5GPPP Brochure Web1 With Author Credits

User Manual: 5GPPP-brochure--web1-with-Author-credits

Open the PDF directly: View PDF PDF.
Page Count: 18

Download5GPPP-brochure 5GPPP-brochure--web1-with-Author-credits
Open PDF In BrowserView PDF
Supported by the

innovations
for new business
opportunities

CONTENT

01
02
03
04
2

BUSINESS AND STAKEHOLDERS ROLES
TRANSFORMATIONS WITH 5G

5

UNPRECEDENTED PERFORMANCE
FOR IMPROVED USER EXPERIENCE

7

5G NETWORK AS A SECURE, RELIABLE AND
FLEXIBLE ORCHESTRATION PLATFORM

11

RECOMMENDATIONS

15

Property of the 5G Infrastructure Association

Executive
summary

5G is the next generation mobile network that enables
innovation and supports progressive change across all
vertical industries and across our society1. Through its
Radio Access Network (RAN) design and its orchestrated
end-to-end architecture, it has the potential to
boost innovation and generate economic growth
in the European economy. The 5G service models
support agility and dynamicity, thereby impacting the
granularity, duration and trustworthiness of business
relationships. The ability to combine private and public
networks and data centres across multiple domains in
a secure and controlled way facilitates collaborative
business processes. It reshapes the digital business
ecosystem with new value chains linking stakeholders
from the telecommunications world and the vertical
industries in win-win situations. New stakeholders
emerge in this evolved ecosystem, for example cloud
companies and software houses that profit from the
cloudification and virtualization of the infrastructure,
and brokers that facilitate sharing of spectrum and
trading of connectivity and processing resources. Small
and medium-sized enterprises and start-ups are able to
embed 5G in their innovative products and services for
existing and new customers and markets, leveraging on
the Anything as a Service (XaaS) model.
These opportunities are conditioned by the ability of 5G
architecture and technologies to deliver the performance
levels required for vertical industry stakeholders to engage in
the 5G digital business ecosystem. This white paper highlights
the technological innovations of the first phase of the 5G
Public Private Partnership (5G PPP) and how they contribute
to the key performance targets for the 5G service classes:
enhanced Mobile Broadband (eMBB), Ultra-Reliable Low
Latency Communications (URLLC), and massive Machine Type
Communications (mMTC). The performance levels ensure
an unprecedented experience for end users including high
data rates, reduced end-to-end latency, massive connectivity,
ultra-reliability and support for very high mobility. The 5G PPP
innovations go far beyond what is announced for early 5G
deployments. For eMBB service, the integration of mm-wave
and frequencies below 6 GHz, along with ultra-dense networks
and nomadic nodes, ensure the targeted performance levels
with ubiquitous coverage and in high mobility scenarios, in
contrast with standalone deployments of mm-wave networks,
suitable for fixed usage. The innovations related to the
transport network allow also translating the peak throughputs
available at the air interface into perceived user experience
at affordable deployment cost for operators. In addition to
this, the envisioned 5G air interface serves simultaneously all
service classes (eMBB, mMTC and URLLC) in a cost effective
way, paving the way for new business opportunities with and
for verticals.

3

1

The 5G concept developed in this paper is not limited to the
RAN; it covers the end-to-end path and allows the 5G network
to act as a secure, reliable and flexible orchestration platform
across multiple domains. The 5G PPP innovations converge
towards the vision of 5G as a holistic orchestration platform
that integrates networking, computing and storage resources
into one programmable and unified infrastructure. The 5G
PPP innovations on multi-domain orchestration enable quick
end-to-end service deployment and dynamic sharing of
networking and processing resources among stakeholders.
The 5G security architecture is built on a baseline trust model
as a fundamental feature, and provides tools to analyse trust
and make it explicit in specific scenarios. The 5G architecture
ensures resilience of the network against attacks and its
availability during failure incidents. Availability and reliability
are achieved by mechanisms such as error recovery, fault
detection and fault resolution. These security, reliability and
flexibility properties, along with the multi-service air interface,
ensure that the 5G network is not just an enhanced air
interface as for pre-5G early deployments, announced for the
period 2018-2020, but also an open platform for new business
opportunities.
The architecture and protocols are designed to adapt to a
wide range of deployment scenarios including deep indoor,
hot spots, urban areas, rural areas, maritime areas and in an
aeronautical context. The 5G concept combines various access
technologies, such as cellular, wireless, satellite and wireline,
for delivering reliable performance for critical communications
and improve area coverage.
Standardization and spectrum regulation are critical
elements for avoiding fragmentation of future deployments
and increasing efficiency by eliminating redundant options.
Spectrum regulation must ensure the early availability of
a limited number of frequency bands, which eases the
development of the necessary equipment and facilitates faster
preparation of tests and trials. As of standardization, 5G PPP
projects contribute to 5G standards development by building
consensus among European industry, leading to individual
and concerted actions towards standardization bodies. In
contrast to early announcements of 5G deployments, the 5G
results are aligned with the standardization trends in 3GPP,
ensuring a global impact of European 5G innovations.
Once the first 5G standards are released and the frequency
bands are available, deployments of 5G networks will start,
adopting cost efficient upgrade paths building on existing 4G
infrastructure. Networking and processing resource sharing
strategies between stakeholders can be implemented for
delivering the performance targets, e.g. for URLLC use cases,
at an affordable cost. This resource sharing/integration is
enabled by the multi-domain orchestration advocated by 5G
PPP projects and aim at achieving win-win situations for all
the stakeholders involved in the 5G service. Regulation must
facilitate such flexibility in infrastructure sharing in order to
foster the development of the 5G digital business ecosystem.

https://5g-ppp.eu/wp-content/uploads/2015/02/5G-Vision-Brochure-v1.pdf

Property of the 5G Infrastructure Association

INTRODUCTION
5G is the next generation mobile network that enables innovation and supports progressive
change across all vertical industries and our society1. Through Radio Access Network (RAN)
design and an orchestrated end-to-end architecture, it has the potential to boost innovation
and generate economic growth across all verticals. 5G network deployments and market
evolution are subject to the technology achieving the performance targets that accelerate
adoption by vertical industries. This white paper highlights the technological innovations
developed in the 5G Public Private Partnership (5G PPP) program2 and how they help
reaching the key performance targets for the 5G service classes: enhanced Mobile Broadband
(eMBB), Ultra-Reliable and Low Latency Communications (URLLC), and massive Machine Type
Communications (mMTC)3. These performance levels ensure an unprecedented experience
for end users including high data rates, reduced end-to-end latency, massive connectivity,
ultra-reliability and support for very high mobility, ubiquitously. This white paper shows
how the 5G PPP innovations go beyond what is announced for early 5G deployments for the
eMBB service class, and how all 5G service classes are delivered over a scalable and cost
efficient network. It then explains how 5G technological innovations transform the network
into a secure, reliable and flexible orchestration platform across multiple technology and
administrative domains. Multi-domain orchestration allows a quick end-to-end service
deployment and a dynamic sharing of infrastructure resources among stakeholders, offering
new business opportunities and paving the way for new business models.
1

2
3

5G Vision – The 5G Infrastructure Public Private Partnership: the next generation of communication networks and
services https://5g-ppp.eu/wp-content/uploads/2015/02/5G-Vision-Brochure-v1.pdf
https://5g-ppp.eu/5g-ppp-phase-1-projects/
5G empowering vertical industries, https://5g-ppp.eu/wp-content/uploads/2016/02/BROCHURE_5PPP_BAT2_PL.pdf

15 PPP PHASE 1 Golden Nuggets
5G Spectrum Requirements,
Evaluation and Candidate
Bands

and Physical Architectures

5G Flexible RAN

5G Multi-Service

5G Massive Channel

5G Flexible Interference

5G Performance

Waveform

Access

Mitigation and RRM

Evaluation Framework

5G Integrated Transport

5G Network

Networks (FH/BH)

Management

Technology Enablers for
5G RAN Platforms
(HW & SW)

Flexible and Agile
Service Deployment

4

5G System, Functional, Logical

E2E Orchestration
in Single and MultiDomains 5G Virtualized
Networks

Network Softwarization
and Programmability
integrating SDN and
NFV Technologies

5G Networks Security

Programmable

and Integrity

Industrial Networks

Property of the 5G Infrastructure Association

01

BUSINESS AND STAKEHOLDERS ROLES
TRANSFORMATIONS WITH 5G

FIGURE 1

5G offers new business opportunities on a global level through
enhanced performance, flexibility and individualization. Compared
to previous generations of mobile networks the changes are
more radical. 5G technologies address today’s limitations and the
future capabilities, such as data rate, end-to-end latency, coverage,
softwarization, virtualization, network computing and promise to
create hyper-connectivity for delivering unprecedented services in
a secure and controlled way. The service levels are able to match
the different needs for the benefit of the individual end-customers
segments and vertical industries. 5G paves the way for innovative
business opportunities for exploiting multiple new Business-toConsumer (B2C), Business-to-Business (B2B) and Business-toGovernment (B2G) business models.

