kom (2016) 0587 - Ingen titel

Europaudvalget 2016
KOM (2016) 0587
Offentligt

EUROPEAN

COMMISSION

Brussels, 14.9.2016

SWD(2016) 300 final

COMMISSION STAFF WORKING DOCUMENT

Accompanying the document

COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN

PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL

COMMITTEE AND THE COMMITTEE OF THE REGIONS

Connectivity for a Competitive Digital Single Market - Towards a European Gigabit

Society

{COM(2016) 587 final}

kom (2016) 0587 - Ingen titel

Table of Contents

1. Introduction

1.1 Growing demand for connectivity

1.2 The user's perspectives

2. State of play

2.1 Current state of connectivity in Europe

2.2 A growing need for connectivity

2.3 Baseline analysis: from today to 2020 and beyond

2.4 The gap between bandwidth demand and network capacity deployed

3. Cost and benefits of very high capacity connectivity

3.1 Costing the networks for a Gigabit Society

3.1.1 Costing the gap and the financial endowment of current initiatives

3.1.2 The trajectory of private infrastructure investments

3.2 The importance of defining objectives for broadband

4. Strategic objectives for 2025

4.1 The importance of setting realistic political guiding objectives

4.2 Leveraging mutually reinforcing objectives and existing infrastructures

4.3 Benefits of the strategic objectives for 2025

4.3.1 Gigabit connectivity for all main socio-economic drivers such as schools, transport hubs and

main providers of public services as well as digitally intensive enterprises

4.3.2 High performance 5G connectivity: by 2020 a fully-fledged commercial service in at least one

major city in each of the 28 Member States and by 2025 uninterrupted 5G coverage of all urban

areas and major terrestrial transport paths

4.3.3 All European households, rural or urban, to have access to Internet connectivity offering a

download speed of at least 100 Mbps, upgradable to Gigabit speed.

4.4 The need to prioritise investments

Annex: Technological developments

II.

Which technology?

What will technology deliver

III. Copper based technologies

IV. Optical fibre – a new generation of networks

kom (2016) 0587 - Ingen titel

Fibre-to-the-Business

Fibre-to-the-Home

Outlook for the 2025 horizon

Investing in FTTH

Will fibre-based technologies be able to meet growing demand?

Wireless technological developments

kom (2016) 0587 - Ingen titel

1. Introduction

This Staff Working Document accompanies the Communication "Connectivity for a

Competitive Digital Single Market – Towards a European Gigabit Society". The purpose of

this document is to provide the background information which underpins the Commission's

choice of new proposed long term strategic objectives for 2025 to focus on Europe's

connectivity needs beyond 2020.

1.1

Growing demand for connectivity

The creation of a connected Digital Single Market is one of the top priorities of the European

Commission. One of the three pillars of the European Commission's Digital Single Market

Strategy of May 2015 aims to create the right environment and conditions for the deployment

of very high-capacity networks. Both the European Parliament

and the European Council

have recently emphasized the need for more Internet connectivity.

The prerequisite to achieve a fully functional Digital Single Market is to ensure access to

ubiquitous, very high-capacity fixed and mobile infrastructures. The increase in data

consumption and usages (see section 2.2) and the process of aggregation and convergence

between wireless and fixed networks will require the provision of very high-capacity (VHC)

networks

ever closer to the end-user.

This is particularly relevant in view of Europe's competitive position in the Global Digital

Economy. Gigabit connectivity is already a reality in countries such as Japan and South

Korea, and is translating into increasing usage of video and high bandwidth applications.

Bandwidth usage in Korea has historically been considerably higher than in Europe. It now

also seems that Japan, which initially had limited bandwidth usage despite high fibre

coverage, is starting to take off and would overtake Europe according to forecasts from

IDATE

. In mid-2015, VHC networks represented around 70% of total fixed broadband in

Japan and South Korea. Comparing the dynamics of investment into high capacity networks

it can be concluded that South Korea and Japan have already entered into the Gigabit era, and

China and Russia are also pursuing network rollout at a similar level of ambition. In

comparison, by mid-2015, in the EU NGA networks represented 9% of the total fixed

broadband subscriptions and e.g. Fibre to the Premises (FTTP) coverage was 20.8%, although

in some Member States, such as Estonia, Portugal, Spain and Sweden, deployment was

taking place at a much more significant scale.

1 European Parliament Resolution of 19 January 2016 on Towards a Digital Single Market Act (2015/2147(INI))

2 Conclusions of the European Council meeting of 28 June 2016 (EUCO 26/16).

3 "Very high-capacity network" means an electronic communications network which either consists wholly of optical fibre

elements at least up to the distribution point at the serving location or which is capable of delivering under usual peak-time

conditions similar network performance in terms of available down- and uplink bandwidth, resilience, error-related

parameters, and latency and its variation. Network performance can be considered similar regardless of whether the end-

user experience varies due to the inherently different characteristics of the medium by which the network ultimately

connects with the network termination point.

