Road to 5G:
Introduction
and Migration
April 2018
ROAD TO 5G: INTRODUCTION AND MIGRATION
About the GSMA
Future Networks Programme
The GSMA represents the interests of mobile operators
worldwide, uniting nearly 800 operators with almost 300
companies in the broader mobile ecosystem, including
handset and device makers, software companies,
equipment providers and internet companies, as well as
organisations in adjacent industry sectors. The GSMA also
produces industry-leading events such as Mobile World
Congress, Mobile World Congress Shanghai, Mobile World
Congress Americas and the Mobile 360 Series of
conferences.
The GSMA’s Future Networks is designed to help operators
and the wider mobile industry to deliver All-IP networks so
that everyone benefits regardless of where their starting
point might be on the journey.
For more information, please visit the GSMA corporate
website at www.gsma.com. Follow the GSMA on Twitter:
@GSMA.
The programme has three key work-streams focused on:
The development and deployment of IP services, The
evolution of the 4G networks in widespread use today, The
5G Journey developing the next generation of mobile
technologies and service.
For more information, please visit the Future Networks
website at: www.gsma.com/futurenetworks
Document Editor
Dongwook Kim, 5G Project Manager
Michele Zarri, Technical Director – Networks
Acknowledgement (5G Introduction Project Members)
Applied Communication Sciences
ARM Ltd.
AT&T Mobility
Axiata Group Berhad
China Mobile Limited
China Telecommunications Corporation
China Unicom
Deutsche Telekom AG
DISH Network Corporation
Ericsson
Guangdong OPPO Mobile Telecommunications Corp.,Ltd.
Huawei Technologies Co Ltd
iBasis, Inc
Intel Corporation
KDDI Corporation
KPN B.V.
KT Corporation
Kuwait Telecom Company (K.S.C.)
LG Electronics Inc
LG Uplus, Corp.
MediaTek Inc.
Nokia
NTT DOCOMO, Inc.
Orange
Qualcomm Incorporated
SingTel Mobile Singapore Pte. Ltd.
SK Telecom Co., Ltd.
SoftBank Corp.
Sprint Corporation
Syniverse Technologies, Inc
Telecom Italia SpA
Telefónica S.A.
Telenor Group
Telia Finland Oyj
T-Mobile USA, Inc
Turkcell Iletisim Hizmetleri A.S.
United States Cellular Corporation
Videotron Ltd
Vodafone Roaming Services S.à r.l
ZTE Corporation
ROAD TO 5G: INTRODUCTION AND MIGRATION
Contents
Executive Summary
1
Introduction
5
8
4.3.4
Impact on device and
network
Impact on voice including
service continuity
NSA Option #3 to NSA
Option #7 and SA Option
#5
Description
Feasibility of the path in
meeting 5G use cases
22
Deployment considerations
Impact on device and
network
Impact on voice including
service continuity
NSA Option #3 to NSA
Option #3 and SA Option
#2
Description
Feasibility of the path in
meeting 5G use cases
Deployment considerations
23
23
Impact on device and
network
Impact on voice including
service continuity
NSA Option #3 to NSA
Option #4 and SA Option
#2
Description
24
Feasibility of the path in
meeting 5G use cases
Deployment considerations
28
Impact on device and
network
Impact on voice including
service continuity
27
4.7
Other migration steps
27
5
Recommendations for
collaborative actions
28
1.1
1.2
1.3
Scope
Abbreviations
References
9
9
11
4.3.5
2
2.1
2.2
2.3
Why 5G?
The advent of the 5G era
Drivers of 5G
Expectations on 5G
12
13
13
13
4.4.1
4.4.2
2.4
Use cases and spectrum
bands for 5G
14
4.4
4.4.3
4.4.4
4.4.5
Part I
3
3.1
3.2
5G Network Deployment
Options
Standalone (SA) and Nonstandalone (NSA)
Evolved Packet Core (EPC)
and 5G Core Network (5GC)
15
15
16
16
4.5
4.5.1
4.5.2
4.5.3
4
5G Network Introduction &
Migration Paths
18
4.5.4
4.1
General observations and
assumptions
EPS to SA Option #2
Description
Feasibility of the path in
meeting 5G use cases
19
4.5.5
19
19
19
4.6
Deployment considerations
Impact on device and
network
19
20
Impact on voice including
service continuity
EPS to NSA Option #3
Description
Feasibility of the path in
meeting 5G use cases
Deployment considerations
20
4.2
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5
4.3
4.3.1
4.3.2
4.3.3
21
21
21
21
4.6.1
4.6.2
4.6.3
4.6.4
4.6.5
22
23
23
23
24
24
24
24
24
25
26
26
26
27
ROAD TO 5G: INTRODUCTION AND MIGRATION
Part II
6
6.1
6.1.1
6.1.2
6.1.3
6.2
6.3
6.4
6.5
Part III
7
7.1
7.1.1
Study of 5G Network
Introduction Case
Why 5G?
5G technology and service
trial activities
Other business/social
drivers in the nation
Deployment purpose and
spectrum at disposal
Analysis of 5G migration
options
Rationale behind migration
option
Challenges in actual
migration
Lessons learned from
migration case study
Core network
considerations
Core network solutions
EPC (Evolved Packet Core)
30
30
31
31
7.1.2
7.2
5GC (5G Core)
Comparison of EPC and
5GC
42
44
8
Detailed considerations on
5G deployment options
Standalone considerations
45
8.2
Non-standalone
considerations
46
9
Impact on voice including
service continuity
Options for operator
voice/video
communications service in
5G
49
Recommendations for
voice/video
communications service in
5G
51
8.1
32
46
33
34
9.1
35
36
9.2
38
40
40
41
41
Annex A
Other considerations for operators
A.1
Status Icon related to 5G
50
52
53
ROAD TO 5G: INTRODUCTION AND MIGRATION
Executive Summary
Unlike previous generations of mobile networks,
the fifth generation (5G) technology is expected
to fundamentally transform the role that
telecommunications technology plays in the
society. 5G is also expected to enable further
economic growth and pervasive digitalisation of a
hyper-connected society, where not only are all
people connected to the network whenever
needed, but also many other devices/things
virtually creating the society with everything
connected (i.e. Internet of Everything). 5G will
therefore enable new use cases such as smart
cities, smart agriculture, logistics and public safety
agencies. Furthermore, there are a variety of
spectrum bands available for deployment of 5G,
which can be sub-divided in three macro
categories: sub-1GHz, 1-6GHz and above 6GHz.
In addition, 3GPP (3rd Generation Partnership
Project) has specified new 5G radio access
technology, 5G enhancements of 4G (fourth
generation) network, and new 5G core network.
The variety of requirements and spectrum needs
show that there are many options of 5G
introduction and different spectrum bands will be
needed to support all use cases. Operators must
therefore consider the feasibility of different
options in meeting their intended initial use cases
and interoperability of their choice with other
options to ensure their networks deliver the use
cases effectively while supporting global
interoperability.
This document therefore analyses the alternatives
available for operators intending to introduce a
3GPP compliant 5G system migrating from their
3GPP 4G network. The structure of the document
is as follows: Part I provides an high level
description of the 5G introduction and subsequent
migration along recommendations for
collaborative actions while Part II delves into real
operator case study and Part III provides
advanced technology considerations requiring a
reasonably advanced level of understanding of the
3GPP system.
While 3GPP is defining both a new 5G core
network (5GC) as well as a new radio access
technology called 5G “New Radio” (NR), it is
possible to integrate elements of different
generations in different configuration with 5G: SA
(standalone) and NSA (non-standalone). SA
scenario uses only one radio access technology
(5G NR or the evolved LTE (Long Term Evolution)
radio cells) and the core networks are operated
alone. NSA scenario combines NR radio cells and
LTE radio cells using dual-connectivity to provide
radio access and the core network may be either
EPC (Evolved Packet Core) or 5GC (see Table 1
for details).
Table 1: Comparison of 5G radio access and core networks
Advantages
Disadvantages
SA
• Simple to manage
• Inter-generation handover
between 4G-5G
• Not able to leverage existing
LTE deployments if NR is used
in SA
NSA
• Leverage existing LTE
deployments
• Tight interworking between LTE
and NR required
• May impact end user experience
EPC
• Leverage existing EPC
deployment
• Cloud support is optional
5GC
• Cloud native
• Easier to support multiple
access
• New deployment required
Radio
access
network
Core
network
5
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 1: 4G and 5G Deployment Options
Consequently, five deployment options are
available for 5G as depicted in Figure 11. Red
colour denotes NR and 5GC.
NOTE: for simplicity, the figure does not
depict control and user plane connection
As there can be multiple paths to reach the final
target configuration that an operator intends to
deploy, it is essential to consider migration steps
that would comprise different paths. This
document analyses the following migration steps
that the 5G Introduction project group members
believe to be more likely.
• EPS (Evolved Packet System) to SA Option
#2
• EPS to NSA Option #3
• NSA Option #3 to NSA Option #7 and SA
Option #5
1
For completeness, the figure also includes “Option 1”
representing today’s 4G deployments
6
• NSA Option #3 to NSA Option #3 and SA
Option #2
• NSA Option #3 to NSA Option #4 and SA
Option #2
The analysis of the paths (see Table 2) take four
perspectives. First, feasibility of use case refers to
the ability of the path in addressing 5G use cases.
Second, deployment considerations consider the
support for early 5G devices and core network &
radio access network deployment considerations
of the migration step (i.e. the core network
solution adopted as a result of the path and the
extent of leveraging existing LTE deployment).
Third, impact on device and network considers
the impact of the migration step on device and
network. Finally, impact on voice service including
service continuity refers to the ease of voice
service provisioning and continuity offered by the
migration step.