Stakeholders around 5G
END
CUSTOMERS

VERTICALS

NETWORK
OPERATORS

START-UPS
SMES

SERVICE
PROVIDERS

1.1
New value chains
in the 5G ecosystem
The 5G service models support agility and dynamicity far beyond
what is possible today, thereby impacting the granularity, duration
and trustworthiness of business relationships. Network Slicing – a
key concept of the 5G architecture – enables such capabilities and
allows Network Service Providers (NSP) to develop new offerings
using the Anything as a Service (XaaS) model, including IaaS
(Infrastructure as a Service), PaaS (Platform as a Service) and NaaS
(Network as a Service). The NSPs can flexibly allow the co-existence
of multiple tenants on their infrastructure. The tenants, who
represent a whole range of different vertical industry stakeholders
– called Online Service Providers (OSP) – offer products tailored to
the specific needs of their end users. The ability to combine private
and public networks and data centres across multiple domains in a
secure, controlled and provable way paves the way for collaborative
business processes. This enables flexible value chains and value
added services in a cost efficient way.
Table 1 provides an illustrative representation of the main
stakeholder roles that can schematically be identified, indicating
also how their business relationships could potentially evolve as the
5G ecosystem is being developed. In such an ecosystem 5G acts as a
catalyst for the development of new business relationships providing
opportunities for NSPs (06), a new generation of Communication
Service providers (07), Network Infrastructure Manufacturers (02),
IT Service Providers (05) and business customers including SMEs.
In this new model, partnerships are established across multiple
layers of services ranging from infrastructure sharing to network
capabilities.

TABLE 1

VENDORS

5

Stakeholder roles and interactions

Description

Roles and interactions

01

IT HW/SW equipment
manufacturers

Provide IT HW/SW equipment to IT service providers, Network
Infrastructure manufacturers, Network Service Providers and Online
Service Providers

02

Network Infrastructure
Manufacturers

Provide infrastructure to NSPs (06)

03

Software Network Function
Providers

Provide software network functions to NSPs (06), CSPs (07) and OSPs
(08)

04

Device Manufacturers

Provide devices to NSPs (06), OSPs (08) and end-customers (10)

05

IT service providers

Provide IT service to NSPs (06), CSPs (07) and OSPs (08)

06

Network Service Providers
(NSP)

Provide network services to CSPs (07) and OSPs (08) via Brokers (09), to
IT SPs (05), to other NSPs and to end-customers (10)

07

Communication Service
Providers (CSP)

Provide communication services to OSPs (8) and end-customers (10).

08

Online Service Providers
(OSP)

Provides on-line services to end-customers and can receive content
from them. Vertical stakeholders (automotive, energy, factories, health,
media) can take this role.

09

Brokers

Intermediary between OSPs and NSPs, and between NSPs in their effort
to dynamically establish the most effective solution meeting their needs

10

End-customers

B2C, B2B and B2G

Property of the 5G Infrastructure Association

The ability to combine private and public networks
and data centres across multiple domains in a
secure and controlled way facilitates collaborative
business processes.

1.2
Evolved roles for
network service providers

1.3
New business roles
and new actors

Main challenges and uncertainties in this changing ecosystem are
related to how the NSPs evolve their current business models, to
enable the offering of specialized services. Telecom operators are
currently facing several dilemmas with respect to business model
evolution, multi-stakeholder coordination, alignment of incentives,
regulation and competition4. A fruitful evolution of the telecom
operator oriented business models towards integration of verticals
in win-win partnerships is instrumental in bootstrapping and
enabling the ecosystem evolution. The major challenge for the NSP
is to deliver the needed level of service to a vertical (SSLA: Service
and Security Level Agreement), while keeping the control of its own
and whole infrastructure (sovereignty).

Vertical industry stakeholders’ involvement in the 5G value chain
marks the most important change compared to 4G. Stakeholders
from vertical industries such as automotive, energy, factories,
health, media, public transportation, aeronautics and other
sectors, can take the role of OSPs providing services directly to endcustomers on top of the infrastructure and connectivity services of
NSPs. Manufacturing companies producing vertical industry specific
equipment may play the role of device manufacturer.

Changes are induced in the relationships between NSPs, content
providers and content delivery providers. The changes enable new
service experiences such as immersive media or health services,
enabled by orchestrating, controlling, using and monitoring
infrastructure resources in an end-to-end coordinated approach.
An NSP may negotiate specific wholesale agreements with content
providers and content delivery providers for the provision of
services allowing greater customer choice and control. In this same
context, these agreements can extend to the deployment of innetwork content caches, thus enhancing the user experience while
mastering network and cache deployment costs.
Revenues for Business-to-Consumer mobile data services in
combination with revenues from wholesale relationships will
increase when the new services and products are deployed across
the value chains. The goal is to create value by detecting new
demand for services, enriched by digital platforms, addressing new
consumer and business needs.

4

Service Level Awareness and open multi-service internetworking,
http://www.networld2020.eu/wp-content/uploads/2016/07/NetWorld2020_
WP_Service-Level-Awareness_Final_June-16.pdf

Property of the 5G Infrastructure Association

The introduction of the cloud computing model into the telecom
industry enables the emergence of new stakeholders from the IT
world (5) into the 5G ecosystem. IT Service Providers and Network
Functions Providers can deliver new services such as cyber security
services or big data analytics to other stakeholders in the ecosystem
such as NSPs, OSPs or directly to end-customers.
The ecosystem and regulatory evolution enables new business role,
such as the Broker (9) role that offers services to help its customers
be more effective. A broker acts as intermediary between OSPs
(including verticals) and NSPs, and between NSPs, in their effort
to dynamically establish the most effective solution meeting
their needs. Among these brokers we can cite spectrum brokers
that facilitate spectrum sharing between NSPs, and connectivity/
processing brokers that, e.g., manage marketplaces for trading
connectivity and processing resources between NSPs and cloud
providers for the purpose of setting up end-to-end services with
guaranteed Service Level Agreements (SLAs). Brokers may also
act as intermediaries between end-customers and NSPs/OSPs
providing services with similar characteristics as for the wholesale
market.
To sum up, new business opportunities emerge for telecom/network
operators, manufacturers and solution providers as well as for a
range of new stakeholders such as OSPs, software houses, brokers,
start-ups and SMEs that use 5G for creating innovative products and
services for existing and new customers and markets, leveraging on
the XaaS model. These opportunities are conditioned by the ability
of 5G technologies to provide the targeted performance levels that
convince vertical stakeholders and allow the creation of this new
dynamic ecosystem around 5G networks.

6

02

UNPRECEDENTED PERFORMANCE
FOR IMPROVED USER EXPERIENCE

2.1
A flexible 5G RAN design
The promise of 5G starts with the ability to offer an unprecedented
experience for end users, in both Business-to-Business and
Business-to-Consumer relationships. This includes extremely high
data rates, very low latency for devices and support of very high
mobility speeds and massive connectivity. This section describes
how flexible 5G RAN design, innovations on the physical layer, radio
resource management and protocol design, enable reaching these
targets.

FIGURE 2

Due to the diverse requirements of the 5G service classes, the
5G RAN is designed to operate over a wide range of spectrum
bands, from 700 MHz to tens of GHz, with diverse characteristics,
such as channel bandwidth and propagation conditions. The 5G
RAN allows integrating Long-Term Evolution Advanced (LTE-A)
technologies, novel 5G radio and WiFi evolutions. Multi-antenna
concepts and advanced multiuser detection techniques are
integrated at system level and help achieving extremely high data
rates across the coverage area. Novel densification strategies
lead to the deployment of ultra-dense networks, with fixed small
cells and nomadic nodes, such as relays mounted on vehicles.
Further capabilities are the native and efficient support of multiconnectivity, Vehicle-to-Anything (V2X), network-controlled Deviceto-Device (D2D) and satellite communication. The 5G RAN supports
a wide range of physical deployments, from distributed base
stations to centralized cloud-RAN deployments or distributed edge
clouds. Self-backhauling is an important feature, in which devices
act as base stations and self-establish wireless backhaul links to
suitable donor base stations [METIS-II]. Support for heterogeneous
backhaul technologies is important to maximise 5G availability,
resilience and coverage.

Illustration of the 5G RAN

Integration of spectrum below 6 GHz with mm-wave
bands results in an 11-fold cell edge throughput
improvement compared to 4G. A 100-fold area
capacity increase can be obtained by using small
cells and applying an flexible spectrum management
framework.