4 See Study on "Regulatory, in particular access, regimes for network investment models in Europe", by WIK Consult,

Deloitte and IDATE; SMART 2015/0002.

kom (2016) 0587 - Ingen titel

Figure 1: VHC coverage in the European Union, June 2015

When considering the growing capacity needs in terms of network infrastructures, there is a

need to take into account how new devices are enabling the development of new applications

(e.g. smartphones) and affecting customer bandwidth requirements (e.g. larger screens and

higher resolution). In addition, a single connection/subscription often serves simultaneously

multiple users, in particular for households with children, SMEs and organisations like

schools and libraries, further increasing the need to ensure speed and quality of experience.

The widespread adoption of cloud services

, the number of connected devices (IoT) and the

booming Machine-to-Machine (M2M) industry

is also contributing to further increase the

traffic load on communications networks, including mobile networks

In particular the growing use of smartphones and tablets increases the traffic per wireless

access point (Wi-Fi and/or 4G/LTE base stations) and will require increasingly smaller cells

to deliver the planned 5G connectivity performance. The volume of data that will transit

through future 5G small cells will reach several gigabits per second (Gbps) and be further

aggregated into multi-Gbps traffic streams.

This requires appropriate high capacity backhaul communications, which will mostly be

fibre-based links, making 5G a complement but not a replacement to fixed VHC networks in

areas with such a dense concentration of connected devices and users. In addition,

developments in 5G are likely to lead to massively improved fixed wireless (as opposed to

cellular wireless) solutions.

1.2

The user's perspectives

The results

of the public consultation launched last year by the European Commission on the

need for Internet speed and quality beyond 2020 show the perception of respondents – mostly

individual and business users – that important improvement in connectivity features is needed

Communication from the Commission on a European Cloud Initiative-Building a competitive data and knowledge

economy in Europe, COM(2016) 178 final.

Communication from the Commission on Digitising European Industry. Reaping the full benefits of a Digital Single

Market COM(2016) 180 final.

Commission Staff Working Document Europe's Digital Progress Report 2016, SWD(2016) 187 final.

Full synopsis of the results available at https://ec.europa.eu/digital-single-market/en/news/full-synopsis-report-public-

consultation-needs-internet-speed-and-quality-beyond-2020

kom (2016) 0587 - Ingen titel

for the future. The public consultation was addressed to all types of users in all sectors. Some

1550 respondents from across Europe contributed to the public consultation, of which around

85% as individuals and 15% as (public or private) organisations. In particular, the results of

the public consultation show the following main trends:



Good connectivity is seen as a

necessary condition to achieve the Digital Single

Market, for citizens, for the industry, for businesses

(whatever their size and sector of

the economy),

for schools, for research and innovation centres,

as well as – more

generally - for competitiveness, jobs and growth. Inadequate connectivity is considered as

a risk or a high risk for around three quarters of the respondents – affecting in particular

employment, cohesion, education and learning, research and data driven activities,

consumer welfare, and accessibility.

While

respondents pointed to download speed as the most important feature

of fixed

connectivity today,

other connectivity features will gain significant importance

in the

future - notably upload speed, reliability and uninterrupted access.

Contributors have a

low trust that future connectivity would spontaneously emerge

a level of quality and speed that would fulfil their needs: more than half declared

themselves sceptical.

Among the respondents who are pessimistic or very pessimistic about the fulfilment of

their future fixed and/or mobile connectivity needs in 2025, close to

90% think that

measures by public authorities are needed to promote investment in and take-up of

connectivity networks and services

in line with (or even beyond) identified future

needs.



The results of a representative Eurobarometer survey

, carried out in the 28 Member States of

the European Union between 17 and 26 October 2015, confirm that when subscribing to an

Internet connection, respondents are increasingly likely to consider quality criteria. They are

important criteria for 70% of customers, which makes them the second most important. In

terms of trend, quality criteria are those that gained most in importance compared with 2014:

e.g. +7 points for maximum download or upload speed, +6 points for maximum amount of

data downloaded or uploaded.

Figure 2 summarises the main trends as far as households' Internet access is concerned.

27,822 EU citizens from different social and demographic categories interviewed face-to-face at home in their native

language

kom (2016) 0587 - Ingen titel

Figure 2: Percentage of households with fixed and/or mobile Internet access in EU

Base: All respondents (n = 27822)

These trends underline the growing needs for Internet connectivity for all the sectors of the

European economy and show that expanding individual usage aggregates into quickly

growing demand for VHC fixed and mobile broadband infrastructures that can underpin the

Digital Single Market. When considering these trends in Europe, it should further be

emphasised that future needs should in particular take into account current practices of

today's "digital natives" (below 25 years old) since they will increasingly influence

mainstream data consumption patterns as this generation grows in demographic terms and

becomes part of tomorrow's workforce.

In addition, the availability of higher capacity connectivity networks in itself drives the

development of innovative and value-adding services: failing such favourable conditions,

providers have to adapt their services or launch them elsewhere. Updating and enhancing

existing network infrastructures to meet this anticipated growing demand for higher capacity

will however require significant additional investments.