ROAD TO 5G: INTRODUCTION AND MIGRATION
Table 2: High-level overview of migration step analysis results
Path
Deployment
Device/Network
Voice
• Full 5G use cases
• 5G Core benefits
• Needs to retain
EPC
• Little impact on
4G
• 4G/5G system
interworking
required
• IMS Voice
supported
• No CS
interworking from
5GS
• Limited support
for 5G use case
• Leverage LTE
• Quick time-tomarket
• No 5G Core
benefits
• EPC procedures
• Impact on 4G
• Leverage existing
VoLTE service
NSA#3 to
NSA#7 / SA#5
• Full 5G use cases
• Initially limited
• 5G Core benefits
• Needs to retain
EPC
• Impact on NR,
LTE
• Impact on IMS
• 5GC deployment
• IMS Voice
supported
• No CS
interworking from
5GS
NSA#3 to
NSA#3 / SA#2
• Full 5G use cases
• Initially limited
• Core migration
• 5G Core benefits
• Needs to retain
EPC
• Wide area NR
• Impact on NR,
LTE
• Impact on IMS
• 5GC deployment
• IMS Voice
supported
• No CS
interworking from
5GS
NSA#3 to
NSA#4 / SA#2
• Full 5G use cases
• Initially limited
• Core migration
• 5G Core benefits
• Needs to retain
EPC
• Impact on NR,
LTE
• Impact on IMS
• 5GC deployment
• IMS Voice
supported
• No CS
interworking from
5GS
EPS to SA#2
EPS to NSA#3
1
Use Case1
Limited support for 5G use case denotes focus on enhanced mobile broadband use case
NOTE: CS stands for circuit-switched
The availability of options and migration steps
indicate that the mobile industry needs to respect
the need of different operators to follow different
migration strategies with some degree of
divergence. Nevertheless, customers and the
wider industry will be able to benefit as a whole if
the following collaborative actions are taken by
operators in order to guarantee service continuity,
services and network interoperability and to
unlock economies of scale.
• Global profile for the industry to address
•
•
•
Basic connectivity of devices to a 5G
network (including interoperability
between 5G and LTE networks)
IMS (IP Multimedia Subsystem)
services delivered over NR
Support roaming to/from networks
with same/different 5G deployment
option
• Global issue registry and resolution group
that receives issues encountered during 5G
commercialization and collaboratively
resolves the issues.
7
ROAD TO 5G: INTRODUCTION AND MIGRATION
1
1. Introduction
8
ROAD TO 5G: INTRODUCTION AND MIGRATION
1.1 Scope
CSCF
Call Session Control Function
This document analyses the alternatives available
for operators intending to introduce a 3GPP
compliant 5G system migrating from their 3GPP
4G network. For each of the options and
consequent migration steps, this paper considers
the technical, economic and strategic perspective.
The structure of the document is as follows: Part I
provides a high level description of the 5G
introduction and subsequent migration along
recommendations for collaborative actions. Part II
delves into real operator case study of deploying
5G while Part III outlines advanced technology
considerations requiring a reasonably advanced
level of understanding of the 3GPP system.
CSFB
Circuit Switched Fallback
CUPS
Control Plane – User Plane Separation
E2E
End-to-End
eMBB
enhanced Mobile BroadBand
EMM
EPS Mobility Management
eNB
Evolved Node B
EN-DC
E-UTRA-NR Dual Connectivity
EPC
Evolved Packet Core
ePDG
Evolved Packet Data Gateway
EPS
Evolved Packet System
ESM
EPS Session Management
EU
European Union
E-UTRA
Evolved Universal Terrestrial Radio
Access
E-UTRAN
Evolved Universal Terrestrial Radio
Access Network
EVS
Enhanced Voice Services
FIFA
Federation Internationale de Football
Association
gNB
Fifth Generation NodeB
GSM
Global System for Mobile
Communications
GW
Gateway
HEVC
High Efficiency Video Coding
HO
Handover
It is assumed that the mobile network operator
deploying a 5G network is already operating a 4G
network. This document covers 3GPP mobile
accesses (4G and 5G) only. Non-3GPP accesses
are not covered in this version of the document.
1.2 Abbreviations
Term
Description
2G
Second Generation (Mobile Network)
3G
Third Generation (Mobile Network)
3GPP
Third Generation Partnership Project
4G
Fourth Generation (Mobile Network)
5G
Fifth Generation (Mobile Network)
5GC
5G Core (network)
HSS
Home Subscriber Server
5GPPP
Fifth Generation Private Public
Partnership
IMS
IP Multimedia Subsystem
5GS
Fifth Generation System
IMT
International Mobile
Telecommunications
AMF
Access & Mobility Management Function
IOC
International Olympic Committee
AP
Access Point
IoT
Internet of Things
API
Application Programme Interface
IP
Internet Protocol
AS
Application Server
IPTV
Internet Protocol Television
ASN.1
Abstract Syntax Notation One
ISD
Inter-Site Distance
B2B
Business-to-Business
ITU
International Telecommunication Union
CA
Carrier Aggregation
LCS
Location Services
CAPEX
Capital Expenditure
LIPA
Local IP Access
CN
Core Network
LS
Liaison Statement
CP
Control Plane
LTE
Long Term Evolution
CS
Circuit-Switched
9
ROAD TO 5G: INTRODUCTION AND MIGRATION
MCG
Master Cell Group
SIPTO
Selected IP Traffic Offload
MCPTT
Mission Critical Push-To-Talk
SM
Session Management
mIoT
Massive IoT
SMF
Session Management Function
MME
Mobility Management Entity
SMS
Short Message Service
mMTC
Mobile Machine Type Communications
SN
Secondary Node
MN
Master Node
SRVCC
Single Radio Voice Call Continuity
MPS
Multimedia Priority Service
TB
Terabyte
MR-DC
Multi-RAT Dual Connectivity
TCO
Total Cost of Ownership
NAS
Non Access Stratum
TV
Television
NB-IoT
Narrowband-IoT
TWAG
Trusted Wireless Access Gateway
NE-DC
NR E-UTRA Dual Connectivity
TX
Transmission
NF
Network Function
U-20
Under 20
NFV
Network Function Virtualization
UDM
User Data Management
NGEN-DC
NG-RAN E-UTRA-NR Dual Connectivity
UE
User Equipment
NR
New Radio
UL
Uplink
NRF
Network Repository Function
UMTS
NSA
Non-Standalone
Universal Mobile Telecommunications
System
NWDA
Network Data Analytics
UPF
User Plane Function
OEM
Original Equipment Manufacturer
URLLC
PCF
Policy Control Function
Ultra-Reliable Low Latency
Communications
PCRF
Policy and Charging Rules Function
USA
United States of America
P-CSCF
Proxy Call Session Control Function
ViLTE
Video over LTE
PDCP
Packet Data Convergence Protocol
VoLTE
Voice over LTE
PGW
PDN (Packet Data Network) Gateway
VoWiFi
Voice over WiFi
PGW-C
Control plane of the PGW
WLAN
Wireless Local Area Network
PGW-U
User plane of the PGW
QCI
QoS Class Identifier
QoE
Quality of Experience
QoS
Quality of Service
RAN
Radio Access Network
RAT
Radio Access Technology
RRC
Radio Resource Control
RX
Reception
SA
Standalone
SBA
Service Based Architecture
SCG
Secondary Cell Group
SGW
Serving Gateway
SIG
Special Interest Group
SIP
Session Initiation Protocol
10
ROAD TO 5G: INTRODUCTION AND MIGRATION
1.3 References
Ref
[1]
[2]
Title
“5G New Wave”, 5G Forum
“5G PPP use cases and performance
evaluation models”, 5GPPP
[3]
“5G Spectrum Public Policy
Position”, GSMA
[4]
“5G White Paper”, NGMN
[5]
“Mobile Broadband Transformation
LTE to 5G”, 5G Americas
[6]
“The 5G Era”, GSMA
[7]
“Understanding 5G”, GSMAi
[8]
“White Paper on 5G Vision and
Requirements”, IMT-2020 PG
[9]
3GPP TS 23.214 Architecture
enhancements for control and user
plane separation of EPC nodes
[10]
3GPP TS 23.401 General Packet
Radio Service (GPRS) enhancements
for Evolved Universal Terrestrial
Radio Access Network (E-UTRAN)
access
[11]
3GPP TS 23.501 System
Architecture for the 5G System;
Stage 2
[12]
3GPP TS 23.502 Procedures for the
5G System; Stage 2
[13]
3GPP TS 36.300 Evolved Universal
Terrestrial Radio Access (E-UTRA)
and Evolved Universal Terrestrial
Radio Access Network (E-UTRAN);
Overall description; Stage 2
[14]
3GPP TS 37.340 Evolved Universal
Terrestrial Radio Access (E-UTRA)
and NR
Multi-connectivity; Stage 2
[15]
3GPP TS 38.300 NR; Overall
description; Stage 2
[16]
ITU-R M.2083 “IMT-2020 Vision”,
ITU-R
[17]
R2-1713952 LS Reply to SA WG2 on
Status Icon related to 5G
[18]
S2-178933 LS Reply to SA WG2 on
Status Icon related to 5G
11
ROAD TO 5G: INTRODUCTION AND MIGRATION
2
2. Why 5G?
12
ROAD TO 5G: INTRODUCTION AND MIGRATION
Table 3: Key performance requirements of IMT-2020
2.1 The advent of the 5G era
2.2 Drivers of 5G
Unlike previous generations of mobile networks,
the fifth generation (5G) technology is expected
to fundamentally transform the role that
telecommunications technology plays in society.
5G is also expected to enable further economic
growth and pervasive digitalisation of an hyperconnected society, where not only are all people
are connected to the network whenever needed,
but also many other devices/things virtually
creating the society with everything connected
(i.e. Internet of Everything) [1], [7]. It is not a
coincidence that governments around the world
(especially in the most advanced economies such
as China, EU, Japan, Korea and USA) are
demanding acceleration in the introduction of 5G
technology in their respective markets.
While previous generations of mobile networks
were purpose built for delivering communications
services such as voice and messaging (e.g. 2G) or
mobile broadband (e.g. 4G), 5G will have
flexibility and configurability at the heart of its
design to enable mobile operators to serve IoT
(Internet of Things) use cases and to support
ultra-reliable, low latency connections as well as
enhanced mobile broadband [4], [5], [6], [16].
Particularly, new use cases designed to support
smart cities, smart agriculture, logistics and public
safety agencies will deeply impact for the better
every aspect of our lives.
The variety of business models and services that
5G systems will support, however, leads to
numerous strategies a mobile operator has
available to introduce 5G services. As global reach
and economies of scale for network equipment as
well as devices remain a vital component for the
success of a mobile telecommunications networks,
it is important to ensure that although different
operators may follow different 5G introduction
and deployment options, those two tenets are
preserved.
2.3 Expectations on 5G
Technically, 5G is a system designed to meet the
requirements of IMT-2020 set by the ITU-R
specification M.2083 [16]. 5G will provide more
advanced and enhanced capabilities compared to
4G LTE (IMT-Advanced). The following table
summarises some of the key performance
parameters.
It can be noted that 5G will aim to provide 20
times the peak data rate (speed), 10 times lower
latency (responsiveness) and 3 times more
spectral efficiency than 4G LTE.
13
ROAD TO 5G: INTRODUCTION AND MIGRATION
2.4 Use cases and spectrum bands for 5G
5G has three major use case classes: enhanced
Mobile Broadband (eMBB), mIoT and ultra-reliable
low latency (URLLC). The requirements for the
use case classes and the use cases within each
class vary significantly [2], [4], [8]. For example,
smart meters will require only periodic
transmission of relatively small sized traffic while
enhanced mobile broadband will require bursty
/continuous transmission of large size traffic.