2.2
Extremely high data rates
The high data-rate requirements for 5G call for a substantially larger
amount of spectrum, higher spectral efficiency, and significantly
denser deployments of base stations. A particular challenge is the
inhomogeneous distribution of traffic over time and space. This
requires the network to react quickly and dynamically to fulfil the
increased service demand during a time period at a certain region.
These requirements are addressed by a set of key innovations
proposed for the 5G Radio Access Network (RAN), covering
spectrum usage, flexible Radio Resource Management (RRM),
ultra-dense network deployments and highly efficient and low cost
transport network.
New spectrum will be available for 5G, including in the millimetre
wave (mm-wave) band. The joint usage of high and low spectrum
frequencies combines the benefit of higher bandwidth and
beam-forming capabilities to increase area capacity using higher
frequencies, while maintaining good area coverage capabilities
using low frequencies. The flexible and dynamic resource allocation
between different radio accesses enables offloading functionality
in high resource demand situations. As a result of better resource
utilization, the cell-edge user throughput is improved, for example
from 15 Mbps to 170 Mbps – an 11-fold improvement compared
to baseline LTE [mmMAGIC]. The availability of new spectrum is
accompanied by a flexible spectrum management framework that
applies new spectrum access schemes in addition to conventional
spectrum management methods. For example, License Assisted
Access (LAA) and License Shared Access (LSA) can provide more
spectrum capacity to end users [Coherent]. A spectrum manager
allocates then radio spectrum and energy according to the needs
of the traffic, considering the conditions of the surrounding radio
environment and the available resources, yielding a 100-fold area
capacity increase over current state of the art of LTE small cells
[SPEED-5G].
A further innovation that increases data rates is the 5G flexible RRM
framework that steers base-stations to use particular channels and
spectrum, makes use of dynamic Time Division Duplex (TDD), and
allows coordinated operation of access points. Dynamic TDD is
the flexible use of spectrum for uplink or downlink based on the
instantaneous requirements of burst traffic. In order to contest
cross-interference among base stations and among devices,
advanced interference coordination ultimately yields an estimated
60 % gain in uplink performance for dynamic TDD compared to

7

Property of the 5G Infrastructure Association

baseline, static TDD. Combined with efficient cancelation of very
strong cross-interference an additional 50 % gain is obtained
[METIS-II]. The flexible RRM framework includes support for finegrain radio resource allocation and node cooperation across the
cell boundaries, improving the performance of coordinated multipoint transmission and reception (CoMP), multi-connectivity and
other cooperative functions in the RAN [Coherent]. Aggregating the
bandwidth of several network access technologies (cellular, WiFi,
satellite and wireline) through traffic splitting/steering techniques
is also part of this flexible RRM and helps providing the desired
quality of service level throughout the coverage area. These RRM
techniques are facilitated by the development of resource pooling
techniques (centralized RAN), allowing centralization of RRM
decisions. A new architectural paradigm, called Virtual RAN (VRAN)
can ensure hardware re-configurability through the utilization of
open source platforms and general purpose hardware, thereby
extending radio resource allocation flexibility to the processing
resources [Flex5GWare].
A further complementary advancement is the deployment of ultradense networks, including intelligent small cell nodes and the use
of nomadic nodes. Intelligent small cell nodes have the ability to
make local decisions on the usage of resources or spectrum bands,
improving scalability and reducing signalling loads in ultra-dense
environments. The flexible RRM optimises then the deployment
of neighbour cells and matches the wireless link requirements
[Speed5G]. The 5G infrastructure can integrate complementary
nomadic access nodes that increase network capacity and extend
the coverage area. For example, by selecting a nomadic node out
of 20 randomly distributed nomadic nodes in a macrocell closer
to a hotspot, the downlink 10 percentile user throughputs can be
improved by around 150% relative to a fixed picocell [METIS-II].
Integration of cellular with satellite solutions ensures the service
continuity for certain moving nodes. Note that the increase of data
rates through network densification may increase network energy
consumption. Therefore it is essential to anticipate the dynamic
activation and deactivation of network nodes, such as small cells,
to improve the overall energy efficiency of the network. Centralized
traffic scheduling and multi-cell coordination schemes, such as
dynamic point selection and joint transmission can be used to
operate an optimal number of nodes in the network. Relative to
no multi-cell coordination, the network power consumption can be
reduced by 50% at low user generated traffic [METIS-II].
In analogy to the innovations on the radio interface, the transport
network, for backhaul and fronthaul applications, requires much
higher capacity. To meet higher bandwidth demand, new functional
splitting models are defined, leveraging the distribution of radio
functions between radio unit nodes and centralized processing
nodes. A single efficient transport network has been designed that
is able to support backhaul and fronthaul at the higher data rates
at a lower cost [5G Crosshaul]. For example, a 200 MHz frequency
band, using 256 QAM (Quadrature Amplitude Modulation), and 8x8
beamforming is capable of generating a peak throughput at the radio
splitting interface of tens of Gbit/s, resulting in an aggregate traffic
of several hundreds of Gbit/s. This traffic volume is only sustainable
through advance optical transmission technologies and a flexible
high-speed low-cost transmission using a passive wavelengthdivision multiplexed network (WDM-PON) was demonstrated
[5G-XHaul]. High levels of centralization maximizes the advantages
of pooling for transport and processing nodes, while an efficient use
of spectrum increases the amount of available bandwidth per user.
A trade-off between spectral efficiency and transport optimization
resulting from centralization of processing must be found. A novel
application for managing energy consumption has been defined,
which allows switching transport nodes on or off depending on the
activation or deactivation of radio nodes [5G Crosshaul].
While these technologies are tailored for the eMBB service class,
they go beyond what is announced for early 5G deployments. The
integration of mm-wave and frequencies below 6 GHz, along with
nomadic nodes ensures a ubiquitous coverage with high data rates,
in contrast to standalone deployments of mm-wave networks. The
innovations related to the transport network ensure that the peak
throughputs available at the radio interface translate into perceived
user experience and affordable deployment cost for operators.

Property of the 5G Infrastructure Association

2.3
A very low end-to-end
latency for time-critical services
Low end-to-end latency is a primary requirement driving the
5G development. Many critical use cases, such as Augmented
Reality, Precision Medicine and remote assisted robotic surgery
in Health, road safety and autonomous driving in connected
vehicles, factory automation, etc., require very low response times
in the communication between the respective parties. The delay
targets range from tens of milliseconds to 1 millisecond. Network
conditions, computing load and congestion induce variability in the
end-to-end latency. We show in the following how the latency can
be reduced by an optimization of the radio access, the backhaul but
also the processing time that incurs in providing the specific service
(including the availability of data).

A latency as low as 1 ms can be obtained on the air
interface by applying a new frame structure with the
possibility of using short TTI and advanced HARQ. Mobile
Edge Computing reduces end-to-end latency for critical
services.

A low latency air interface
New waveforms that are robust against time-offset do not require
signalling associated with time alignment, allowing for reduced
latency. A flexible frame design allows multiplexing of Transmission
Time Intervals (TTI) with different lengths on the same spectrum
resources, reducing latency for short critical messages [FANTASTIC5G]. Furthermore, advanced Hybrid Automatic Repeat Request
(HARQ) schemes have been developed to reduce latency on the
air interface based on early detection of packet error without full
decoding, or estimation of the number of re-transmissions needed
for successful decoding [mmMAGIC]. Multi-mode connectivity, a
concept that allows a mobile user to be connected to multiple access
nodes at the same time, introduces also spatial diversity allowing
information transmission via the shortest path [FANTASTIC-5G].
A further key enabler for reduced latency is to improve control plane
procedures and related signalling [METIS-II]. For example, a critical
requirement in 5G is that an inactive device with the sudden need to
send mission-critical data can quickly access the system, establish
a Radio Resource Control (RRC) connection, and send the data. One
proposed innovation in this respect is to allow differentiating critical
from less critical services through a novel form of preamble use of
the Random Access Channel (RACH), which allows increasing the
reliability of mission-critical system access at the first attempt by
two orders of magnitude, which results in improved system access
latency. In addition, a new RRC state is proposed, which allows
temporarily inactive devices and the infrastructure to maintain
the context information related to a previous RRC connection,
for example the security. With this new state, devices can switch
about 4-10 times faster from inactive to active state. The exact
gains depend on the radio and core network signalling latency.
Similar gains are estimated for the downlink. The new state helps
to substantially reduce the delay between the paging request to an
inactive device and the actual data transmission to the device.

8

When operating at mm-wave frequencies the objective is to
concurrently fulfil high throughput and low latency requirements.
mm-wave allows the use of short TTIs, which leads to inherently
low latency from a frame structure point of view. Responding to
the challenge of processing large amount of data in a short time,
high throughput and robust decoders for Low-Density Parity-Check
(LDPC) codes to reduce decoding latency are being implemented
[mmMAGIC]. The time required for initial access when a user joins
the network or performs handover involves a procedure with
considerable overhead. Fast beam-alignment schemes that exploit
advanced beam codebook design, context information and multinode coordination have been proposed to reduce the initial access
latency [mmMAGIC].

A new architecture that reduces
the end-to-end latency

FIGURE 3

The communication delay caused by the physical distance between
the source and the destination is reduced by the introduction of
Mobile Edge Computing (MEC) in 5G networks. MEC offers additional
processing capacity near the base station for local application level
processing. The distance herein refers to the end-to-end network
path and rarely refers to the geographic proximity. Decreasing
the number of entities along the network path is an established
technique to significantly reduce latency. The integration of
distributed cloud resources with a cluster of small cells at the
network edge assists in achieving low latencies by removing the
overhead of backhaul to the core network and enables services at
the network edge [SESAME].