2. State of play

2.1 Current state of connectivity in Europe

Today virtually all EU citizens have access to basic broadband networks

(97% have access

to fixed broadband connections according to the DESI index 2016

) and increasing numbers

At a download speed of at least 2Mbps at end-user level.

kom (2016) 0587 - Ingen titel

of citizens and businesses have access to networks allowing at least 30 Mbps download

speed: 70.9% have general coverage at Next Generation Access (NGA) connectivity level

in the EU, according to DESI 2016

However, only some countries, such as Malta, Lithuania, Belgium and the Netherlands,

already enjoy nearly comprehensive coverage of NGA networks; in most of those cases, this

is probably due to the competitive impulse provided by cable networks, which can be

upgraded at relatively low cost to NGA levels of connectivity

. Elsewhere, NGA coverage

has been slow to develop – notably in countries that lacked extensive cable in their legacy

networks (Italy or Greece being examples).

However, the roll-out of NGA networks has improved significantly since the DAE objectives

were adopted in 2010. As shown in Figure 3, some Member States in particular have been

able to leap frog others in increasing network capacity.

Figure 3: 30Mbps and more (NGA) coverage of households per Member State (2011 vs 2015)

100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

MT BE NL LT LU DK PT LV UK AT EE CY DE IE SI HU ES SE FI CZ BG RO SK PL HR FR IT EL

2011

2015

NGA broadband coverage

(as a % of households)

The Digital Economy and Society Index (DESI) is a composite index developed by the European Commission (DG

CNECT) to assess the development of EU countries towards a digital economy and society. It aggregates a set of relevant

indicators structured around 5 dimensions: Connectivity, Human Capital, Use of Internet, Integration of Digital

Technology and Digital Public Services. For more information about the DESI please refer to http://ec.europa.eu/digital-

agenda/en/digital-agenda-scoreboard

NGA broadband coverage/availability (as a % of households) with Next Generation Access including the following

technologies: FTTH, FTTB, Cable DOCSIS 3.0, VDSL and other superfast broadband allowing at least 30 Mbps

download).

Several studies highlight the role played by cable networks in stimulating NGA deployments including SMART

2015/0002, WIK-Consult (2015) for Ofcom ‘Competition and Investment: analysing the drivers of superfast broadband’,

and the EP (2013) study ‘Entertainment X.0 to boost broadband deployment’.

kom (2016) 0587 - Ingen titel

Yet

there are still substantial differences between Member States

in terms of take-up of high

speed (NGA) broadband subscriptions.

Figure 4: fast broadband (at least 30 Mbps) household penetration, July 2015.

In addition, the connectivity challenge in rural areas remains acute, with patchy basic NGA

coverage exemplified by 28% fixed line coverage and 36% mobile 4G household coverage

(both coverages being partially overlapping).

Figure 5: NGA broadband coverage in EU

Basic NGA broadband coverage in the EU

100%

Total

80%

60%

40%

20%

End 2010

End 2011

End 2012

End 2013

End 2014

Mid 2015

Rural

Source: IHS and VVA

A key development driving the increase of network capacity is that legacy telephone and

cable (coaxial) networks, including the copper ‘local loops’, are in the process of being

upgraded to improve their broadband performance.

In many countries, including the UK or Germany, a large part of the NGA coverage beyond

the cable footprint has been achieved through only partial upgrades of the legacy copper loop

(FTTC), rather than through full VHC upgrades (e.g. FTTH/B). As investigated in the Study,

For instance according to the Digital Economy and Society Index, updated in July 2015, 22 % of European homes

subscribed to broadband access of at least 30 Mbps but this represented only 1% of households in Hungary and 60% in

Belgium. Similarly, subscriptions to broadband access of at least 100 Mbps reached on average 11% in the EU, but less

than 1% in Croatia, Cyprus, Greece, and Italy (where networks capable of such speeds are largely absent) while it

exceeded 40% in Latvia, Romania and Sweden.

kom (2016) 0587 - Ingen titel

"Regulatory, in particular access regimes for network investment models in Europe"

, the

former approach is most unlikely to be sufficient to cope with data consumption under the

most ambitious scenario forecast. As a result

, t

he current state of broadband connectivity in

Europe and the actual trends in its modernisation will not fulfil the growing needs for

Internet speed and quality beyond 2020.

Figure 6: Projections of NGA penetration by technology 2015 – 2025 (% of households,

EU28)

Despite some technological progress

, today not all NGA networks can deliver 100 Mbps.

This remains an important challenge for the achievement of the 2020 DAE objective for

100 Mbps subscriptions of one European household in two: it will require both a wide

availability and affordable offers to increase demand for connectivity of at least 100 Mbps by

2020.

See Study by WIK-IDATE-Deloitte;"Regulatory,

in particular Access regimes for network investment and business

models in Europe";

SMART 2015/0002.