In terms of spectrum bands earmarked for
deployment of 5G, they can be sub-divided in
three macro categories: sub-1GHz, 1-6GHz and
above 6GHz.
Sub-1GHz bands are suitable to support IoT
services and extend mobile broadband coverage
from urban to suburban and rural areas. This is
because the propagation properties of the signal
at these frequencies enable 5G to create very
large coverage areas and deep in-building
penetration. The 1-6GHz bands offer a reasonable
mixture of coverage and capacity for 5G services.
There is a reasonable amount of existing mobile
broadband spectrum identified with this range
which could be used for initial 5G deployments.
Spectrum bands above 6GHz provide significant
capacity thanks to the very large bandwidth that
can be allocated to mobile communications and
thus enable enhanced mobile broadband
applications [3]. The downside of using high
spectrum bands (so called “millimetre wave”2) is
the much reduced coverage size of each cell and
its susceptibility to blocking.
This variety of requirements and spectrum needs
show that there are many options of 5G
introduction and different spectrum bands will be
needed to support all use cases. Operators must
therefore consider the feasibility of different
options in meeting their intended initial use cases
and interoperability of their choice with other
options to ensure their networks deliver the use
cases effectively while supporting global
interoperability.
Figure 2: Capacity and coverage considerations of spectrum
categories
2
As the wavelength is defined as the ratio between the speed
of light and the frequency of the wave, a 1 mm wavelength is
14
obtained at 30 GHz, however it is common practice to use this
terminology for spectrum bands above 25 GHz.
Part I:
3. 5G Network
Deployment Options
ROAD TO 5G: INTRODUCTION AND MIGRATION
3.1 Standalone (SA) and Non-standalone (NSA)
As with the previous generations, 3GPP is defining
both a new 5G core network, referred to as 5GC,
as well as a new radio access technology called
5G “New Radio” (NR). Unlike previous generations
that required that both access and core network
of the same generation to be deployed (e.g.
Evolved Packet Core (EPC) and LTE together
formed a 4G system), with 5G it is possible to
integrate elements of different generations in
different configurations, namely:
• Standalone using only one radio access
technology and
• Non-Standalone combining multiple radio
access technologies.
Dual Connectivity: Operation where a given
UE consumes radio resources provided by at
least two different network points (e.g. NR
access from gNB and LTE access from eNB).
Three variations of NSA are defined in 3GPP:
• Option 3 using EPC and an LTE eNB acting
as master and NR en-gNB acting as
secondary;
• Option 4 using 5GC and an NR gNB acting
as master and LTE ng-eNB acting as
secondary; and
• Option 7 using 5GC and an LTE ng-eNB
acting as master and an NR gNB acting as
secondary.
3.2 Evolved Packet Core (EPC) and 5G Core
Network (5GC)
In a standalone scenario, the 5G NR or the
evolved LTE radio cells and the core network are
operated alone. This means that the NR or
evolved LTE radio cells are used for both control
plane and user plane. The standalone option is a
simple solution for operators to manage and may
be deployed as an independent network using
normal inter-generation handover between 4G
and 5G for service continuity.
5G deployment options are being defined in 3GPP
using either the existing EPC (Evolved Packet
Core, specified in 3GPP TS 23.401 [10]) or the
5GC (5G Core network, specified in 3GPP TS
23.501 [11]).
Three variations of SA are being defined in 3GPP:
While EPC could be considered an evolution of
previous generation packet core networks, the
5GC has been designed from its inception to be
“cloud native”, that is inheriting many of the
technology solutions used in cloud computing and
with virtualisation at its core. 5GC also offers
superior network slicing and QoS features.
Another important characteristic is the separation
of the control plane and user plane that besides
adding flexibility in connecting the users also
allows an easier way to support a multitude of
access technologies, better support for network
slicing and edge computing.
• Option 1 using EPC and LTE eNB access
(i.e. as per current 4G LTE networks);
• Option 2 using 5GC and NR gNB access;
and
• Option 5 using 5GC and LTE ng-eNB access
In non-standalone (NSA) scenario, the NR radio
cells are combined with LTE radio cells using dualconnectivity to provide radio access and the core
network may be either EPC or 5GC depending on
the choice of operator. This scenario may be
chosen by operators that wish to leverage existing
4G deployments, combining LTE and NR radio
resources with existing EPC and/or that wish new
5GC to deliver 5G mobile services. This solution
will require tight interworking with the LTE RAN.
The end user experience will be dependent on the
radio access technology(ies) used.
16
The two architectures follow a very different set
of design principles and the main differences are
discussed more in detail in section 7.2.
NOTE: from 3GPP Release 14 onwards, the
separation of control plane and user plane is also
an available option in 4G.
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 3: Overview of SA and NSA Options
Standalone LTE under EPC
(option 1)
Standalone NR under 5GC
(option 2).
Standalone LTE under 5GC (option
5)
Non-standalone LTE and NR
under EPC (option 3)
Non-standalone NR and LTE
under 5GC (option 4).
Non-standalone LTE and NR under
5GC (option 7)
ROAD TO 5G: INTRODUCTION AND MIGRATION
Part I:
4. 5G Network
Introduction &
Migration paths
18
ROAD TO 5G: INTRODUCTION AND MIGRATION
4.1 General observations and assumptions
As discussed in the previous section, 5G can be
introduced either in standalone mode (option 2
and option 5) using 5GC or in non-standalone
mode, using EPC (option 3) or 5GC (options 4 and
7).
There are therefore several possible “paths”
operators can follow to first introduce 5G and
then migrate it to the target configuration(s). This
section presents a selection of what are thought
to be likely introduction and migration scenarios.
Readers should be aware that many aspects need
to be considered when deciding on the best
migration strategy, including spectrum allocation,
support for other service and industries, and
support of 5G capabilities in terminals.
2017 (see
http://www.3gpp.org/specifications/releases).
4.2.3 Deployment Considerations
Compared to EPC-based deployment options such
as Option 1 and Option 3, this option provides an
open, flexible, and service-based network
architecture for 5G which can fully exert the
ability of 5G. In this context, it may be considered
as the deployment choice for operators who need
to fulfil the market requirements, especially of
vertical industries. It is also a long term network
architecture as it already uses both the newly
defined radio and core network.
Direct interworking between 5GS and 2G/3G CS
domain is not considered at the beginning of
deployment for this option.
In this section, it is assumed that the operator:
• has deployed a full 4G system comprising
an EPC and LTE access
• plans to migrate in mid- or long-term to
5GS.
4.2 EPS to SA Option #2
4.2.1 Description
In this scenario the operator migrates directly
from EPS (Option 1) to the standalone Option 2
with inter-RAT mobility mechanisms used to move
devices between 4G LTE under EPC coverage and
5G NR under 5GC coverage.
4.2.2 Feasibility of the path in meeting 5G use
cases
One of the key benefit of this option is that SA
architecture can take full advantage of 5G end-toend network capabilities supported by NR and
5GC, providing customized service, especially to
vertical industry, in an effective and efficient way.
New features, including service-based
architecture, end-to-end network slicing, and
MEC, can be enabled according to specific
requirement of each service, providing customized
superior user experience. The 3GPP specifications
of Option 2 will be completed by June 2018 with
the 5GC stage 2 specification completed in Dec.
19
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 4: Description of EPS to SA Option #2
4.2.4 Impact on Device and Network
SA Option #2 envisages the deployment of both
NR gNB based NG-RAN as a new radio access and
5GC as new core along with new features on LTE
eNB based E-UTRAN to support inter-RAT
mobility. Option #2 requires the device to support
both a radio front end capable of receiving and
transmitting data over NR as well as new
procedures for the 5GC.
Since SA operator provides services that are
delivered over standalone 5GS, interworking
between 5GS and EPS for service continuity for
those services may be required. Based on
operator strategy, single registration solution with
or without AMF-MME interface as defined in 3GPP
TS 23.501/502 can be deployed [11], [12].
Converged nodes of SMF+PGW-C, UPF+PGW-U,
UDM+HSS and PCRF+PCF are needed
accordingly. MME is also impacted to support 5GS
and EPS interworking for converged SMF+PGW-C
selection or for dual registration without AMFMME interface.
Since Dual-connectivity is not required for Option
#2, workload and cost for 4G existing eNB
upgrade and modification is relatively low, with
only minor upgrades needed to support
interworking with 5G.
20
The UE supports complete set of functionalities
for Control-Plane and User-Plane and for all
interfaces to the network.
To avoid the coverage limitations in UL due to
different TX/RX configurations and network
deployment between downlink and uplink, high
power UE can be used to enhance UL coverage.
4.2.5 Impact on voice including service continuity
NOTE: a detailed description of the technical
implications for providing service continuity in this
scenario is provided in Section 9.
Depending on whether the operator supports
voice services over IMS and whether it provides
national coverage (i.e. 100% of its coverage) or
less, the feasibility of voice service continuity in
this migration step will differ.
If the operator provides VoLTE (Voice over LTE)
with national coverage, then the operator can
either provide IMS voice service over 5G network
(5GS) or utilize existing VoLTE service. If IMS
voice service over 5G network (5GS) is not
provided, the operator may still be in the position
to provide voice by adopting non-standardised
solutions, hence requiring experimental approach.
If the operator only offers VoLTE service over
EPS, then interworking of 5GS with EPS is
ROAD TO 5G: INTRODUCTION AND MIGRATION
necessary for subscribers to fall back from 5G to
4G when voice service is required.
If the operator provides VoLTE with partial
coverage and CS voice complements VoLTE to
support national coverage, then the operator
needs to utilize existing VoLTE service and ensure
that continuity between LTE and CS access is
implemented (i.e., SRVCC: Single Radio Voice Call
Continuity). In the case where the VoLTE service
does not provide continuity between LTE and CS
access, then the voice call continuity would not be
guaranteed for regions where VoLTE is not
supported.
If the operator does not provide VoLTE and
provides CS voice with national coverage only,
then this migration step would involve some
investment. The operator would need to provide
national coverage IMS voice service over 5G
network (5GS). Otherwise, the migration step
would not guarantee voice service continuity as
there is no standardized solution for 5G networks
(5GS) equivalent to the SRVCC operation defined
for 4G networks (EPS) in initial release of 5G
(3GPP Release 15). Note that there is also no
Circuit Switched Fallback (CSFB) operation
available for 5GS.
finishing the specification for other options (see
3GPP news article on approval of NR spec and
industry support for 3GPP NR)
Besides the accelerated time to market, as the NR
will augment the existing capability of the LTE
radio network, this option allows flexible “on
demand” deployment where capacity is needed
using the same or different vendors for LTE and
NR. Furthermore, this option is going to be
maintained in future releases of 3GPP (beyond
release 15) and therefore can be used in longerterm, even if other options are deployed in
parallel.
