March 2019.
[26] GSMA, https://www.gsma.com/iot/, accessed in October 2021.
[27] Ericsson, “Cellular networks for Massive IoT”, January 2020.
[28] ITU-T Series Q, supplement 71, “Testing methodologies of Internet related
performance measurements including e2e bit rate within the fixed and mobile operators'
networks”, October 2019.
[29] ITU-R M.2445-0, "Intelligent transport systems (ITS) usage", November 2018.
[30] 3GPP TR 21.914 V14.0.0, “Release 14 Description; Summary of Rel-14 Work Items
(Release 14)”, May 2018.
[31] 3GPP TR 21.916 V16.0.1, “Release 16 Description; Summary of Rel-16 Work Items
(Release 16)”, September 2021.
[32] 5GAA, “5G Automotive Association Status update”, September 2020.
[33] ITU-T Rec. E.806, “Measurement campaigns, monitoring systems and sampling
methodologies to monitor the quality of service in mobile networks”. June 2019.
[34] ITU-R M.2083-0, "IMT Vision – Framework and overall objectives of the future
development of IMT for 2020 and beyond", September 2015.
[35] ITU, World Radiocommunication Conference 2015 (WRC-15): Agenda and Relevant
Resolutions, 2015.
[36] Intel, “Alternative LTE Solutions in Unlicensed Spectrum: Overview of LWA, LTE
LAA, and Beyond”, 2016.
[37] Ofinno, “3GPP: The Unlicensed Journey”, November 2020.
[38] ITU-T Rec. E.805.1, "Quality of service operational strategy for improved regulatory
supervision of providers of mobile telecommunication services", January 2021.
[39] ITU, “QoS Regulation Manual”, 2017.
[40] Ericsson, "Mission Critical Push-to-Talk - Solution overview release 3", 2021.
[41] 3GPP TS 22.280 V18.0.0, "Mission Critical Services Common Requirements
(MCCoRe); Stage 1 (Release 18)", September 2021.
[42] 3GPP TS 23.280 V17.8.0, "Common functional architecture to support mission
critical services; Stage 2 (Release 17)”, September 2021.
[43] GSMA, “The Mobile Economy 2021”, 2021.
[44] ITU, "Global ICT Regulatory Outlook 2020", 2020.
[45] Ericsson, “Ericsson Mobility Report”, June 2021.
[46] ITU Broadband Commission, "The State of Broadband: People-Centered
Approaches for Universal Broadband", September 2021.
[47] ITU, “ICT Price Trends”, 2020.
[48] ITU, “The economic impact of broadband and digitization through the COVID-19
pandemic”, 2021.

68