In July 2010, only 0.5% of lines in the EU had average speeds at or above 100 Mbps. By mid-2015, an estimated 11% of

fixed broadband subscriptions in the EU were able to provide at least 100 Mbps.

kom (2016) 0587 - Ingen titel

Figure 7: Fixed broadband subscriptions in EU by speed

Take-up projections of NGA in a 5-10 year timeframe vary, and show significant differences

across countries and connectivity parameters.

For example, based on announcements from operators, Government subsidy initiatives and

trends in demand, IDATE has projected that overall NGA take-up will surpass the 50% mark

by 2020 in the EU on average. However, it is notable that a high proportion of take-up is

projected to be on the basis of FTTC technologies which may not in all cases meet the

100Mbit/s quality of service objective. When such technologies are excluded from the

calculations, take-up is projected to be around 45% of households in 2020, falling short of the

DAE objective. Besides, gaps between countries are expected to persist, with a significant

proportion of countries expected to lag behind the DAE take-up objective, even if all

technologies are considered.

As far as FTTH/B is concerned, projections by experts

suggest that only 20% of households

would be subscribing to an FTTH offer by 2020 leading to a take-up of 31% by 2025. Due to

different levels in the projected coverage of FTTH, take-up is also expected to vary widely by

country, with high take-up in several Eastern European countries as well as in Portugal,

Sweden, Spain and to a lesser extent France, contrasting with low take-up in several countries

including the UK, Germany, Italy and Poland (See Figure 8).

See Study

WIK-IDATE-Deloitte: "Regulatory, in particular access regime" ;

SMART 2015/0002

kom (2016) 0587 - Ingen titel

Figure 8: Projections for FTTH/B penetration by country (% of total households, 2015-2025)

IDATE

Projections by 2019 from other sources such as Heavy Reading

(see Figure 9), also suggest

low take-up levels in the UK, Italy, Poland and Germany.

Figure 9: Projections for household take-up of FTTH/B in December 2019 (Heavy Reading)

http://www.ftthcouncil.eu/documents/Webinars/2015/Webinar_14April2015.pdf

kom (2016) 0587 - Ingen titel

Closer analysis of the developments reveals a trend that, as business and household services

and applications depending on high quality connection are becoming more popular,

subscriptions to 100 Mbps or more are growing sharply, albeit from a low base. Especially

pronounced in the Member States which already have the highest 100 Mbps subscription rate,

this suggests both important emulation effects on demand and increasing supply of attractive

services which exploit such higher capacity connectivity – the so-called next generation

applications. Figure 10 shows how dramatically the take-up rate of connections offering at

least 100 Mbps is progressing in countries where VHC networks are widely available.

Figure 10: Increase in subscriptions offering at least 100 Mbps connectivity in EU and in

selected countries (fixed broadband)

Fixed broadband subscriptions to at least

100 Mbps at EU level

50%

45%

40%

35%

12%

10%

Jan-11

Jan-12

Jan-13

Jan-14

Jan-15

Jul-10

Jul-11

Jul-12

Jul-13

Jul-14

Jul-15

30%

25%

20%

15%

10%

Jul-10

Jul-11

Jul-12

Jul-13

Jul-14

Jan-11

Jan-12

Jan-13

Jan-14

Jan-15

Jul-15

Source: Communications Committee

Concerning mobile connectivity, the deployment of 4G has continued to increase sharply

after a very slow start in many Member States and is now available from at least one operator

to 86% of homes; however, some Member States lag significantly behind this average, and

even in the better-served countries, away from population centres mobile data coverage

continues to be subject to very significant gaps, including along major transport routes.

kom (2016) 0587 - Ingen titel

Figure 11: LTE household coverage by at least one operator in EU countries

Source: IHS and VVA

In addition, connectivity in Europe is still overwhelmingly asymmetric, while uplink speeds

are increasingly important for services such as cloud computing.

2.2 A growing need for connectivity

The number of Internet users, devices and applications continues to grow and will generate

more and more traffic over the core networks, creating greater need for connectivity.

Mobile data traffic in Western Europe is expected to grow six fold from 2015 to 2020

. The

European Parliament report "Reforming

EU telecoms rules to create a Digital Union"

(2016)

points out a "tremendous

expected increase of mobile data traffic in Europe - from

0.98 Exabytes per month in 2015 to 7.23 Exabytes per month"

by 2020.

In addition, these high levels of anticipated mobile data traffic are still expected to fall well

below projected traffic to fixed locations, given that already in 2012 the cellular mobile data

traffic was identified as "only

the tip of a much larger iceberg"

if one takes account of the

data used via Wi-Fi access points

.According to Cisco, the overall IP traffic is expected to

grow to 194 Exabytes per month by 2020, up from 72.5 Exabytes per month in 2015,

i.e.

compound annual growth rate (CAGR) of 22 percent

Commission Staff Working Document Europe's Digital Progress Report 2016, SWD(2016) 187 final.

ISBN: 978-92-823-8882-2

Study on Impact of traffic off-loading and related technological trends on the demand for wireless broadband spectrum

(SMART 2012/0015).