Intelligent function placement for customized latency
BASE STATION

NF1

EDGE CLOUD

NF1
NF2

ULTRA LOW LATENCY

CENTRAL CLOUD

NF3

NF1
NF2

LOW LATENCY

NF2

HIGH LATENCY

A further enabler for low latency is flexible network architecture,
where the various networking functions, such as PHY algorithms,
scheduling, HARQ, handover, routing and others, execute in a
central or in an edge cloud hardware depending on the specific
service requirements. A low-latency service is provided by network
functions belonging to both radio access and core network, where
the most latency critical functions are moved from the central cloud
to the edge cloud, to avoid backhaul latency and unnecessary hops
along the path to the central cloud. A service creation request is
mapped to network functions that are realised in the instantiation
of a suitable network slice configuration taking into account the
service requirements. For example, the most critical network
functions are moved from the core network into the edge cloud,
close to the user. For even more stringent latency requirements, the
network functions can be placed on an edge cloud located directly
in the base station [NORMA].

A low latency processing in the hardware/
software platforms
In addition to the networking delay between the source and
destination, the telecommunication hardware itself introduces
latency [Flex5GWare]. The dynamic reconfiguration of hardware
and software platforms is being developed to reduce such delays.
A hardware-agnostic mechanism guides the reconfiguration
of the underlying heterogeneous hardware infrastructure and
‘stitches’ software and hardware components at runtime, removing
repetition of functionality among multiple layers, technologies
and elements thereby offering a flexible, reprogrammable and
reconfigurable functional composition. A cognitive dynamic
hardware and software partitioning contributes to the reduction of
the overall latency via choosing configurations that balance latency
against other key metrics, such as energy efficiency. This solution
considers the latency incurred by all components implemented in

9

either hardware or software, communication between components
and external interfaces.

In-network caching as an enabler for low
latency communications
Video and augmented reality applications from the media and
entertainment vertical industry require low latency coupled with
large data rates. A subset of these applications involves videos
that are not generated online and can profit from efficient content
caching techniques near the end user, reducing the latency
associated with content retrieval from remote locations. 5G PPP
projects are working towards a flexible content caching solution
relying on Network Functions Virtualisation (NFV) and a Software
Defined Network (SDN) enabled traffic optimization. In addition
extending the cache of the operators to a large virtual cache spanning
multiple operator networks is enabled by inter-slice, multi-tenant
cache peering technologies and supplements the advantage of
caching in the edge cloud. A related innovation reduces directly the
latency introduced by network nodes using a special optimization of
IPv6 addresses to speed-up the routing process. In a hierarchically
structured network the IPv6 address prefix is evaluated gradually
by the routers, eliminating the need for a routing lookup-table,
which is the most time expensive operation in IP-routing. Cache
efficiency is increased by mechanisms for content pre-fetching at
the edge nodes including base stations, small cells, nomadic nodes
and even User equipment (UE), using context-aware information.
Content prefetching and cache updates may be performed using
multicast/broadcast capabilities of terrestrial networks and
satellites, alleviating the signalling and traffic loads in the middle
and last mile [Charisma].
The latency-reduction techniques are complementary and may be
used to meet the requirements of the targeted service. For example,
when considering tele-operation of moving robots with haptic
feedback, the latency reduction techniques are applied for handling
control and haptic feedback signals on the radio interface and along
the path between the two end points, while fog computing near the
robots is used for image processing and local control algorithms.
Many of the technologies used to increase the data rate, contribute
to the reduction of latency. Flexible RRM allows multiplexing timecritical services with other services, ultra-dense networks ensure
the availability of a close access point and RAT integration and
multi-connectivity allow reaching the destination by the shortest
path. However, centralization of processing – for interference
management and resource pooling – may be conflicting with the
need of performing application-level processing near the end user.
A hybrid and flexible architecture where processing is centralized
for some services and distributed for others is needed.

2.4
A massive connectivity of devices
The commercialization and deployment of 5G systems is driven
by the need to support very high connection densities to make
the Internet of Things serviceable. Connection density is defined
as the average number of simultaneously active connections that
can be supported in a given area, measured in connections per
square kilometre. Example use cases include crowded spaces,
i.e. stadiums or conference venues, as well as massive MTC in
cities, agriculture or factories, etc., where sensors, actuators, and
controllers are wirelessly inter-connected. Massive connectivity is
supported by new air interfaces that should optimise the available
radio and infrastructure resources, spanning areas from protocol
enhancements and radio resource management to waveform
design [mmMAGIC, FANTASTIC-5G, METIS-II].
One of the enablers for massive connectivity is the handling
of the transition of UE modes. Transitions between idle mode
and connected mode must be simplified or even avoided. In this
direction, connection-less transmission of small packets from
UEs once registered and authenticated in the network reduce
significantly the required number of signalling messages. In
this case, the (small) packet must comprise both source and
destination addresses. A component contributing to the reduction
of signalling upon connection establishment is the addition of new
Radio Resource Control (RRC) states such as the RRC extant state

Property of the 5G Infrastructure Association

[FANTASTIC-5G, METIS-II], which is a hybrid state between RRC idle
and RRC connected states. From RRC idle it inherits its behaviour,
i.e. UE controlled mobility and from RRC connected it retains most
of the UE specific access stratum information.
A further innovation is related to the channel access for sensors.
Classical channel access protocols comprise two stages: the access
notification stage and the data delivery stage. This leaves room for
feedback and resource allocation to the UE from the base station,
for instance, related to power control and timing alignment.
New, “one-stage” access protocols are being developed, in which
access notification and data delivery are performed in a single
transaction by means of one or more consecutive packets or in a
single transmission thereby reducing signalling overhead for short
messages [FANTASTIC-5G]. Although initially designed to meet very
low latency requirements, the “one stage” protocol is also of interest
for longer latency channels like satellite links since it minimises the
handshakes.
Additionally, new random access schemes are proposed, where,
each device is allowed to contend with a predefined sequence
of preambles over multiple Physical Random Access Channels
(PRACHs), denoted as the device’s signature [FANTASTIC-5G]. This
signature is constructed from information unique to each device,
such as the device’s identity. The proposed schemes result in a
significant reduction of message exchanges in the access protocol
and can be complemented by appropriate collision resolution
techniques, exploiting sparsity properties.
A new waveform design is proposed for asynchronous small
packet transmissions in the uplink [FANTASTIC-5G]. Because of the
superior spectral properties of certain waveforms, the need for
tight temporal synchronization of users can be relaxed. This allows
compressing or even avoiding broadcast messages, thus leading to
energy and radio resource savings.
Finally, interoperability between the aforementioned innovations
supporting the mMTC service class and other solutions for data rate
increase or latency reduction is ensured by considering a flexible
frame structure, allowing multiplexing of short and long TTIs and
coexistence of different waveforms on the same frequency band
[FANTASTIC-5G].

2.5
A high performance in
high mobility scenarios
Over the past decades, digital communication technologies
introduced significant advances in mobility support, such as
seamless handover between cells and efficient coverage of
highways and railways. Challenges remain for very high mobility
scenarios and for supporting applications such as road safety,
assisted driving, autonomous driving as well as other business and
infotainment applications.

New waveform designs allow resistance to Doppler
effect and enable perfect transmission for vehicle
speeds up to 600 Km/h.
Robustness to Doppler Effect – a must have
for high mobility scenarios
Fast moving mobile nodes suffer from the Doppler shift and spread.
The latter leads to severe inter-carrier interference in 4G and the
network access for Intelligent Transport Systems operating at 5GHz
(ITS-G5), because these systems rely on the orthogonality property
of subcarriers in the orthogonal frequency-division multiplexing
scheme. This interference induces reception errors and imposes
retransmissions that lead to increased latency and require the use
of robust modulation and coding schemes with reduced spectral
efficiency. The development of new waveforms that provide better
spectral containment of the signal power, reduce the effects of
inter-carrier interference induced by the Doppler effect [FANTASTIC-

Property of the 5G Infrastructure Association

5G]. 5G is being designed with a flexible configuration of radio
resources in the time and frequency dimensions, allowing for larger
subcarrier spacing in sub-bands, further alleviating the impact of
inter-carrier interference. Initial demonstrations [FANTASTIC-5G]
provide evidence of the high potential of the proposed solution. In
the case of an uplink transmission of an image at a vehicle speed
of 600 km/h and using the proposed new waveforms the image
was perfectly transmitted, while using current 4G technology,
inter-carrier interference caused signal corruption that results in a
failed transmission. These results confirm that new waveforms can
improve transmission quality such that highly efficient V2X (vehicleto-everything) communication is enabled.