The capability of deploying NR while anchoring
the communication to the EPC network offers the
opportunity of making optimal use of the
spectrum above 6GHz where operators will have
available the large bandwidths necessary to
deliver the high throughput in hotspots but that
cannot be provided easily over large areas due to
the fast signal attenuation. However, NR in option
3 can also be deployed in spectrum bands below
6GHz and the example above should be
considered as illustration of one of the possible
deployment scenarios.
4.3.3 Deployment considerations of option 3
4.3 EPS to NSA Option #3
4.3.1 Description
This section covers the migration from EPS
(Option 1) to non-standalone Option 3 with the EUTRA extended to allow compatible devices to use
dual connectivity to combine LTE and NR radio
access.
4.3.2 Feasibility of the path in meeting 5G use
cases
One of the key advantages of this option is that it
only requires the development of specifications of
NR as non-standalone access as part of E-UTRAN
connected to EPC rather than the specification of
the full 5G system as it is the case for standalone
NR in section 4.2 and other 5GS cases. In Dec.
2017, 3GPP completed the specifications of Option
3 with the corresponding ASN.1 encoding due in
March 2018 as an intermediate milestone before
Depending on the EPC features defined by 3GPP
in Release 15 and future releases, the EPC
capabilities may represent a possible bottleneck
(e.g., latency) that limits the performance that
could otherwise be extracted from NR.
Data throughput per 5G connected subscriber
(e.g. in SGW/PGW) is expected to increase via NR
and LTE in dual-connectivity. While the increase
of data requires additional consideration in
network planning (for example, adding user plane
nodes for SGW/PGW), adoption of additional
features in the latest releases of 3GPP
specification may resolve the challenge.
Finally, as the 5GC is not going to be used in this
option, none of the differentiating capabilities of
the new architecture described in Section 7.1.2
are available to the operator.
21
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 5: Description of EPS to NSA Option #3
4.3.4 Impact on Device and Network
NSA Option #3 requires deployment of NSA NR
en-gNB in E-UTRAN and new features on LTE eNB
to support EN-DC procedures, hence has impacts
on E-UTRAN. NSA Option #3 has also impact on
UE, but limited impact on EPC and HSS depending
on operators’ choice, and no impact on IMS.
From the point of view of the device, the
attractiveness of this solution is that it only
requires the additional support of specifications of
NR as non-standalone access as part of E-UTRAN
connected to EPC. The device will communicate
with the core network using the same EPC
procedures used by currently available devices
either under only LTE or under both LTE and NR
radio coverage. It should be noted however that
combining of LTE and NR radio interfaces for split
bearers may increase memory requirements.
4.3.5 Impact on IMS voice including service
continuity
NOTE: a detailed description of the technical
implications for providing service continuity in this
scenario is provided in Section 9.
Depending on whether the operator supports
voice services over IMS and whether it provides
national coverage (i.e. 100% of its coverage) or
22
not, the feasibility of voice service continuity in
this migration step will differ.
If the operator provides VoLTE with national
coverage, then there is no negative impact
associated with migrating from EPS to NSA Option
#3. The operator can utilize existing VoLTE
service.
NOTE: The operator can also choose to upgrade
VoLTE to utilize NR
If the operator provides VoLTE with partial
coverage and CS voice complements VoLTE to
support national coverage, then the operator
needs to utilize existing VoLTE service and ensure
that continuity between LTE and CS access is
implemented (e.g., SRVCC). However, this is
something an operator would consider
independently from the deployment of 5G to
ensure voice service continuity.
If the operator does not provide VoLTE and
provides CS voice with national coverage only,
then the operator needs to utilize existing CS
voice service and ensure CSFB is implemented. As
is the case of VoLTE with complementary CS
voice, the decision to adopt this technical solution
is independent from the introduction of 5G.
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 6: Description of NSA Option #3 to NSA Option #7 and SA
Option #5
NOTE 1: Early 5G device refers to devices that support only NSA Option #3 5G deployment option
NOTE 2: New 5G device refers to devices that support NSA Option #3 and also other 5G deployment option(s)
4.4 NSA Option #3 to NSA Option #7 and SA
Option #5
4.4.1 Description
This section covers the migration from nonstandalone Option 3 to non-standalone Option 7
and standalone Option 5 with RAN connectivity to
5GC in parallel to EPC connection for early 5G
devices.
4.4.2 Feasibility of the path in meeting 5G use
cases
Whereas the network was not able to leverage
the advantages of 5GC in NSA Option #3, in this
path 5GC is deployed so that the full advantage of
5G end-to-end network capabilities can be
delivered to the users. This path enables
operators to provide initial 5G use cases (e.g.,
mobile broadband) leveraging LTE and EPC
installed base, while new use cases can be
addressed on clean slate 5GS architecture.
4.4.3 Deployment considerations
Since the 5GC is now used, this path requires that
devices support the new protocol stack to access
this core network. This is problematic not only for
legacy 4G-only devices, but also for early 5G
devices supporting only NSA Option #3 that only
have an EPC protocol stack. Mobile network
operators are likely to maintain Option #3 support
after the introduction of Options #7/5. LTE RAN
also needs upgrade to connect to 5GC and more
LTE base stations (eNode B) may need to be
upgraded to interwork with NR. This path also
requires tight interworking between LTE and NR.
This path allows operators to continue to
selectively deploy NR only where needed. As LTE
is already offered in wide-area coverage in initial
condition (NSA Option #3), the network can still
leverage the wide-area coverage LTE network and
deploy NR only when intended use case requires
it.
4.4.4 Impact on Device and Network
Option #7/5 upgrade from Option 3 requires
deployment of 5GC and upgrade of LTE eNB to
support 5GS session, mobility, QoS management
and MR-DC procedures and 5GC N2/3 RAN-core
interfaces along with upgrade of NR gNB to
support 5GC N3 RAN-Core user plane interface. In
addition to the need to deploy the new 5GC and
(for Option 7) DC, devices will need to support
5GC NAS messages, devices and E-UTRA need to
support extensions to the LTE RRC layer, E-UTRA
23
ROAD TO 5G: INTRODUCTION AND MIGRATION
will need upgrades to terminate the 5GC N2/3
RAN-Core interfaces and both E-UTRA and IMS
will require upgrades to support changes to QoS
model. N2 interface is the interface between the
5G RAN and AMF, while N3 interface is the
interface between the 5G RAN and UPF.
4.4.5 Impact on IMS voice including service
continuity
NOTE: a detailed description of the technical
implications for providing service continuity in this
scenario is provided in Section 9.
Depending on whether the operator supports
voice services over IMS and whether it provides
national coverage (i.e. 100% of its coverage) or
not, the feasibility of voice service continuity in
this migration step will differ.
If the operator provides VoLTE with national
coverage, then there is no significant impact to
service continuity as a result of migrating from
NSA Option #3 to NSA Option #7 and SA Option
#5. The operator can either provide IMS voice
service over 5G network (5GS) or utilize existing
VoLTE service. If IMS voice service over 5G
network (5GS) is provided, the operator will need
to ensure that adequate tests are performed. If
VoLTE service is utilized, then EPC is necessary
for subscribers when voice service is used.
If the operator provides VoLTE with partial
coverage and CS voice complements VoLTE to
support national coverage, then the operator
needs to utilize existing VoLTE service and ensure
that continuity between LTE and CS access is
implemented (e.g., SRVCC and CSFB). In the case
where the VoLTE service does not provide
continuity between LTE and CS access, then the
voice call continuity would not be guaranteed for
regions where VoLTE is not supported.
If the operator does not provide VoLTE and
provides CS voice with national coverage only,
then this migration step would involve some
investment. The operator would need to provide
national coverage IMS voice service over 5G
network (5GS). Otherwise, the migration step
would not guarantee voice service continuity as
24
there is no standardized solution for 5G networks
(5GS) equivalent to the SRVCC operation defined
for 4G networks (EPS) in initial release of 5G
(3GPP Release 15).
4.5 NSA Option #3 to NSA Option #3 and SA
Option #2
4.5.1 Description
This section covers the migration from having
only non-standalone Option 3 to adding
standalone Option 2 with inter-RAT mobility
mechanisms used to move devices between 5G
NSA LTE plus NR under EPC coverage and 5G NR
under 5GC coverage.
4.5.2 Feasibility of the path in meeting 5G use
cases
Whereas the network was not able to leverage
the advantages of 5GC in NSA Option #3, in this
path 5GC is deployed so that the full advantage of
5G end-to-end network capabilities can be
delivered to the users. This path enables
operators to address all use cases on clean slate
5GS architecture. However, the operator may
need to consider migration of initial use cases
served by EPC to 5GC if all use cases are to be
supported by 5GC.
4.5.3 Deployment considerations
Since there is no tight interworking at radio level
between 4G and 5G, this path works best when
NR has been deployed to support wide-area
coverage: while operators can offer seamlessly
handover traffic from 5GC to EPC, this transition
requires also a change of core network
architecture and QoS model and therefore should
not be used too frequently.
Early 5G devices supporting only NSA Option #3,
and capable of communicating only with EPC, will
be able to use their 5G radio capabilities in the
target scenario provided the gNB is able to
support both option #2 and #3 devices
simultaneously.
4.5.4 Impact on Device and Network
Option #2 requires deployment of 5GC and
update of NR gNB to support both NSA (option
ROAD TO 5G: INTRODUCTION AND MIGRATION
#3) and SA (option #2) in parallel. Option #2 has
impacts on E-UTRAN connected to EPC to support
inter-RAT mobility, IMS to support 5GS QoS
management and also on UE.
Figure 7: Description of NSA Option #3 to NSA Option #3 and SA
Option #2
NOTE 1: Early 5G device refers to devices that support only NSA Option #3 5G deployment option
NOTE 2: New 5G device refers to devices that support NSA Option #3 and also other 5G deployment option(s)
4.5.5 Impact on IMS voice including service
continuity
NOTE: a detailed description of the technical
implications for providing service continuity in this
scenario is provided in Section 9.
Depending on whether the operator supports
voice services over IMS and whether it provides
national coverage (i.e. 100% of its coverage) or
not, the feasibility of voice service continuity in
this migration step will differ.
If the operator provides VoLTE with national
coverage, then there is no significant impact
associated with migration from NSA Option #3 to
SA Option #2. The operator can either provide
IMS voice service over 5G network (5GS) or utilize
existing VoLTE service. Prior to the adoption of
IMS voice service over 5G network (5GS), the
operator may need to perform extensive testing.