Source: CISCO VNI index, see: http://www.cisco.com/c/en/us/solutions/service-provider/visual-networking-index-

vni/index.html

kom (2016) 0587 - Ingen titel

Figure 12: IP Traffic volumes 2015-2020

Source: Cisco VNI Global IP Traffic Forecast, 2015–2020

IP traffic growth is explained by the increase in the number of users but also by the

substantial growth of IP traffic per capita over the past decade. Globally, monthly IP traffic is

expected to reach 25 GB per capita by 2020, up from 10 GB per capita in 2015

Business Internet traffic is estimated to grow 2.6-fold from 2015 to 2020, a compound annual

growth rate of 21% and, in 2020, it will reach 27.9 Exabytes per month. In particular, as

businesses and consumers exchange their data with the cloud, this will also lead to a modified

demand pattern for upload traffic. Hence, while most of the traffic will still be in download,

demand for upload will increase, as well as the need for lower latency (i.e. higher

responsiveness) for applications such as cloud computing, connected driving and e-health.

This capacity demand is fuelled by the desire of users to enjoy better quality online services

including online video and cloud applications, as well as enabling multi-screen viewing,

which is becoming increasingly prevalent in European households. Projections by Deloitte

see very high capacity connections as a requirement to meet the aggregate demand from

dozens of connected devices in a home. This is becoming the norm in European households

where several users consume bandwidth from several devices at once. Deloitte further notes

that “demand

for connectivity has evolved symbiotically: as faster speeds have become

available, the range of applications supported has increased and the viable number of

devices per person has steadily risen.”

Cisco, The Zettabyte Era: Trends and Analysis, June 2016.

Deloitte Technology, Media and Telecommunications Predictions 2016.

kom (2016) 0587 - Ingen titel

Figure 13: Europe IP Traffic and Service Adoption Drivers

Source: Cisco VNI Global IP Traffic forecast 2014-2019 – Europe includes Western Europe + CEE, excluding Russia

These trends increase the demand for capacity and other quality of service characteristics of

digital networks. There is an emerging consensus among industry players and investors that

in the medium and long run, fixed and mobile networks converge: for instance, it is expected

that 5G connectivity providers will rely on (nearly) ubiquitous VHC network infrastructures

coming very close to users' premises (i.e. to the building, to the small cell), to support their

business.

As mentioned above, this booming global traffic is the product of the increase in connected

activity at the level of individual users (both the growing number of end users and, to an ever

greater extent, the increase in data traffic per end user). Globally, the average number of

devices and connections per household and Business place is also growing due to M2M

applications, such as smart meters, video surveillance, healthcare monitoring, transportation,

and package or asset tracking. By 2020, M2M connections will represent 46 percent of the

total devices and connections, according to Cisco.

Figure 14: Traffic evolution by the type of devices 2015-2020

Source: Cisco VNI Global IP Traffic Forecast, 2015–2020

kom (2016) 0587 - Ingen titel

New and innovative (so-called "next-generation") applications requiring low latency, high

Internet access speed - often bi-directional - and other improved connectivity parameters are

emerging and will reinforce the demand for better connectivity.

Figure 15: Applications' bandwidth and latency requirements

Next generation TV is likely to be a significant driver of bandwidth demand for households

in the coming years. However, it is not the only driver. Bandwidth may also increasingly be

needed and used for business purposes, both in business premises and in the home office. For

a complete picture therefore, it is helpful to take into account a range of applications (in

addition to TV) and to understand how bandwidth requirements may vary for different types

of users including businesses.

kom (2016) 0587 - Ingen titel

Table 1: Application categories with their capacity and quality requirements 2025

Source: WIK-IDATE-Deloitte; Regulatory, in particular access, regimes for network investment models in

Europe, SMART 2015/0002

Further examples of current applications with particular quality requirements are presented in

table 2.

Table 2: Quality requirements of applications

High Connections Density

High Download Speed

High Upload Speed

High Peak data rate

Video, 3D video, UHD screens

real-time applications

Education / e-Learning

IoT & M2M

Audio / Music

Gaming

e-Commerce

IP Telephony

HD telepresence

Low Packet Loss

High Reliability

High Ubiquity

High Capacity

High Security

Low Latency

Symmetry

Low Jitter

kom (2016) 0587 - Ingen titel

e-Banking

mission critical

applications

massive communications

basic applications

Health Services

Self-driving car

Security / CCTV

e-Government

Smart City

Industry automation

Augmented Reality

Big Data

Work and play in the cloud

Gigabytes in a second

Social Networking

Smart Home / Building

Location-Based Services

Videoconferencing

Browsing

Source: European Commission

2.3 Baseline analysis: from today to 2020 and beyond

In terms of supply of NGA in commercially viable areas, forecasts from IDATE based on

market intelligence (see Figure 16) suggest that upgrades to NGA and VHC networks will

continue, but at a relatively gradual pace. The reasons are notably uncertainties in the market

regarding the rate of pick-up (demand in subscriptions), the materialization of new services

and applications, the durability of intermediate technical solutions based on legacy copper

infrastructures, arguably underpinned by strategic profit-maximizing considerations which

can favour delaying such a transition at the operator level, even at the cost of beneficial

externalities for society as a whole.