Seamless handover and multi-connectivity
– zero interruption time and increased
diversity
High mobility induces frequent network attachment procedures
that may cause annoying service interruptions. High capacity
backhaul and the concept of synchronization of base stations enable
synchronous and random access-less handovers. Base stations
agree on the time a handover will take place and the mobile user
receives an un-interrupted service that is seamlessly transferred
from the source to the target cell at the agreed handover time
without a new network attachment procedure [FANTASTIC-5G].
The generalisation of the multi-mode connectivity concept
introduced above enables simultaneous connectivity of a vehicle
to several base stations at the infrastructure as well as to other
vehicles, allowing for robust and seamless handover [FANTASTIC5G].
The Central Controller and Coordinator (C3) is a logical entity in
charge of centralised network-wide or large area-wide control and
coordination among entities in the Radio Access Network (RAN) –
possibly using different Radio Access Technologies (RATs) – based
on centralised network view [Coherent]. It facilitates the handover
procedure by programmatically acting reactively or proactively
on user mobility events. The proposed seamless handover, multimode connectivity and UE-relaying can easily be implemented and
managed by the C3. The handover decision is based on knowledge
retrieved by the network graphs and hence is more efficient. The
centralized network view allows considering several metrics in the
handover decision, such as received signal strength, interference
level, network load and vehicle speed.

2.6
A UE positioning accuracy
in the range of sub-meter
The new RAN design allows achieving a new kind of performance
indicator, that is UE positioning. A positioning accuracy in the order
of sub-meter is critical for a plethora of Location-Based Services
(LBSs), as input for data analytics, to improve public safety in
emergency scenarios, to introduce services like collision avoidance
in autonomous vehicles and to create new customized experiences
and services for the end user. The target sub-meter range is clearly
beyond what can be achieved with existing systems. 5G RAN has
a few key enablers that can be seized by positioning systems,
including in indoor and urban canyon areas:
The high density of access nodes leads to a high probability
of Line-of-Sight (LoS) conditions between access nodes and
the UE.
The large channel bandwidth available in 5G networks
enables increased positioning accuracy for methods
based on time-of-flight (ToF) distance measurements
[Flex5GWare].
The availability of multiple antennas allows the use of
beamforming techniques for positioning for Angle-ofArrival (AoA)-based techniques, besides minimizing the
multipath propagation effects for ToF-based ranging
methods [5G-XHaul].
The envisioned integration of LTE-A evolution, with novel 5G radio,
WiFi evolution and other technologies such as Global Navigation
Satellite System (GNSS) positioning is a key to achieve better
positioning accuracy.

10

03

5G NETWORK AS A SECURE, RELIABLE AND
FLEXIBLE ORCHESTRATION PLATFORM

3.1
A flexible architecture that integrates
natively networking, computing and
storage resources
5G is a holistic orchestration platform that integrates networking,
computing and storage resources into one programmable and
unified infrastructure. This vision requires a flexible multi-tenant
architecture where computing resources are distributed within the
network including sites of the vertical industry stakeholders, within
the base stations, in edge clouds at central offices, in regional
and central clouds, and managed by different stakeholders. The
new architectural paradigm will need to support heterogeneous
hardware resources in the same holistic vision: custom ASICs
(Application-Specific Integrated Circuit) are still the way to implement
fast network processing, and heterogeneous components such
as FPGAs (Field-Programmable Gate Array), GPUs (Graphics
Processing Unit) can be added to commodity hardware machines
to dramatically increase the performance of certain workloads.
Therefore, the Virtual Network Function (VNF) concept defined in
current NFV architecture should be extended to support a more
granular decomposition of the functionality and the mapping into
heterogeneous hardware execution platforms.

A new 5G security architecture includes a baseline
trust model as a fundamental feature and
addresses also security on an end-to-end fashion,
to support applications that require coordination
across multiple domains.

3.2
A secure and
trustworthy network
The security architectures for current 3G/4G networks are defined
by 3GPP5. However, there are few main drivers that are pushing for
a new security architecture for 5G:
Growing threat levels and increasing abuse of peripheral
devices for attacks on back ends and entire networks (for
example through DDOS attacks);

The extremely high data rate capabilities combined with the
availability of storage and processing resources in the edge nodes
allows local processing and data analytics close to the users or within
the vertical industry stakeholder premises, offering advantages
such as low latency, security and confidentiality. Processing capacity
in the edge nodes allows software to be transferred and executed
near the data (Software to Data, S2D).

Introduction of new technologies such as virtualization,
edge-computing, fog-computing, in network data and
video caching and purpose-specific hardware;

FIGURE 4

A flexible architecture integrating networking,
processing and storage resources.

Partial lack of coverage of management aspects;
Absence of an explicit and complete trust model for 3G and
4G networks; and
New business models involving more complex trust
relationships enabled by open access via network slicing.

RADIO ACCESS
NETWORKS

FIXED & WIRELESS ACCESS
NETWORKS

AGGREGATION & CORE
NETWORKS

EDGE
CLOUDS

RADIOS

NETWORK
CLOUDS

S2D

S2D

FRONT/BACKhaul
NETWORK

Optical access
NETWORK

Optical METRO
NETWORK

Mobile Edge Computing

11

Performance levels for supporting safety and security
of critical services in different vertical domains such as
eHealth, transportation and industrial automation;

Optical core
NETWORK

core
CLOUDS

Trusted and Trustworthy 5G Architecture
Without a well-defined trust and governance model, it is unclear
which stakeholders have what responsibilities and liabilities in
the new business ecosystem. The current ad-hoc approach works
well for a small number of network operators. The proliferation
of operators is already causing concerns, making network
infrastructure more open and posing risks such as impersonation
on signalling interchange networks. In 5G networks these problems
will become more significant, as will the possibility that security
issues are not addressed by any stakeholder, creating opportunities
for attacks.
The security architecture being developed in 5G PPP will extend and
influence the 3GPP security standards and architecture to capture
virtualization and network slicing aspects as well as to include
a baseline trust model as a fundamental feature [5G-ENSURE].
While some stakeholders and trust relationships exist in every
5G application, there is no one trust model to fit all situations.
Trust and trustworthiness must be geared to the needs of vertical

5

TS 33.401, 3GPP System Architecture Evolution (SAE); Security architecture.

Property of the 5G Infrastructure Association

applications, taking account of the business stakeholders and
relationships involved, and the level of risk for each stakeholder
from network-related security threats. In the future, vertical
industry consortia must agree trust models to meet business
and regulatory requirements in a range of critical and non-critical
applications including scenarios that cross borders or span multiple
physical infrastructure domains.
To support this security architecture, a security management
framework that relies on autonomic network management is
proposed. Autonomic management solutions leverage insights
from real-time analytics, and actuation on network resources in
real-time to minimize or prevent the effects of detected threats
[CHARISMA].

Multi-Domain and Multi-Layer Security
5G networks rely on network virtualization, implemented by VNFs,
forming network slices executed on a shared infrastructure. The 5G
security architecture must follow the design principles of the overall
5G architecture, and needs to be logical rather than physical. Slicing
must isolate resources and data on shared infrastructure.
Thereby, 5G security must address threats in an end-to-end fashion,
to support applications that require coordination across multiple
domains. This requires monitoring security in a cross-domain
fashion, between physical domains and layered virtual domains. A
proposed solution is to use a hierarchical management architecture
matched to the trust network, allowing monitoring and control
between mutually trusting and trustworthy stakeholders [Selfnet].

Security as a service
5G networks support new business models, enhanced connectivity
services and enriched network functionality based on a combination
of network operator and vertical industry stakeholder assets and
capabilities. In many new business models, the role of a virtual
network operator may be fulfilled by a vertical industry-focused
organisation such as a manufacturer or a health care provider. Such
organisations may not have the capability of managing network
security, and may not want to invest in acquiring expertise in areas
outside their core business. Security services are therefore needed
in conjunction with virtualised network provisioning services to
support virtual network operators manage their networks.
Key requirements for security services have been identified
and enablers are being developed to meet these requirements
[5G-ENSURE]:
Trust enablers to provide users with possibly certified
information about trust dependencies and trustworthiness
of stakeholders and technology components;
Enablers to support authentication, access control and
accountability, such as group authentication of Internet of
Things (IoT) devices;
Privacy enablers to improve subscriber identity protection,
during network connection and authentication procedures;

3.3
A Reliable and
resilient network
Availability is related to the service coverage and is defined as
the probability that a service request is accepted with the target
Quality of Service (QoS). Reliability is the capability of the system to
offer a continuous and consistent service quality while operating
in dynamic conditions. Quality is associated with performance
indicators that depend on the targeted service, such as cell edge
data rate for eMBB and latency and packet error rate for critical IoT.
In a reliable network the target performance indicator must be met
using mechanisms that are intrinsically unlikely to fail, for example
redundancy or error correction mechanisms.
Network resilience refers to the ability of a network to recover from
harm caused by an event or situation that degrades network QoS.
Reliability measures help preventing some types of threats, while
resilience measures help recovering from threats that could not be
prevented. Network security measures may also be considered in
this way, given that QoS in a 5G network also covers confidentiality
and integrity characteristics as well as availability and performance.
The use of 5G networks in safety and security critical applications
means that threats may have a far higher impact. Measures to
prevent or mitigate such threats are of paramount importance,
to ensure the effects do not adversely impact critical applications.
Measures to ensure a reliable and resilient network for critical
applications have to cover networking aspects, such as ensuring a
high reliability and availability of the connection [FANTASTIC-5G],
and security aspects, such as ensuring isolation between network
slices [CHARISMA].