If VoLTE service is utilized, then it is necessary to
continue to support EPC for subscribers using
voice services.
If the operator provides VoLTE with partial
coverage and CS voice complements VoLTE to
support national coverage, then the operator
needs to utilize existing VoLTE service and ensure
that continuity between LTE and CS access is
implemented (e.g., SRVCC). In the case where
the VoLTE service does not provide continuity
between LTE and CS access, then the voice call
continuity would not be guaranteed for regions
where VoLTE is not supported.
If the operator does not provide VoLTE and
provides CS voice with national coverage only,
then this migration step would involve some
investment. The operator would need to provide
national coverage IMS voice service over 5G
network (5GS). Otherwise, the migration step
would not guarantee voice service continuity as
there is no standardized solution for 5G networks
(5GS) equivalent to the SRVCC operation defined
for 4G networks (EPS) in initial release of 5G
(3GPP Release 15).
25
ROAD TO 5G: INTRODUCTION AND MIGRATION
4.6 NSA Option #3 to NSA Option #4 and SA
Option #2
4.6.1 Description
This section covers the migration from nonstandalone Option 3 to non-standalone Option 4
and standalone Option 2 with RAN connectivity to
5GC in parallel to EPC connection for early 5G
devices.
Figure 8: Description of NSA Option #3 to NSA Option #4 and SA
Option #2
NOTE 1: Early 5G device refers to devices that support only NSA Option #3 5G deployment option
NOTE 2: NSA 5GC device refers to devices that support NSA Option #4
NOTE 3: SA 5GC device refers to devices that support SA Option #2
4.6.2 Feasibility of the path in meeting 5G use
cases
In this path, unlike in NSA Option #3, the 5GC
core network is used to replace the EPC in serving
5G use cases (if 5GC replaces EPC completely
then legacy 4G UEs or UEs with 4G only
subscription will no longer be served). This means
that users can takes full advantage of 5G end-toend network capabilities. This path enables
operators to address all use cases on clean slate
5GS architecture. However, the operator may
need to consider migration of initial use cases
served by EPC to 5GC if all use cases are to be
supported by 5GC.
4.6.3 Deployment considerations
Early 5G devices supporting only NSA Option #3
and only able to connect to EPC will imply that
26
operators will keep maintaining the EPC for longterm. In addition, LTE RAN needs upgrade to
connect to 5GC and more LTE base stations
(eNodeB) may need to be upgraded to interwork
with NR. This path also requires tight interworking
between LTE and NR.
This path allows operators to continue to
selectively deploy NR only where needed.
However, compared with the migration described
in section 4.4., this path may require the
deployment of a larger number of more NR gNB
since NR acts as is the master node with LTE
secondary node in the area where option #4 is to
be used. As the starting point of this migration is
scenario #3 where NR and eLTE are already
tightly interworking, it is possible to continue to
leverage the wide area coverage of LTE and
ROAD TO 5G: INTRODUCTION AND MIGRATION
supplement it with selective NR deployment based
on demand.
4.6.4 Impact on Device and Network
Option #4/2 upgrade from Option 3 requires
deployment of 5GC and upgrade of NR gNB to
support 5GS session, mobility, QoS management
and MR-DC procedures and 5GC N2/3 RAN-core
interfaces along with upgrade of LTE eNB to
support 5GC N3 RAN-Core user plane interface
(only required if option 4A is adopted). N2
interface is the interface between the 5G RAN and
AMF, while N3 interface is the interface between
the 5G RAN and UPF. This step also has impacts
on IMS to support 5GS QoS management and also
on UE (refer to Section 7 for more details).
4.6.5 Impact on IMS voice including service
continuity
NOTE: a detailed description of the technical
implications for providing service continuity in this
scenario is provided in Section 9.
Depending on whether the operator supports
voice services over IMS and whether it provides
national coverage (i.e. 100% of its coverage) or
not, the feasibility of voice service continuity in
this migration step will differ.
If the operator provides VoLTE with national
coverage, then there is no significant impact
associated with migrating from NSA Option #3 to
NSA Option #4. The operator can either provide
IMS voice service over 5G network (5GS) or utilize
existing VoLTE service. Prior to the adoption of
IMS voice service over 5G network (5GS), the
operator may need to perform extensive testing.
If VoLTE service is utilized, then EPC is necessary
for subscribers when voice service is used.
If the operator provides VoLTE with partial
coverage and CS voice complements VoLTE to
support national coverage, then the operator
needs to utilize existing VoLTE service and ensure
that continuity between LTE and CS access is
implemented (e.g., SRVCC). In the case where
the VoLTE service does not provide continuity
between LTE and CS access, then the voice call
continuity would not be guaranteed for regions
where VoLTE is not supported.
If the operator does not provide VoLTE and
provides CS voice with national coverage only,
then this migration step would involve some
investment. The operator would need to provide
IMS voice service over 5G network (5GS).
Otherwise, the migration step would not
guarantee voice service continuity as there is no
standardized solution for 5G networks (5GS)
equivalent to the SRVCC operation defined for 4G
networks (EPS) in initial release of 5G (3GPP
Release 15).
4.7 Other migration steps
This section lists the migration steps that are not
covered in this version of the document for
reader’s information:
•
•
•
•
NSA Option #7 to NSA Option #4
SA Option #5 to NSA Option #4
EPS to NSA Option #4
EPS to NSA Option #7
NOTE: The list above is not exhaustive and
includes only the migration steps that are
considered as more likely by the project group
27
ROAD TO 5G: INTRODUCTION AND MIGRATION
Part I:
5. Recommendations
for collaborative
actions
28
ROAD TO 5G: INTRODUCTION AND MIGRATION
The variety of options and migration steps
available for deploying 5G allow operators to
adopt different strategies that suit the specific
market situations, business models and
competition needs. Also, customers and the wider
industry will benefit as a whole from operators
taking certain collaborative actions that are aimed
to guarantee service continuity, service and
network interoperability and to unlock economies
of scale. Similar to the work carried out earlier for
IMS services (e.g. VoLTE, VoWiFi, ViLTE), a profile
should define the minimum requirements that
need to be fulfilled in order to ensure
interoperability, roaming and portability across the
various 5G deployment options. These profiles
should aim to address:
• Basic connectivity of devices to a 5G
network including interoperability between
5G and LTE networks
• IMS services delivered over NR
• Support roaming to/from networks with
same/different 5G deployment option
The creation of a database of issues to be
encountered during the 5G commercialisation, as
it was done for IMS services, will allow the
industry to efficiently address them and find
solutions that every stakeholder will benefit from.
29
ROAD TO 5G: INTRODUCTION AND MIGRATION
Part II:
6. Study of 5G
network introduction
case
30
ROAD TO 5G: INTRODUCTION AND MIGRATION
This section contains 5G case studies contributed by KT to share their 5G network introduction
cases to provide insights and foresight for challenges in 5G commercialisation in the near future.
This case study addresses KT’s 5G trial activities, tentative 5G commercialization plan and
roadmap, issues in 5G deployment/migration and KT’s challenges from technology and business
perspectives.
6.1 Why 5G?
6.1.1 5G technology and service trial activities
To develop early technical specifications for 5G
trial system and services for the PyeongChang
Winter Olympics that will be held in February
2018, KT launched ‘5G SIG (Special Interest
Group)’ along with global mobile manufacturers
(Ericsson, Intel, Nokia, Qualcomm, and Samsung)
in September 2015 and then completed the 5G
SIG specifications in November 2016.
Furthermore, KT successfully completed the IOC
Olympic Test Events in 2016 and 2017 based on
5G trial base stations and terminals. KT is
currently carrying out lab and field tests for
debugging and verifying the functionalities and
performance of 5G as well as service demos
during major events, as shown in Figure 9. During
the FIFA U-20 World cup held in South Korea,
interactive broadcasting of soccer games using
360-degree VR and Time Slice technology was
offered through 5G connections.
The 5G trial system including core and radio
access network, along with the 5G trial devices,
will be deployed by the end of 2017 in Seoul and
Olympic venue areas. 5G trial services will be
offered during the Olympic Games in
PyeongChang, GangReung, JeongSeon, and
BoGwang areas as well as in Seoul.
In PyeongChang Olympics, various 5G trial
services will be offered by KT such as fixed
wireless access service, mobile services, and
autonomous vehicle services.
Figure 9: KT’s 5G Trials and Commercialization Activities
31
ROAD TO 5G: INTRODUCTION AND MIGRATION
In PyeongChang Olympics, various 5G trial
services will be offered by KT such as fixed
wireless access service, mobile services, and
autonomous vehicle services. Some key examples
of the 5G services are shown in Figure 10 and
Figure 11.
• Sync View enables viewers to enjoy the
game with high-resolution player-view
cameras on players.
• 360-degree Live VR offers capability for
viewers to watch interactive UHD-class
media in 360-degrees.
• Hologram Live enables mobile users to
transfer next-generation 3D hologram
media contents.
• Time Slice Broadcasting offers capability for
viewers to watch realistic and close slowmotions of Olympic events by using
multiple connected cameras.
• Smart Surveillance offers perfect safety and
security for visitors and on-site operators
by using connected, intelligent surveillance
fixed cameras and drones.
• 5G-connected Autonomous Vehicles offer
autonomous driving experience and
immersive in-vehicle next-generation
media.
6.1.2 Other business/social drivers in the nation
South Korea is fully covered with nation-wide LTE
network and the LTE traffic already comprises
more than 99% of the mobile traffic. Total
number of mobile subscribers is around 60 million
and smartphone adoption rate is 85% in South
Korea.
Figure 10: KT’s 5G Trial Services for Olympics
32
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 11: KT’s 5G-connected Autonomous vehicle
Besides, the annual growth rate of mobileconnected IoT devices in South Korea is around
20%. It is also expected that subscribed IoT
devices will grow in number as commercial IoT
network based on 3GPP NB-IoT specifications was
deployed for KT LTE network in July 2017.
Furthermore, KT is also concentrating on
enterprise and B2B IoT segments like home/office
IoT, smart factory, smart city and smart energy.
In South Korea, video traffic occupies the biggest
portion of the traffic (i.e. 56% as of June 2017),
where mobile IPTV and video streaming services
over LTE connection are widely spread. That is,
smartphones/tablets are starting to substitute
traditional TV.