kom (2016) 0587 - Ingen titel

Figure 16: - Projected take-up of NGA by technology (to 2025)

Source: WIK-IDATE-Deloitte; Regulatory, in particular access, regimes for network investment models in

Europe SMART 2015/0002

IDATE projections above suggest that by 2020, even under very optimistic assumptions

(assuming FTTC/VDSL delivers 100Mbit/s in practice

), around 16 countries may miss the

DAE objective of 50% households taking up at least a 100 Mbps connection, and within the

16 affected countries the objective may be missed by around 25m households. In reality other

advanced hybrid copper-based solutions may deliver the required speed provided the local

loop is sufficiently short

. Countries with limited competition between legacy networks

(copper lines / coaxial cables), such as Italy and Greece, are included amongst those

considered likely to miss the objectives (though there has been a fairly recent spurt in activity

in Italy), while incountries which have been characterised by strong FTTC coverage could

fail to meet objectives under the stricter assumption that FTTC technologies may not in all

cases meet the 100Mbit/s objective

This pace of development may be sufficient to meet the needs of some users, but is likely to

limit the potential for more demanding users, including small business and home office users

and may not be sufficient to enable Europe to fully benefit from a connected economy and

society. As explained in more detail in Chapter I of the Study "Support

for the preparation of

the impact assessment accompanying the review of the regulatory framework for e-

communications",

SMART 2015/0005, the demand for data is booming and the scenarios

considered are mostly rather conservative.

Rural NGA deployments vary across the EU and within Member-States, as shown in various

case studies. If the current varying practices remain, the current status of uneven rural

deployment is likely to persist, resulting in patchy access in rural communities to broadband

capable of reaping the benefits from the social and economic integration that digitisation may

If FTTC/VDSL is excluded (as this technology is less likely than the other technologies considered to be offered at speeds

of 100Mbps and above), then only between 42% and 45% of all households in Europe would subscribe to 100 Mbps-

capable networks in 2020.

See figure 26 in Technological developments' Annex.

For additional deployment forecasts see, SMART 2015/0002.

kom (2016) 0587 - Ingen titel

bring. This process is likely to have repercussions on public finances, especially if

accompanied by ageing population. Challenge areas could in theory be addressed through

public subsidies, but these are by no means sufficient.

2.4 The gap between bandwidth demand and network capacity deployed

In Asia, affordable very high capacity (Gigabit) connectivity has already been available as a

consumer service in Japan

, Singapore and Korea for some years. In 2014, Korea’s SK

Telecom announced trials of 10 Gbps

and the Korean National Broadband Plan (Ultra

Broadband Convergence Network

), already launched a 1 Gbps objective in 2010.

In the US, very high capacity (Gigabit) connectivity is also available to households and small

businesses, notably in the cities served by Google Fibre,

and recent reports suggest that

AT&T is responding to the competitive challenge with more widespread urban Gigabit

deployments of its own

KDDI launches Gbps service 2008 http://www.japantoday.com/category/technology/view/kddi-to-launch-1gbps-fiber-

optic-service-in-oct

SK Telecom showcases 10Gbps service http://www.businesskorea.co.kr/english/news/ict/6789-100x-faster-internet-sk-

broadband-offer-10-gbps-internet

UNESCAP

http://www.unescap.org/sites/default/files/4.1%20Korean%20Broadband%20Policies%20and%20Recommendations.pdf

https://fiber.google.com/cities/kansascity/plans/

See for example http://www.latinpost.com/articles/101338/20151210/google-fiber-vs-att-gigapower-likely-to-win-gigabit-

race-thanks-to-google.htm

kom (2016) 0587 - Ingen titel

Figure 17: Percentage of FTTB connections on total subscriptions (OECD)

Percentage of fibre connections in total broadband subscriptions, June 2015

Japan

Korea

Sweden

Estonia

Norway

Slovak Republic

Iceland

Portugal

Slovenia

Denmark

Turkey

Spain

Hungary

Czech Republic

Switzerland

Netherlands

Luxembourg

United States

Mexico

New Zealand

Australia

Canada

Italy

Poland

Chile

France

Finland

Austria

Germany

Ireland

Belgium

Greece

OECD

Latvia

Colombia

10%

20%

30%

40%

50%

60%

70%

80%

Figure 17 illustrates the state of transition from copper to fibre inside and outside of the EU.

In the European Union, some Member States, such as Latvia, Sweden or Estonia already

compare well with Japan on a range of NGA metrics (although Swedish fixed rural coverage

remains relatively limited),.

Several other EU countries, including Portugal, Spain, France, Romania, which benefit from

an expanding FTTH/B footprint, albeit at different pace of deployment, may become

Europe’s leading countries for VHC connectivity in the years to come

. However, large

European countries which have so far been experiencing limited or incremental NGA

deployment may lag behind European and global leaders on VHC broadband.