Availability and reliability on the air
interface
On the radio interface, reliability and availability must be ensured by
means of ubiquitous coverage and error free transmission. A set of
innovations on the radio interface is proposed that ensures meeting
this target. Coverage extension is ensured by a new control channel
design, exploiting beamforming for control channels. To support low
end devices in remote areas, asymmetric link operation is proposed
in which cell edge devices can use long transmissions for an
appropriate signal decoding. To enable robust communication with
low latency, advanced error correction and recovery mechanisms
are proposed. On a system level, 5G utilises multi-connectivity, in
which messages can be sent simultaneously over several radio links
allowing spatial diversity and possibly combined at the receiver for
a robust decoding [FANTASTIC-5G]. Satellite-terrestrial network
integration provides a relevant contribution to the overall set of
deployment options and helps increasing network availability.
Combining several technologies at the radio interface and at the
backhaul may be needed to improve the service reliability especially
to support mission critical applications.

Security monitoring enablers to detect security breaches
including malicious or compromised devices using network
function and traffic monitoring and analysis;
Network management and virtualization enablers such as
platform integrity attestation.
Such security enablers must be supplied as commodity components
and must be able to handle the scalability and diversity of virtual
networks needed by vertical industry.

Property of the 5G Infrastructure Association

12

Reliability in virtualized networks
The reliability of end-to-end services is hard to assess in an NFV
environment in which network functions are dynamically deployed
and share the same hardware infrastructure during their execution.
This is due to the multi-layer dependencies introduced by the usage
of a common infrastructure and virtualization platforms by multiple
network functions which splits the management concerns in
multiple and transparent to each other layers. Although managed
separately all these layers influence the reliability of the end-toend service. The addition of the virtualization layer between the
hardware and the software that acts as a broker for available
resources results in a less stable infrastructure compared to the
previously used physical network equipment infrastructure.
To fully benefit from system dynamicity and elasticity in deployments
over virtualised infrastructures, a reliability framework – based on
machine learning – is proposed that works in two main directions
[Cognet]:
Anomaly detection of patterns in data that do not conform
to expected behaviour and enablement of system
adaptation to unforeseen conditions. Anomaly detection
employs a semi-supervised learning approach, which
constructs a model of normal behaviour from a training
dataset representing normal operation. The deviations
from the normal behaviour are used to detect potential
anomalies.
Fault detection, isolation and resolution of network
malfunctions. This function correlates the number of
detected incidents and identifies the fault for the observed
malfunctions. The fault removal actions defined in the
policy specification resolve the identified errors and
failures.

Resilience against security threats
Malicious attackers and associated threats are inherently present
in any system designed for mass use. We cannot assume that all
subscribers are trustworthy – some try to cheat the system or have
other motives to degrade it. The network must deploy measures
to detect and prevent security intrusions fast, so that the number
of intruders in the network remains small. Furthermore it must be
intrusion tolerant, which means that it should degrade gracefully
rather than abruptly or catastrophically in the presence of
intrusions.
Isolation between network slices is an important property in
delivering resilience and intrusion tolerance. The proposed security
management solution manages interference across multiple slices
and allows preventing tenant-on-tenant intrusions with a high
level of assurance [Charisma]. Resilience to security threats may
involve security measures that could degrade QoS, for example by
increasing latency or dropping packets. The objective is achieving
and maintaining 5G performance while assuring security. For cases
that the primary attack cannot be prevented the best solution is
to predict cascades of secondary effects caused by security threats
[5G-ENSURE], in order to identify measures that could prevent the
propagation of adverse effects.

for UEs attached to a UE-Relay, in areas with poor or no network
coverage, such as tunnels, caves, valleys; in-door, such as in
buildings, basements; and in emergency situations such as lifesaving disaster relief missions [Coherent]. Depending on the
technology properties, service requirements, mobility information
and type of environment, the C3 can decide how to improve
mobility management, how to perform relay selection and how
to configure the system for coverage extension in out-of-coverage
situations. The C3 configures different network entities for relay
selection, coverage extension and handover. This reconfiguration
enables providing different service priorities to relay firemen and
policemen under network out-of-coverage situations in both urban
and rural environments.
These innovations drastically enhance reliability and increase
resilience, although in a wireless context performance degradation
may be unavoidable. The strict safety requirements of critical
applications, imposed by legal constraints, motivate the definition
of graceful degradation options for 5G performance indicators such
as data rate, latency, reliability, security and trustworthiness, in
order to survive the worst case scenarios.

3.4
A quick end-to-end
service deployment
Vertical industry applications require customised access for
different stakeholders. Providing customized access to different
stakeholders is labour- and time-intensive during installation and
commissioning as well as during operations and maintenance
(O&M) activities [VirtuWind). Programmable networks and multitenant capabilities in 5G ensure fast deployment and new services.
This includes the ability to create, sell and provision composite
services in multi-domain environments; technology domains (intraoperator) as well as administrative domains (inter-operator). The
5G PPP target time in reaching a complete service deployment is
less than 90 minutes.
The following sections present approaches for reducing the service
deployment time, starting from a multi-domain orchestrator
that enables service creation over multiple administrative and
technology domains, before detailing intra-operator orchestration
and management tools.

Multi-domain orchestration allows a drastic decrease
of the service deployment time. Provisioning time for
a new antenna in the edge up to its operational stage
can be reduced from 120 hours to 90 minutes. The
setup time for a service published in a marketplace,
spanning across multiple administrative domains, can
be as low as few seconds.

Service continuity during disaster scenarios
CContinuity of service during hazard and emergency events is
crucial. A number of techniques are being explored, among which
UE-relaying and satellite networks. Based on their native multicast/
broadcast capabilities, these latter can offer solutions to “thundering
herd events” where huge number of data and video requests need
to be served by 5G networks.
As of UE-relaying, envisaged by 3GPP as a key technique for
supporting coverage extension for public safety applications, it
can profit from the Central Controller and Coordinator introduced
above. The C3 can improve mobility and network management

13

A Multi-domain Orchestrator for reduced
end-to-end service deployment time
The enablement of cross-domain orchestration of services over
multiple administrative or technology domains through a Multidomain Orchestrator (MdO) [5GEx] enables end-to-end network
and service elements to mix in multi-vendor heterogeneous
technology and resource environments, realizing a full end-to-end
service deployment within a reduced time.

Property of the 5G Infrastructure Association

FIGURE 5

A testbed composed of several testbed sites was set up across
Europe, in order to experiment with the MdO at large scale and
to measure the service deployment time. The experiment is
performed in two steps: first, the service provider defines a new
service as a chain of multiple VNFs, and publishes it in a service
catalogue; the marketplace. Then a customer ‘buys’ the service from
the marketplace, and the MdO deploys the corresponding service
chain across the testbed sites. Each MdO can be an entry point for
customer service requests (Figure 5 is just a simplified example).
If the resources to provision the service are not available in the
same administrative domain, the MdO seeks them in its neighbour
domain by communicating with its peer MdO. This process can be
cascaded through more domains until all the resources have been
allocated. The service deployment time is evaluated as the sum of
the time the MdO needs for creating the service and publishing it in
the marketplace (service creation), plus the time to instantiate the
service across the testbed sites upon a customer request (service
provisioning). The objective of the experiment is to demonstrate
that the MdO is able to reduce the service deployment time from
hours to seconds. [5GEx].

END CUSTOMER

Multi-Domain
orchestrator

Management &
orchestration
plane

(legacy)
control plane
Data plane

Light weight virtualization for a fast VNF
deployment

Multi-domain orchestration for a flexible
service deployment.

Mgmt/orch
Business
Control (legacy)
data

Inter-operator
Orchestration API

Administration C

Multi-Domain
orchestrator

Inter-operator
Orchestration API

Administration A

B2B

B2B

Domain
orchestrator

Domain
orchestrator

Domain
orchestrator

Domain
orchestrator

Network
controller

Network
controller

Network
controller

Network
controller

satellite
network

SDN
NETS.

OPERATOR C Administration

PACKET/
OPTO

LEGACY
Nets.

OPERATOR A Administration

Multi-Domain
orchestrator
Administration B

Domain
orchestrator

Network
controller

Network
controller

Data center
PACKET

OPERATOR B Administration

Reducing service deployment time within
the technology domain
Within the domain of the same operator, the service deployment time
depends on the availability of physical and software infrastructure
and, once the infrastructure is ready, on the performance of the
network management and control framework.
First, when the demanded service does not yet exist, a Service
Development Kit (SDK) helps reducing development time based
on two capabilities: (i) the possibility of reusing VNFs previously
validated and stored in a function catalogue in the construction of
new services, and (ii) tools such as profiler or emulator, that enable
a feedback cycle to the developer of the service, who can test and
optimise the service being developed before deployment. Speed
of deployment is achieved by providing a service platform that is
highly configurable, with an architecture based on micro-services
and a message broker that can accept service or function specific
managers, which can adapt the default behaviour of the platform
[Sonata].