Mobile data usage in South Korea is steadily
increasing. As of June 2017, total data volume per
month exceeds 290,000 TB and monthly data
volume per subscriber exceeds 5 GB. This is
around 5 times the average volume of mobile
traffic of the past 5 years. Besides, the growth
rate of mobile traffic from the previous year is
around 40%. It can be estimated that mobile
traffic of around 700,000 TB and 1,000,000 TB
will be reached in the end of 2019 and 2020,
respectively. Of course, connected vehicles may
accelerate the growth of mobile traffic in the
future. Figure 12 shows the trend and forecast of
mobile data traffic in South Korea from 2012 to
2019. Note that projected data during December,
2017 to December, 2020 are estimated based on
the recent mobile status reports of KCC (Korea
Communications Commission), South Korea.
Since mobile capacity in 2020’s must be expanded
to accommodate the growth of mobile traffic as
well as the adoption of mobile broadband services
and IoT devices (especially in urban and hot-spot
areas), the early introduction and deployment of
5G mobile broadband network is inevitable.
6.1.3 Deployment purpose and spectrum at
disposal
In South Korea, social needs and market
competition for reducing retail cost per mobile
data used (bit) is growing, as mobile technology is
widespread and is commoditized in everyday life.
Therefore, timely response from mobile operators
becomes urgent. It is expected that epoch-making
cost reduction for mobile data can be achieved by
introducing relatively low-cost, ultra-broadband
millimetre-wave bands to mobile network.
33
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 12: Exploding Mobile Traffic in South Korea
South Korean government released ‘K-ICT’
national spectrum plan in January 2017 to serve
the world’s first commercial 5G mobile network in
South Korea, so that new broadband spectrum up
to 300 MHz bandwidth in 3.5 GHz and 1,000 MHz
bandwidth in 28 GHz is planned to be allocated in
2018~2019 timeframe. Therefore, KT is positively
considering using these new bands including LTE
bands for 5G network deployment.
KT’s 5G network will be deployed to serve
primarily eMBB services by offering ultra-high
speed and sufficient capacity in the initial stage,
and then will be extended to accommodate URLLC
and mMTC services. Especially, the large-scale
deployment of small cells in Seoul urban areas is
required due to the accommodation of large
traffic and smaller millimetre-wave band cells
(i.e., cell ISD of 100~200 meters). Therefore, the
acquisition, deployment, and operation of ultradense cells become more important. This means
that the development of small-sized and lowweight base stations, multi-vendor interoperability
of end-to-end system, and next-generation SON
technology are urgently required for commercial
5G network.
34
6.2 Analysis of 5G migration options
Several 5G network options for non-standalone
and standalone architectures are specified in
3GPP, as described in the Section 3. The technical
comparison between 5G NSA and SA options for
mobile network can be summarized as Table 4.
For simplicity, we consider only the NSA Option
3/7 and the SA Option 2 here, so that we can
further discuss and analyse major technical issues
of NSA and SA options.
• Spectrum: Due to the coverage limitation
and cost burden of mmWave band cells, it
is best used only for sub-6GHz SA or NSA
mobile networks.
• Core network: Since Option 7 and Option 2
require new 5GC system, the development
and verification of 5GC is needed. Besides,
enhancement of EPC like CUPS of EPC for
Option 3 is desirable to accommodate large
NR capacity. To support inter-system
mobility in SA network, interconnection
interface between 5GC and EPC may be
required which is not necessary for NSA
network. When it comes to 5G-specific
services, 5GC can naturally support them
through new QoS framework and network
slicing features.
In terms of devices, since Option 3 requires
ROAD TO 5G: INTRODUCTION AND MIGRATION
relatively small investment. However, this
deployment requires upgrade of LTE eNB
and EPC based on EN-DC for Option 3.
Furthermore, significant upgrade of LTE ngeNB based on NGEN-DC and introduction of
5GC are required for Option 7.
single EPC-NAS protocol, UE complexity is
lower than that of Option 7 or SA Option 2.
• RAN: Because the NSA networks can fully
leverage existing LTE infrastructure as
coverage layer, it is ideal in quickly
deploying full-coverage 5G network with
Table 4: Technical comparison between 5G NSA and SA options
NSA
3GPP 5G specification
5G spectrum
CN
CN interworking
Network slicing and 5G QoS
UE impact (for 5G/LTE dual-mode)
Leverage of LTE
LTE upgrade
RAN interworking
Inter-RAT data session
continuity
Forward compatibility with SA
or Release-16 onwards
Op. 3
Rel-15 (’17.12) – 1st
prioritised,
Sub-6GHz and mmWave
bands are feasible
EPC,
Not required
Sub-6GHz and mmWave
bands are feasible
5GC
Not required
Not supported
EPC-NAS
Full
Required (eNB and EPC)
EN-DC
MR-DC and intra-system
handover
Low
Supported
5GC/EPC-NAS
Full
Required (ng-eNB and 5GC)
NGEN-DC
MR-DC and intra-system
handover
Mid
6.3 Rationale behind migration option
In this section, we also discuss and analyse major
economic and business issues of NSA and SA
options. The key comparison between 5G NSA
and SA options for mobile network in terms of
cost, time, and service aspects can be
summarized as follows. Note that the estimation
of cost and time can be different depending on
various technical and business variables.
• System/Device availability: From its
experience in LTE commercialisation, KT
spent 18 months to develop LTE systems
and devices and then deployed commercial
LTE network in major cities and provinces
in South Korea after 3GPP LTE
specifications were ready. Therefore, it is
expected that KT will be able to
commercialise 5G network during 2019 as
3GPP 5G specifications will be completed
during 4Q of 2017 as Phase 1 (NSA Option
3) and 2Q of 2018 as Phase 2 (SA) of
Release-15. Of course, the exact 5G
commercialisation plan may change
depending on different factors.
Op. 7
Rel-15 (’18.6)
SA (Op. 2)
Rel-15 (’18.6)
Sub-6GHz band is desirable
5GC
Required (with or without N26
interface between 5GC and EPC)
Supported
5GC/EPC-NAS
Partial (Reattach)
None or minor
NR-DC (Intra-RAT)
Inter-system handover (N26)
High
• Deployment: Since NR cells can be
deployed in urban hotspot areas as capacity
and speed boosting cells in NSA network,
deployment time will be shortened and NR
CAPEX will be lower than that of SA
network. However, more investment will be
required to upgrade existing LTE RAN for
Option 7. In addition, acquisition cost for
5G spectrum for SA is higher than that for
NSA since sub-6GHz band is required for
nation-wide coverage.
• Migration: KT has a plan to migrate to SA
network from initial NSA network in short to
mid-run. Therefore, initial deployment of
NSA network requires additional cost and
time for migration to mature SA network.
When it comes to the upgrade aspects of
LTE system, Option 7 requires more costs
because of the major upgrade of legacy
LTE base stations as well as the
introduction of 5GC than other options.
• Services: Option 3 has a big disadvantage
as it is limited in supporting 5G-specific
services due to the use of legacy EPC. On
the other hand, nation-wide SA network will
give best NR coverage quality although the
coverage of initial SA network is limited. In
35
ROAD TO 5G: INTRODUCTION AND MIGRATION
dealt in the previous sections indicate that
smooth, phased migration of 5G network based
on NSA and SA architectures is desirable for quick
and stable introduction of 5G networks within
2019 timeframe. First of all, MR-DC plays key role
in 5G-LTE interworking and migration. Of course,
standalone EPC/LTE network has to be supported
for a considerable time, especially to support inbound roaming users.
addition, since Option 7 has evolved, 5Ggrade LTE access and core network
capabilities, overall service quality is higher
than that of Option 3.
6.4 Challenges in actual migration
We observe that there are significant trade-offs
among deployment time, customer experience,
TCO, and system/UE availability for 5G
introduction. The technical and business aspects
Table 5: Economic comparison between 5G NSA and SA options
NSA
Op. 3
System availability
SA (Op. 2)
Op. 7
’19.2Q
’19.4Q
’19.4Q
Short
Medium
Long
Low
Low
Mid
High
High
Low
Medium (mmWave bands can
Medium (mmWave bands can
High (Sub-6GHz band is required for
be used for SN cells)
be used for SN cells)
coverage cell)
High
Not supported
Medium
Supported
None
Supported
Medium
High
Low
NR coverage quality
Low
Low
Voice service for 5G UE
CSFB and VoLTE
CSFB and VoLTE
High (Sub-6GHz + mmWave),
Medium (Sub-6GHz or mmWave)
VoLTE and VoNR
(estimated from Spec.+18
months)
Required time for
deployment
Deployment cost (NR)
Cost for LTE system
upgrade
Acquisition cost for 5G
spectrum
Migration cost to SA
Support of new 5G
services
Overall service quality
(initial deployment)
KT is currently considering both 5G NSA Option 3
and/or SA Option 2 networks as the prioritized
options for our nationwide 5G network. The
adoption of Option 7 evolving from Option 3 is to
be decided. The SA Option 5 is not considered
and the NSA Option 4 requires further analysis.
Figure 13 shows KT’s 5G mobile network
migration plan that consists of three deployment
phases:
possible that the SA Option 2 network
coexists with the NSA network or replaces
the NSA network.
EPC+ may be upgraded to NFV-based EPC
(CUPS of EPC) and support inter-CN
interworking between EPC and 5GC.
Voice over EPS and/or 5GS is also required
as KT provides nation-wide VoLTE
coverage; however, the support of fallback
to 3G CS voice may be for further
consideration.
• Phase 1 (Early 5G): The NSA Option 3
network is deployed where NR and LTE
cells interwork through EN-DC. NR cells act
as data boosting cells in hot-spot areas
under nationwide LTE cells.
• Phase 2 (Full-scale 5G): The NSA Option 3
network can migrate to NSA Option 7
network based on NGEN-DC. LTE eNB will
be upgraded to support the LTE evolution
(eLTE) of Release-15 onwards. It is also
• Phase 3 (All-5G): As a long-term migration
path of 5G, 5G unified network based on
the SA Option 2 along with standalone LTE
network will be operated. Refarming of LTE
bands can be performed at this stage.
Further, 5GC may control mobile network
as well as WiFi Aps through 5G’s accessagnostic core feature.
36
ROAD TO 5G: INTRODUCTION AND MIGRATION
When it comes to spectrum considerations for 5G,
the combinations of low (sub-3 GHz), mid (3~6
GHz), and high (above-6 GHz) spectrum bands
will be used. Each spectrum band are to provide
wide-coverage, capacity-enhanced coverage, and
full-capacity booster cells, respectively. Therefore,
the promising band combinations of 3.5 GHz and
28 GHz for NSA and SA networks under
consideration are as follows:
• NSA (MR-DC): LTE 1.8 GHz (master) + NR
28 GHz (secondary)
• SA (NR-DC): NR 3.5 GHz (master) + NR
3.5/28 GHz (secondary)
Note that NR 3.5 GHz band, if available, can also
be used as secondary cells in the NSA network.