Although the picture does not take into account the effect of cable subscriptions, it gives an

idea of the different pace of this transition. Furthermore, rural NGA coverage has been

increasing slowly in several countries such as Germany, France, Italy, Austria and Finland,

increasing the risk of a growing urban and rural digital divide as can be seen in Figure 18.

See the Study "Support for the preparation of the impact assessment accompanying the review of the regulatory

framework for e-communications", SMART 2015/0005 and, Study SMART 2015/0002

kom (2016) 0587 - Ingen titel

Figure 18: – Next generation access (FTTP34, VDSL and DOCSIS 3.0 cable) coverage, June

2015

Source:

IHS and VVA

- Digital Scoreboard – Connectivity section

The size of the gap between gradually upgraded network infrastructures and the exponentially

growing usage depends on several factors, such as (1) whether future demand can be met

through incremental upgrades of existing copper and coaxial (cable) networks or only

through FTTH/B; and (2) the extent to which future wireless technologies (5G) will be able

to rely on fixed networks for backhaul and other data transmission needs.

The size of Europe’s bandwidth challenge can be seen most vividly by comparing where we

are today with what would be needed to benefit from all aspects of a connected society in

2025. For instance, Prof. Brett Frischmann observed that current demand expressed by end-

users may fail to reflect the innovation potential in the market, which could be unlocked

through more performant infrastructure

According to the Samknows Survey, average download speeds achieved in Europe in 2014

were 24Mbit/s.

If investment in NGA technologies continues at its current levels, IDATE

has projected that average download speeds would reach around 185Mbps by 2025,

while

upload speeds would reach around 84Mbps. Based on trends in video and cloud usage under

the ‘status quo’, IDATE has also estimated that bandwidth use in the EU may expand from

62GB per line per month in 2025 to 303GB per line.

This may seem a significant

improvement for households used to experiencing restricted bandwidths

but may not be

enough to enable home and business users to benefit from future technological and service

innovations.

FTTP – fibre to the premises – includes fibre to the home and fibre to the building projects. The term premises includes

residential houses, as well as apartment houses. See annex for further reference on the technologies..

Source:

https://ec.europa.eu/digital-single-market/en/download-scoreboard-reports

See the Expert Panel conducted under SMART 2015/005 –Annex 13 for more details.

Page 115 Samknows for EC Oct 2014 Quality of Broadband Services in the EU.

In the context of SMART 2015/0002, IDATE/WIK/Deloitte forecast a likely uptake of NGA by technology to 2025 and

based speeds and speed growth per technology on the basis of Samknows data. According to Akamai speed measurements,

average speeds have been increasing by 16% per annum across a range of geographies. Extending this projection would

result in speeds of around 150 Mbps in 2025.

See Study SMART 2015/0002

Many Internet users are already experiencing challenges with the bandwidth they have available. Almost four in ten

respondents to the Eurobarometer survey of 2014 noted that they had experienced difficulties accessing online content or

applications as a result of insufficient speed of download capacities.

kom (2016) 0587 - Ingen titel

If bandwidth needs are calculated on the basis of what might be required to run certain

applications, a case study of the German market providing a forecast for 2025 suggests that

an average user might require 150-500Mbit/s downstream with more than 100Mbit/s up,

while high-end users including those running small or home offices might require 1Gbit/s in

download and more than 600 Mbps in upload (see SMART 2015/0005). This bandwidth

would be used not only for multi-screen ultra HD video, but also for applications such as

cloud and e-health as well as for teleworking and small business needs.

Figure 19: - Model of market potential – Germany 2025

As shown in Figure 19, data rates required by the most demanding users could reach 1 Gbps

or more on the downstream link by 2025, while a significant proportion of households and

offices could demand download speeds of 500-1000Mbit/s and 300-600Mbit/s upstream by

2025. This scenario therefore sets the upper bounds for potential users (including many

business users) demands in the medium term, while it is worth noting that even a less

ambitious scenario will need the VHC rollout to reach far deeper into most of the present

networks. In addition, it has to be assumed that middle- and larger sized enterprises in the

digital economy will have much greater simultaneous up- and download needs.

There is evidence suggesting that in the telecom sector

demand responds to supply

and that

limited download and upload speeds may limit the types of usage and applications that might

otherwise emerge.



Data from the UK regulator Ofcom for example suggests that download bandwidth

consumption for NGA (FTTC and FTTP) networks was around two times higher than

bandwidth consumption for non-NGA networks, with significantly higher use of

upload capacity.

This evidence of higher usage being associated with the availability of NGA is

supported by the case of Palaiseau in France, which has been the subject of a pilot



kom (2016) 0587 - Ingen titel



trial for the switch-over of Orange copper customers to FTTH networks: it was

observed that the average Internet traffic of Orange’s broadband customers was

multiplied by a factor of three. Importantly, this trial also resulted in fibre clients’

usage of upload bandwidth being increased 8 times

In Sweden, following an early boost by the central government, one out of every two

municipalities is involved in VHC to the business and VHC to the home deployments.

This has led to very high take-up: as of July 2015, 68% of the broadband connections

in Sweden are NGA

achieved predominantly through FTTH and FTTB connections.