Property of the 5G Infrastructure Association

Once the software service is available, one has to ensure that
the physical infrastructure for the targeted service coverage is
deployed and configured for the desired service. A self-organized
autonomic network management framework enables the reduction
of the average service creation time through key innovations
such as: automated physical, virtual infrastructures, and services
deployment; and integrated management and orchestration of SDN/
NFV apps for on-demand service creation [Selfnet]. For example,
for the use case of provisioning a new antenna in the edge up to
its operational stage, measurements have been made to exemplify
the achievable service deployment times. The approach being
implemented for testing the framework performance is gathering
the results for the deployment time for a completely “empty”
set of nodes including the deployment of a base station service,
supported by an existing OpenStack infrastructure. Time figures to
deploy 6 physical machines simultaneously have been gathered in
a process that includes all necessary steps, from edge installation
to configuring the Evolved Packet Core (EPC), and compared to the
figures obtained from a proof of concept prototype. The total time
has been reduced from 120 hours to 90 minutes [Selfnet].

NFV can be used to support highly dynamic scenarios, in which the
VNFs are instantiated “on the fly” following the service requests.
VNFs tend to become small and highly specialized Micro-VNFs,
i.e., elementary and reusable network elements. Complex services
can be built through the “chaining” of these Micro-VNFs. Different
virtualization approaches can be used to support these microVNFs: Tinified VMs and unikernels. Unikernels have very important
properties allowing to reduce the service deployment. They offer
very good performance in terms of low memory footprint and
instantiation time. They have very good isolation and security
properties. The recent measurements using ClickOS, a Xen-based
unikernel, demonstrate a small footprint (around 5 MB when
running) and an instantiation time around 30 milliseconds while
processing up to 10Gb/s of traffic. It furthermore does not need
a disk to work. In this way, it is possible to make very efficient use
of resources, allowing thousands of unikernels to run on a single
physical host and offer a fast end-to-end service deployment
[Superfluidity].

Network programmability for a reduced
service configuration time
The introduction of Software-defined Mobile Network Control (SDMC)
in 5G networks extends the concept of network programmability
beyond SDN. While SDN decouples network control and forwarding
functions, SDMC introduces the separation of logic and agent for
any network function in the network, extending the SDN principles
to all control-plane, user-plane and management functions typically
deployed in mobile networks.
SDMC offers a high level interface to several network functions
ranging from radio control to traffic steering. With SDMC, service
providers are able to configure the equipment to their needs by
simply re-programming the controller using well-defined APIs,
enabling a new service within a reduced implementation, test and
deployment timescale. The definition of a standard northbound
interface simplifies the creation of new network functions, as the
low level and vendor-specific characteristics are managed by the
SDMC controller southbound interface, with a clear advantage in
heterogeneous and dense wireless networks. SDMC can employ
tailored algorithms per network slice they are deployed in and
can also manage on-the-fly deployment of VNFs close to the users
reducing their experienced latency. This feature is desirable for the
verticals market, as several network operators can provide their
services to verticals by using the SDMC approach.

14

04

RECOMMENDATIONS

Sections 2 and 3 provided details of technological and architectural
innovations researched and developed by 5G PPP projects and
covering innovation areas such as:
Flexible air interface design that serves the three main
services classes for 5G
Integration of different radio access technologies within
the 5G transport network
Flexible architecture that allows quick setup of slices on
multi-tenant networks
Integration of fog computing and mobile edge computing
near the end user and within vertical premises.
These innovations are shared with the global 5G research
community in order to avoid fragmentation of standards and
deployed technologies. This is being achieved by the 5G PPP through
international cooperation with other regions such as the Americas,
China, Japan, and South Korea. The 5G PPP projects are building prestandards consensus among their partners and provide relevant
contributions to standardization. The 5G Infrastructure Association
is a market representation partner of 3GPP that brings into 3GPP
a consensus view of market requirements and contributes to
the roadmap of 3GPP. Action towards regulation bodies is being
undertaken for securing sufficient spectrum for 5G and a stable
regulatory framework across all stakeholders. Last but not least,
cost effective deployment strategies are being developed, including
an upgrade path from 4G, a cost-effective transport network and
deployment and operation cost sharing strategies among involved
stakeholders.

4.1
Standardisation
Global standardization is an important element of the long term
sustainability and the widest possible use of the 5G PPP results
and reduces the risk of fragmentation of future deployments.
The benefit of standards is that they foster a wide ecosystem and
provide increased efficiency by eliminating redundant options.
However, standards can limit innovation by preventing use of new
technology and raising entry barriers. Therefore, it is important to
select the right areas to standardise and create standards with the
right properties.
Standardisation has provided profound benefits by the definition
of a common air interface and will likely do so in the future. The
virtualisation of the network and possible sharing of resources
implies the need for standardised interfaces between the virtual
functions and the execution platform. This must also cover
distributed network approaches where some resources are located
at the network edge. Management of future networks is likely
more complex and a standardised approach may be beneficial to
efficiently manage them. Initiatives such as Multiple Operator Core
Networks (MOCN) constitute an enabler for sharing RAN resources.
Such initiatives need to be extended covering the entire ICT edge
infrastructure.
Standards must be flexible to support and sustain the diversity of
business models and deployments of 5G networks. The standards
should cover all use case classes; for example it is important that
eMBB, URLLC and mMTC are covered in the same standards
framework. Standards should be unified and non-fragmented to
ensure global and cost-effective mobility of users and equipment.

15

6

The 5G PPP programme has influenced the current standards
evolution by catalysing the vision what 5G should be about (eMBB,
MTC and URLLC) and articulating the overall 5G key performance
indicators.
Results to date have been proposed to related standards bodies,
most prominently represented by 3GPP, ETSI and ITU. The
contributions of results include:
In 3GPP-RAN specifications of the physical layer of the radio
Interface for UE as well as radio interface architecture and
protocols, radio resource control and management and
the services provided to the upper layers.
In 3GPP-SA specifications of services and features,
definition and evolution of the overall architecture, and
addressing security and privacy by design.
In ETSI contributions to MEC (Mobile Edge Computing),
RRS (Reconfigurable Radio System) and TC CYBER (Cyber
Security).
In ITU contributions to SG15 on network technologies for
transport.
In the future it should be ensured that use-cases important for
Europe are considered and included in the standards, so that
the technology required to realise them is developed. Finally, it is
important that demand for new services requiring new technology
is stimulated so that the equipment is developed and brought to
market, fostering a healthy ecosystem.

4.2
Spectrum
Early access to the necessary frequency bands is critical for Europe
to perform 5G technology tests, trials, pilots and for the early launch
of commercial products services. The news release of the Radio
Spectrum Policy Group (RSPG) states the following on 5G “pioneer
bands”6 :
Low bandwidth spectrum (700 MHz) which can enable 5G
coverage to all areas, ensuring that everyone benefits;
Medium bandwidth spectrum (3.4-3.8 GHz) which will bring
the necessary capacity for new 5G services in urban areas;
and
High bandwidth spectrum (26 GHz) to give ultra-high
capacity for innovative new services, enabling new business
models and sectors of the economy to benefit from 5G.
Those RSPG nominated 5G pioneer bands eases the early
development of the necessary equipment and facilitates faster
preparation of tests and trials.
Sufficiently large radio channels and bandwidths are necessary for
supporting the eMBB use case classes at 3.4-3.8 GHz, in addition to
the 26 GHz band.
The radio frequency channels needed for 5G are of at least 100 MHz
width in the 3 - 4 GHz range rising to 500 MHz in the frequency
range between 5 to 33 GHz and as wide as 1000 MHz at the highest
mm-wave options.

RSPG16-032 “Strategic roadmap towards 5G for Europe - Opinion on spectrum related aspects for next-generation wireless systems (5G)”, Brussels, 09 November 2016.
See also RSPG Chair News Release on 5G Spectrum, 10 November 2016

Property of the 5G Infrastructure Association

4.3
Deployment

4.4
Regulation

The key innovations in 5G depicted in this paper represent a
revolution for user experience, new services and new business
models. The architecture and protocols are designed to adapt
to a wide range of deployment scenarios including deep indoor,
hot spots, urban areas, rural areas, maritime areas and in an
aeronautical context. The 5G concept combines various access
technologies, such as cellular, wireless, satellite and wireline, for
delivering reliable performance for critical communications and
improve area coverage.

Regulation must adopt a facilitating harmonised approach for
supporting 5G deployment. The promotion of investments
requires a stable, consistent and accurate regulatory framework
across all stakeholders. Regulation must increase the consistency
of administrative conditions to facilitate dense cell deployments,
including (i) right-of-way to passive facilities; (ii) supportive
municipal site rental charges; (iii) removal of taxation on sites; and
(iv) predictable and harmonised electromagnetic field emissions
limits.