Figure 13: KT’s 5G Network Migration Plan
When it comes to 5G mobile network deployment
scenarios, we are considering two possible
scenarios for nationwide deployment, as
illustrated in Figure 14. Of course, the scenarios
can be adjusted according to future business and
investment plan, frequency auction, allocation,
regulation, etc.
• Scenario #1 (NSA-First Deployment): NSA
à SA (or SA and NSA): Quick 5G
deployment with low CAPEX and simple
operation
• Scenario #2 (Mixed-First Deployment): SA
in Seoul metropolitan and major cities +
NSA in other areas à nationwide SA:
Traffic-driven deployment with
geographically different options
We have identified and discussed key 5G
deployment and migration issues; however, there
can still be many issues remaining. They are
briefly listed below for further study:
• Migration timeframe from NSA to SA:
Different specification versions can be
applied to NSA and SA networks separately.
For example, NSA in 2019 with Release-15
and then SA in 2020 with Release-16 can
be used.
• Whether to upgrade to Option 7 after the
initial deployment of Option 3 or not:
Option 7 offers better performance than SA
Option 2 with MR-DC and ng-eNB
capabilities. Besides, Option 7 can offer 5Ggrade services even for upgraded LTE
users. Upgrade to Option 7 will be smooth
37
ROAD TO 5G: INTRODUCTION AND MIGRATION
•
•
•
•
when 5GC is deployed and full spec is
available. However, migration to Option 7
from Option 3 may not be necessary as SA
Option 2 network will be quickly deployed
nationwide.
Partial refarming of LTE to NR band can be
considered at the mature 5G stage:
Bandwidth expansion (up to 150 MHz) can
be achieved through 1.8 GHz + 3.5 GHz NR
CA. Co-existence between 5G and LTE
needs to be studied further.
Support of legacy 5G devices: For example,
NSA Option 3 devices should be supported
even for Option 7 and the SA networks
Early adoption of SA network for NR-based
FWA (Fixed Wireless Access) service before
2019 timeframe
Unified control of WiFi Aps through 5G
network: Access-agnostic 5GC in Option 7
and SA networks can provide unified mobile
and WiFi access and control capability
(expected in 3GPP Release-16 onwards).
Therefore, the role of WiFi AP in 5G needs
to be clarified and positioned.
6.5 Lessons learned from migration case study
KT is actively designing and developing
commercial 5G systems and devices by
collaborating with major operators and
manufacturers. This is for KT to pre-emptively
establish next-generation mobile infrastructure to
enable a wide-spectrum of convergence services
in 2020’s through ubiquitous ultra-fast Giga-bit
mobile broadband connectivity in South Korea.
Figure 14: KT’s 5G Network Deployment Scenarios
38
ROAD TO 5G: INTRODUCTION AND MIGRATION
ROAD TO 5G: INTRODUCTION AND MIGRATION
Part III:
7. Core network
considerations
40
ROAD TO 5G: INTRODUCTION AND MIGRATION
EPC handles EMM (EPS Mobility Management)
procedures and ESM (EPS Session Management)
procedures. See Figure 15 for the architecture of
EPC.
Control and User Plane Separation (CUPS) is a
new feature introduced in Release 14 of 3GPP.
CUPS allows operators to flexibly place the
separated control and user plane functions to
support diverse deployment scenarios without
affecting the overall functionality provided by EPC
entities. In addition, EPC will be able to address
latency limits if Control and User Plane Separation
is deployed but not otherwise. CUPS supports
flexible/scalable user plane deployment without
expanding or upgrading the control plane.
Especially in cloud native case, the simpler
configuration and maintenance work of user plane
nodes can be achieved. It is also important to
note that CUPS in EPS can be deployed before or
together with Release 15 Option 3 when needed.
See Figure 16 for the EPC architecture with CUPS
implemented. (Reference: 3GPP TS 23.401 [10]
and 3GPP TS 23.214 [9])
NOTE: For simplicity, the figure does not include
interworking with GSM/UMTS networks
NOTE: For simplicity, the figure does not include
interworking with GSM/UMTS networks
7.1 Core network solutions
7.1.1 EPC (Evolved Packet Core) [10]
EPC is the core network element of 4G network,
which consists of MME (Mobility Management
Entity), HSS (Home Subscriber Server), S-GW
(Serving Gateway), P-GW (PDN Gateway) and
PCRF (Policy and Charging Rules Function). While
MME handles the control plane, S-GW and P-GW
handle both control and user plane. PCRF is also
an integral component of EPC that provides rules
for policy (e.g., QoS) and charging.
NOTE: There may be more network nodes
depending on the features implemented.
Figure 15: EPC architecture – non-roaming
Figure 16: EPC Architecture with CUPS – non-roaming
41
ROAD TO 5G: INTRODUCTION AND MIGRATION
7.1.2 5GC (5G Core) [11]
With the NFV technologies, the mobile network
functions can be virtualized and hosted in a cloud
environment. The physical boundary between
traditional EPC network elements such as MME,
SGW, and PGW will be blurred with virtualization
and software. This promotes the 5GC to be
redesigned and become open and flexible enough
to meet the diversity of service and business
requirement in 5G era.
• AF: Application Function
• AMF: Access and Mobility Management
Function
• AUSF: Authentication Server Function
• NEF: Network Exposure Function
• NRF: Network Repository Function
• NSSF: Network Slice Selection Function
• PCF: Policy Control Function
• SMF: Session Management Function
• SMSF: SMS (Short Message Service)
Function
• UDM: User Data Management
• UPF: User Plane Function
5GC proposes the SBA architecture, which provide
unprecedented efficiency and flexibility for the
network. SBA is an architectural for building
system based on fine-grained, interaction of
loosely coupled and autonomous components
called services. This architecture model is chosen
to take full advantage of the latest virtualization
and software technologies.
Compared to the previous generation reference
point architecture as EPC, the elements of servicebased architecture are defined to be the
NF(network functions), which interconnect with
the rest network functions across a single API
calling interface and provide the authorized
services to them. Network repository functions
(NRF) allows every network function to discover
the services offered by other network functions.
A service is an atomized capability in a 5G
network, with the characteristics of high-cohesion,
loose-coupling, and independent management
from other services. This allows individual services
to be updated independently with minimal impact
to other services and deployed on demand. A
service is managed based on the service
framework including service registration, service
authorization, and service discovery. It provides a
comprehensive and highly automated
management mechanism implemented by NRF,
which greatly reduces the complexity of network
maintenance. A service will interact with other
services in a light-weight manner, e.g. API
invocation.
Figure 17: Service Based Architecture of 5G Core – non-roaming
42
ROAD TO 5G: INTRODUCTION AND MIGRATION
To provide the smooth migration to 5GS, the 5GC
is required to support the interworking with
legacy EPC and IMS system, including N26, Cx,
Sh, Rx interfaces, as illustrated in Figure 18,
Figure 19 and Figure 20. In principle there are
two possibilities to implement the interworking
functionality. The first one is that the NF of SBA
where the services operate provides the legacyreference-point interface to support the
interworking with the EPC/IMS network elements.
An alternative way is that the EPC/IMS network
elements provides the service-based interface that
interworks with the NF of 5GC. Either should be
designed and developed with little impact on the
individual function of the 5GC NF or the EPC
network element. According to the present 3GPP
specifications in Rel-15, the AMF, UDM/HSS and
PCF(5GC network function) supports the legacy
reference-point interface that respectively
interworks with MME(EPC network element) via
N26 interface and AS/CSCF/P-CSCF (IMS network
elements) via Sh/Cx/Rx interface.
Figure 18: 5GC exposes N26
legacy- reference-point
interface to EPC domain
Figure 19: 5GC exposes Sh/Cx
legacy- reference-point
interface to IMS domain
SBA is the natural step that enables 5G network
functionality to become more granular and
decoupled, which shall bring the following
benefits to 5G:
• The network is highly efficient on rolling out
new network features, allowing network
operators to quickly deploy new business
and services. Services that operate with
finer granularity than in legacy networks
and that are loosely-coupled with each
other allow individual services to be
upgraded with minimal impact to other
services. This reduces the time to market
for installing bug fixes, and rolling out new
network features.
• The network is extensible. Each service can
interact directly with other services with a
single light-weighted service based
interface. In such manner, the service
based interface can be easily extended
without introducing new reference points
and corresponding message flows.
• The network will be modular and support
reusability. The network is composed of
modularized services, which reflects the
network capabilities. A service can be easily
invoked by other services (with appropriate
authorization), enabling each service to be
reused as much as possible.
• The network is easily open. The information
about a 5G network can be easily exposed
to external users such as 3rd-parties (e.g.
enterprise) through a specific service
without complicated protocol conversion.
Figure 20: 5GC exposes Rx
legacy- reference-point
interface to IMS domain
43
ROAD TO 5G: INTRODUCTION AND MIGRATION
7.2 Comparison of EPC and 5GC
The following table provides a comparison of main differences between EPC and 5GC that will impact
feasibility of 5G use cases.
Table 6: EPC vs. 5GC (Main differences)
Feature
Access Network
interface
Procedures
EPC (4G Core)
S1 with per UE assigned MME & SGW (in
case of CUPS: single SGW-C and multiple
SGW-U) & multiple PDN GWs
ePDG and TWAG for non-3GPP access
Access dependent procedures.
5GC (5G Core)
N2/3 common to all access with per UE
assigned AMF & multiple N3 to UPFs
Unified registration, authentication, session
management for 3GPP, non-3GPP access
(including untrusted, trusted WLAN and in Rel
16 also fixed access), Common N1/N2/N3 for
3GPP and non-3GPP access, enabling
seamless mobility
Network slicing
Single slice per UE with multiple PGW
E2E multiple slices per UE with shared AMF,
slicing aware RAN and per slice SMF/UPF
(potentially slice specific PCF, NRF, etc.)
QoS model
QCI based bearers
QoS flow based framework, including
reflective QoS support
Cloud native
Possible but node based (box driven)
Native support for cloud based deployment
with service based architecture and servicebased interfaces within 5GC CP; Definition of
NF services
Local applications
Support LIPA/SIPTO based deployment
Support for edge computing. Application
influence on traffic routing.
Session/service
continuity
Supports full IP address preservation for
centralized GW or break before make
solution for local GW (LIPA/SIPTO);
service continuity with 2G/3G PS and
SRVCC
Improved Session model including different
Session and Service Continuity modes.
Support for concurrent (e.g. local and central)
access to a data network. No service
continuity with 2G/3G PS and no support of
SRVCC.
Policy framework
Support SM/QoS based policies
Unified Policy framework for Access and
mobility control, QoS and charging
enforcement, policy provisioning in the UE;
introducing NWDA (Network Data Analytics)
for data analytics support.
Services
supported
SMS over NAS over 3GPP access, IMS
services over 3GPP access and non-3GPP
access, LCS, MPS, MCPTT, mobile
broadband
SMS over NAS (including over Non- 3GPP),
support of IMS services over 3GPP access and
non-3GPP access, LCS, MPS, MCPTT, mobile
broadband
Support for RRC
INACTIVE state
Support not specified
Support for RRC inactive (RRC state machine
includes 3 states: RRC IDLE, INACTIVE,
CONNECTED)
44
Part III:
8. Detailed
considerations on 5G
deployment options
ROAD TO 5G: INTRODUCTION AND MIGRATION
8.1 Standalone considerations
In standalone deployment, MR-DC (Multi-RAT
Dual Connectivity) is not employed and only single
RAT is used to connect UE to the relevant core
network. In other words, different generations of
radio access network elements are not
interconnected with each other but only with
those of the same generation (e.g., gNB and gNB
are interconnected but gNB and ng-eNB are not
interconnected in standalone scenario).
Depending on the deployment option considered,
the following radio access network elements are
relevant:
• eNB: Option #1 (Standalone E-UTRA in
EPS) (3GPP TS 36.300) [13]
• gNB: Option #2 (Standalone NR in 5GC)
(3GPP TS 38.300) [15]
• ng-eNB: Option #5 (Standalone evolved EUTRA in 5GC) (3GPP TS 38.300) [15]
In deployment option #1, eNB is connected to
EPC via S1 interface (3GPP TS 23.401) [10].
In deployment option #2, gNB is connected to
5GC via N2 and N3 interface (3GPP TS 23.501)
[11].
In deployment option #5, ng-eNB is connected to
5GC via N2 and N3 interface (3GPP TS 23.501)
[11].
8.2 Non-standalone considerations
In non-standalone deployment, MR-DC is
employed. MR-DC enables the UE to utilise radio
resources provided by two distinct schedulers in
two different nodes connected via non-ideal or
ideal backhaul, as described in 3GPP TS 37.340
[14]. One scheduler is located in the MN (Master
Node) and the other in the SN (Secondary Node),
and one node provides E-UTRA access and the
other node provides NR access. The MN and SN
are interconnected and at least MN is connected
to the core network. (3GPP TS 37.340) [14]
46
The general principle in MR-DC is that there is one
control plane connection between the MN and a
corresponding core network entity for each MRDC UE. The SN for the MR-DC UE establishes
control plane connection via MN. (3GPP TS
37.340) [14]
For user plane, there are different connectivity
options and hence the user plane bearers are
defined per the entity it connects with and the
type of resource it uses. For the transport of user
plane connection between the core network entity
and master/secondary node, the following types
of bearers are defined, each of which can be
either an MN terminated or SN terminated bearer:
• MCG (Master Cell Group) bearer
• SCG (Secondary Cell Group) bearer
• Split bearer
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 21: Control-plane connectivity for MR-DC
MN/SN terminated bearer corresponds to the user
plane connection to the CN entity that terminates
in the MN/SN respectively. MCG (Master Cell
Group) bearers use only MN radio resources while
SCG (Secondary Cell Group) bearers use only SN
radio resources. Split bearers, however, can use
both MN and SN radio resources. The split bearers
transfer PDCP data between the SN and the MN
via the MN-SN user plane interface (X2/Xn).
While MR-DC is a general concept, the specific
terminology to denote MR-DC depends on the
deployment option considered, see also 3GPP TS
37.340 [14].
• EN-DC (E-UTRA-NR Dual Connectivity):
Option #3 (Non-standalone NR in EPS)
• NGEN-DC (NG-RAN E-UTRA-NR Dual
Connectivity): Option #7 (Non-standalone
NR in 5GS)
• NE-DC (NR-E-UTRA Dual Connectivity):
Option #4 (Non-standalone E-UTRA in 5GS)
For deployment option #3 (EN-DC with the EPC),
the master node is eNB and the secondary node is
en-gNB (i.e. gNB that is connected to EPC), where
eNB is connected to the EPC via S1 interface and
to the en-gNB via the X2 interface. In this
configuration, en-gNB may also be connected to
the EPC via the S1 interface (user plane) and to
other en-gNBs via X2 interface (user plane).
Hence in EN-DC:
• MCG-bearer uses only E-UTRA radio
resources
• SCG-bearer: uses only NR radio resources
• Split bearer can use both E-UTRA and NR
radio resources
For deployment option #7 (NGEN-DC with the
5GC), the master node is ng-eNB and the
secondary node is gNB, where ng-eNB is
connected to the 5GC via N3 interface and gNB is
connected to the ng-eNB via the Xn interface.
Hence in NGEN-DC
• MCG-bearer: uses only E-UTRA radio
resources
• SCG-bearer: uses only NR radio resources
• Split bearer: can use both E-UTRA and NR
radio resources
For deployment option #4 (NE-DC with the 5GC),
the master node is gNB and the secondary node is
ng-eNB, where gNB is connected to 5GC via N3
interface and ng-eNB is connected to the gNB via
the Xn interface. Hence in NE-DC
• MCG-bearer: uses only NR radio resources
• SCG-bearer uses only E-UTRA radio
resources
• Split bearer: can use both NR and E-UTRA
radio resourc
47
ROAD TO 5G: INTRODUCTION AND MIGRATION
Figure 22: User-plane connectivity for MR-DC
(Left: MN terminated; Right: SN terminated)
48
Part III:
9. Impact on voice
including service
continuity
ROAD TO 5G: INTRODUCTION AND MIGRATION
9.1 Options for operator voice/video
communications service in 5G
This section provides overview of Voice/Video
options in different deployment options.
Note: Emergency call is not considered but must
be supported as required by national regulation.
IMS voice and video call will be supported in all
5G deployment options described in this
document, however, there are differences
regarding which core network and which radio
access technology (RAT) is being used for media
and SIP signalling in each of the deployment
options. For Option 3 (E-UTRAN NR Dual
Connectivity), 4 (NR-E-UTRA Dual Connectivity)
and 7 (NG-RAN E-UTRA-NR Dual Connectivity)
where both NR and LTE can be used as RAT, the
decision on which RAT is used for SIP signalling
and media is made in the Radio Access Network
(RAN). In more detail, this means the following
two solutions are possible for IMS voice and video
call:
Solution 1: IMS voice or video call via 5GC,
media and SIP signalling using
• In Option 2: NR via 5GC
• In Option 4: NR and/or LTE via 5GC using
MCG, SCG and/or split bearer (controlled by
NG-RAN)
• In Option 5: LTE via 5GC
• In Option 7: LTE and/or NR via 5GC using
MCG, SCG and/or split bearer (controlled by
NG-RAN)
In addition, one or both of the following
functionalities may be supported in solution 1
during the call establishment
• RAT fallback from NR connected to 5GC in
Option 2/4 to E-UTRA connected to 5GC in
Option 5/7. Any ongoing data transfers
would be maintained using the procedures
defined by 3GPP Release 15 for intra 5GC
PS HO.
• EPS fallback from 5GS in Option 2/4/5/7 to
Option 1 or 3. Any ongoing data transfers
would be maintained e.g. if using a PS HO
with change from 5GC to EPC core
attachment using the procedures defined
by 3GPP Release 15.
50
Solution 2: IMS voice or video call via EPC,
media and SIP signalling
• In Option 3: LTE and/or NR via EPC using
MCG, SCG and/or split bearer (controlled by
E-UTRAN)
For networks that have not deployed IMS Voice
and video and have continued to use the Circuit
Switched Fallback (CSFB) solution defined in 3GPP
Release 8 and subsequent releases, the following
solution can apply.
Solution 3: CS voice via MSC
• In Option 3: CSFB to 2G/3G CS – Devices
currently attached to EPC may use this
solution to initiate or terminate a voice or
video call (the call is established on CS).
Note: CSFB from NR or LTE cell under 5GC is not
specified in 3GPP Release 15.
For an IMS voice and video call, the session
continuity during mobility can be ensured as
follows
• If the UE is using IMS voice or video call via
5GC, then
o PS HO from 5GC to EPC
o if two RANs are connected to 5GC
(one supporting option 2 or 4 and the
other Option 5 or 7): Intra 5GC PS HO
between NR and E-UTRA connected to
5GC
• If the UE is using IMS voice or video call via
EPC, then
o PS HO from EPC to 5GC
o Existing SRVCC to 2/3G (if required),
e.g., after loss of LTE radio coverage
After call completion, the terminal can either stay
on the currently used access / core or one of the
following procedures is possible
• If the UE is on 2G/3G after SRVCC, fast
return to EPS
• If the UE is on EPC, EPC>5GC PS HO
• If the UE is on 5GC, Intra 5GC PS HO
between NR and E-UTRA connected to 5GC
ROAD TO 5G: INTRODUCTION AND MIGRATION
9.2 Recommendations for voice/video
communications service in 5G
When evolving to 5G based on 5GC and if NR
coverage is continuous, then IMS voice and video
call on NR via 5GC (solution 1 with Option 2) is
most applicable, as cost of NR is much lower than
LTE and cell capacity of NR is much more than
LTE.
EVS codec is recommended as default voice codec
and HEVC is recommended as default video codec
for IMS voice and video call on NR capable
terminal for better QoS/QoE.
When evolving to 5G based on EPC, IMS voice
and video call on EPC (solution 2 with option 3) is
most applicable.
ROAD TO 5G: INTRODUCTION AND MIGRATION
Annex A
Other considerations
for operators
52
ROAD TO 5G: INTRODUCTION AND MIGRATION
A.1 Status Icon related to 5G
As 5G network may consist of elements from
evolved 4G networks (e.g., Evolved E-UTRA and
Evolved EPC), the decision on when 5G should be
displayed on the user interface of the terminal is
not as clear as previous generations. Operators
agree that it would be beneficial to have
consistency across device OEMs of the RAT
indicated to the user. Survey and discussion
among operator members of the 5G Introduction
project indicates that different requirements need
to be satisfied for deciding when to present a 5G
status icon depending on the operators. The 5G
Introduction project group sent the results of the
survey and discussion to 3GPP as a reply to the
LS from 3GPP SA2 for information. The liaison
statement can be accessed in the public in 3GPP
document repository (SA2: S2-178933 [17];
RAN2: R2-1713952 [18]).
53
GSMA HEAD OFFICE
Floor 2
The Walbrook Building
25 Walbrook
London EC4N 8AF
United Kingdom
Tel: +44 (0)20 7356 0600
Fax: +44 (0)20 7356 0601