The 5G network deployment has to be cost effective in order
to materialise. Efficient and progressive deployment strategies,
reusing as much as possible 4G infrastructure and exploiting new
available spectrum are currently being elaborated in the 5G PPP
programme [METIS-II]. For example, in dense urban scenarios, 5G
radio base stations should be co-located with 4G base stations and
exploit the newly available spectrum, e.g., at 700 MHz and 3.4-3.8
GHz. Furthermore, in order to provide enhanced capacity, small
cells could be added in frequency bands below 6 GHz (such as
2.6 GHz and 3.4-3.8 GHz) and in mm-wave bands above 24 GHz.
Additionally, complementary use of unlicensed spectrum and the
use of nomadic nodes (see section 2.2) are under consideration for
increasing the average user throughput.

Regulation rules must promote 5G services and avoid restricting
implementation options such as with respect to the slicing concept,
end-to-end virtualization and network sharing. Without this
flexibility, multi-domain slicing and networking and processing
resource sharing cannot be implemented, reducing the economic
value for a wide range of services.

The design of the fronthaul and the functional split have high
impact on the network cost and must be optimised. For example,
for a spectrum bandwidth of 200 MHz and beamforming (8x8
antennas) and if the 5G radio unit is in charge of functionalities from
physical layer to resource element mapping, the peak throughput
for the new radio splitting interface is estimated in some tens of
Gbit/s, generating several hundreds of Gbit/s for the aggregate
signal in fibre [5GCrosshaul]. Current 100 Gbit/s optical interfaces
are too expensive for the fronthaul segment and cost effective
100 Gbit/s direct detection transceivers for Datacom applications
cannot provide sufficient link capacity over desired distances. Such
interfaces are suitable for peer-to-peer connectivity but are not
able to support aggregated traffic in networking scenarios. In view
of these constraints and considering realistic number of add/drop
nodes in the optical network, two potential solutions are discussed:
(i) an optimal functional split that minimizes the Total Cost of
Ownership (TCO) [METIS-II] and (ii) novel technologies, such as
integrated optical chipsets for fibre dispersion compensation and
advanced modulation formats to meet 5G transport requirements
at acceptable cost [5GCrossHaul].

Regulation must ensure equivalent and proportionate privacy
requirements between operators and online service providers, and
remove any roadblocks to development of innovative 5G services
in all vertical industries. For low-latency high-reliability 5G services,
regulation must clarify the liability issues. Privacy aspects require
timely and targeted actions, as stressed by EU Privacy Mandates
(e.g. M/530) and the General Data Protection Regulation (GDPR)
which requires privacy by design.
Finally the public sector should act as an early adopter and promoter
of 5G technologies, for example through public procurement (e.g.
for allowing vertical sectors to adopt 5G). Such initiatives help
building the business case for the necessary investments.

The cost minimization strategies alone will not be sufficient for
ensuring an economically viable 5G network. Considering the
new business models, innovative deployment strategies must be
conceived that involve all stakeholders, from telecommunications,
content and vertical industries, and which collaborate for sharing
the deployment cost and the associated revenue [METIS-II]. The
broker business role introduced in Table 1 facilitates resource
sharing and makes it dynamic in time and space, exploiting thus
networking resources from various access technologies (such as
cellular, wireless, satellite and wireline). The trustworthy and secure
slicing concept and the multi-domain orchestrator presented
in section 3.4 enable the provisioning of an end-to-end service
spanning the infrastructure of multiple stakeholders, ensuring
reliability for critical communications and improving availability
with wide area coverage.

Property of the 5G Infrastructure Association

16

Contributors
List
Editor

Contributors

Salah Eddine El Ayoubi
Orange, salaheddine.elayoubi@orange.com

George Agapiou (OTE research), Ricardo Marco Alaez
(University of the West Scotland) , Jose Alcaraz-Calero
(University West Of Scotland), Maria Barros Weiss
(Eurescom), Pascal Bisson (Thales), Nicola Blefari Melazzi
(University of Rome Tor Vergata), José Bonnet (Altice
Labs), Michael Bredel (NEC), Ömer Bulakci (Huawei),
Carolina Canales (Ericsson), Sonia Castro Carrillo (ATOS),
Fabio Cavaliere (Ericsson), Tao Chen (VTT), Enrique
Chirivella-Perez (University West Of Scotland), Nicolas
Chuberre (Thales AleniaSpace), Alessandro Colazzo
(Azcom), Luis Miguel Contreras Murillo (Telefonica),
Giovanna D’Aria (Telecom Italia), Thomas Deiss (Nokia),
Panagiotis Demestichas (University of Piraeus), Marco Di
Girolamo (HPE), Mark Doll (Nokia), Manos Dramitinos
(Athens University of Economics and Business), Simon
Fletcher (RealWireless), Vassilis Foteinos (Wings ICT
Solutions), Alexander Geurtz (SES), Marco Gramaglia
(UC3M), Jesus Gutierrez Teran (IHP Microelectronics),
Frederic Kermel (Airbus), Michael Kraemer (Intel),
Vivek Kulkarni (Siemens), Orestis Liakopoulos (Wings
ICT Solutions), Hakon Lonsethagen (Telenor), Diego R.
Lopez (Telefonica), Jian Luo (Huawei), Aarne Mämmelä
(VTT), Patrick Marsch (Nokia), Josep Martrat (ATOS),
Francesco Mauro (Telecom Italia), Ioannis Neokosmidis
(Incites), Navid Nikaein (Eurecom), Miquel Payaró (CTTC),
Rauno Ruismaki (Nokia), Dario Sabella (Telecom Italia),
Mehrdad Shariat (Samsung), Martin Schubert (Huawei),
Christoph Thuemmler (Edinburgh Napier University),
Fabio Ubaldi (Ericsson), Panagiotis Vlacheas (Wings ICT
Solutions), Sark Vladica (IHP Microelectronics), Thomas
Walter (Docomo), Qi Wang (University West Of Scotland),
Yue Wang (Samsung), Jean-Philippe Wary (Orange), Mick
Wilson (Fujitsu)

Authors
Jean Sébastien Bedo (Orange)
Salah Eddine El Ayoubi (Orange)
Miltiadis Filippou (Intel)
Anastasius Gavras (Eurescom)
Domenico Giustiniano (Imdea)
Paola Iovanna (Ericsson)
Antonio Manzalini (Telecom Italia)
Olav Queseth (Ericsson)
Theodoros Rokkas (Incites)
Mike Surridge (University of Southampton)
Terje Tjelta (Telenor)

Property of the 5G Infrastructure Association

17

Supported by the

This material has been designed
and printed with support from
the Euro-5G project and the 5G
Infrastructure Association. The
Euro-5G Project has received
funding by the European
Commission’s Horizon 2020
Programme under the grant
agreement number: 671617.
The European Commission
support for the production of this
publication does not constitute
endorsement of the contents
which reflects the views only of
the authors, and the Commission
cannot be held responsible for
any use which may be made of the
information contained therein.

More information at

www.5g-ppp.eu



Source Exif Data:
File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.3
Linearized                      : Yes
Language                        : en-GB
XMP Toolkit                     : Adobe XMP Core 5.6-c137 79.159768, 2016/08/11-13:24:42
Create Date                     : 2017:03:06 18:25:33+01:00
Metadata Date                   : 2017:03:06 18:25:45+01:00
Modify Date                     : 2017:03:06 18:25:45+01:00
Creator Tool                    : Adobe InDesign CC 2017 (Macintosh)
Instance ID                     : uuid:73566fe0-5726-2f48-82f2-9eff0e1f99f7
Original Document ID            : xmp.did:2c29d16c-595c-404b-b639-c891c7511893
Document ID                     : xmp.id:cb8e46c9-8b39-4abf-a7de-31817b31f1e0
Rendition Class                 : proof:pdf
Version ID                      : 1
History Action                  : converted
History Parameters              : from application/x-indesign to application/pdf
History Software Agent          : Adobe InDesign CC 2017 (Macintosh)
History Changed                 : /
History When                    : 2017:03:06 18:25:33+01:00
Derived From Instance ID        : xmp.iid:331d53cd-6f6c-4c25-8ae0-d00d1d457b03
Derived From Document ID        : xmp.did:2c29d16c-595c-404b-b639-c891c7511893
Derived From Original Document ID: xmp.did:2c29d16c-595c-404b-b639-c891c7511893
Derived From Rendition Class    : default
Format                          : application/pdf
Title                           : 5GPPP-brochure.indd
Producer                        : Adobe PDF Library 15.0
Trapped                         : False
GTS PDFX Version                : PDF/X-1a:2003
GTS PDFX Conformance            : PDF/X-1a:2003
Page Count                      : 18
Creator                         : Adobe InDesign CC 2017 (Macintosh)
EXIF Metadata provided by EXIF.tools

Navigation menu