Where FTTH is widespread, extending VHC to base stations is far more feasible and

efficient. This is well illustrated by the example of 4G in Stockholm where the

world’s first 4G deployment took place helped by the virtually 100% VHC

coverage.

Although all technologies are likely to continue to improve, the gradual replacement of the

medium over which connectivity is provided appears inevitable. Leveraging the speed of

light, fibre has an efficiency range for delivering high quality, symmetrical connections of

dozens of kilometres as against 250 meters for the most promising copper developments.

While existing copper-based infrastructure has the advantage of wide territorial reach in most

of the Union, continuous reliance on this infrastructure in all but the very outer reaches of the

network (e.g. within multi-dwelling buildings) may limit the availability – and possible take-

up - of applications which demand the highest quality of connectivity, retarding

developments necessary for the digitalisation of European industry, access to cost-reducing

cloud services for SMEs and spontaneous, game-based skill development for the new

generation.

Europe’s gaps in high speed broadband connectivity are likely to have a significant impact on

productivity and growth: econometric analysis conducted by WIK-Consult/Ecorys, VVA

suggests that broadband speed is positively correlated with Total Factor Productivity across a

range of industrial sectors, while external studies have also identified positive economic

effects from Gigabit technologies.

There is also evidence that a persisting digital divide

between countries and regions of Europe may affect migration, employment and social

inclusion.

In line with literature pointing out that an increase of 10 percentage points in standard

broadband penetration could contribute between 0.25% to 1.38% to GDP growth

a small,

http://www.fibre-systems.com/news/story/orange-plans-full-fibre-coverage-9-french-cities-2016

Source: Vodafone’s call for the Gigabit Society, Dec. 2015

Study supporting the Impact Assessment for the Review of the Framework for electronic communications.

A Study by Analysis Group found that 14 broadband communities which benefited from gigabit connectivity enjoyed

approximately $1.4 billion in additional GDP when gigabit broadband became widely available (Early Evidence Suggests

Gigabit

Broadband

Drives

GDP,

David

Sosa

2015,

http://www.analysisgroup.com/uploadedfiles/content/insights/publishing/gigabit_broadband_sosa.pdf ). Forzati and

Mattsson (2013) estimate the benefit of fibre installation in Stockholm by Stokab at 16 billion SEK.

For example, Forzati and Mattsson (2012) found that a 10% increase in the proportion of the population living within 353

metres from a fibre connected premise corresponds to a positive change in the population after three years. Xiong (2013)

also found that a higher fibre penetration of 10% at workplaces and 13% at residential places in a municipality in 2007

lead to a 0.17% improved population evolution between 2007 and 2010.

Among others: Crandall, R., Lehr, W., and Litan, R. (2007), The Effects of Broadband Deployment on Output and

Employment: A Cross-sectional Analysis of U.S. Data,

Issues in Economic Policy,

6; Czernich, N., Falck, O., Kretschmer

T., and Woessman, L. (2011), Broadband infrastructure and economic growth,

Economic Journal,

121(552); Koutroumpis,

P. (2009). The Economic Impact of Broadband on Growth: A Simultaneous Approach,

Telecommunications Policy,

33;

kom (2016) 0587 - Ingen titel

but expanding body of literature highlights how the effects of faster broadband through VHC

connectivity could boost growth further and offer a new lease of life to rural communities

3. Cost and benefits of very high capacity connectivity

3.1 Costing the networks for a Gigabit Society

3.1.1

Costing the gap and the financial endowment of current initiatives

Some studies have tried to estimate the NGA broadband gap in Europe and to provide

estimates about the cost to fill it. The best known of these studies is probably the one

performed by the European Investment Bank in 2011. The study considers four scenarios for

broadband deployment in Europe. The most

ambitious scenario foresees FTTH/B

roll-out

throughout Europe and the gap was estimated at €221 billion

The same scenario of 100% FTTH/B coverage was analysed by Analysys Mason in a study

for DG CONNECT in 2012

. The amount foreseen is similar (€250 billion, for deployment

of FTTP-only, across Europe). The amount is reduced to €154 billion in case of high duct re-

use. Analysys Mason also estimated the costs associated to a 100% FTTC deployment which

are in the area of €50 billion. In case of high duct re-use, the cost would go down to €31

billion.

Studies suggest that the costs are justified by the benefits. Analysys Mason's study

showed

that the bigger is the intervention, the higher is the consumer surplus.

Table 3: NGA investments and associated benefits under three scenarios

Source: Analysys Mason 2013 Study on the Socio-economic impact of bandwidth

Qiang, C. Z., and Rossotto, C. M. (2009), Economic Impacts of Broadband, In

Information and Communications for

Development 2009: Extending Reach and Increasing Impact,

35–50.Washington, DC: World Bank.

See Study SMART 005/2015

http://www.eib.europa.eu/attachments/efs/eibpapers/eibpapers_2011_v16_n02_en.pdf

Analysys Mason, The socio-economic impact of bandwidth (2013).

ibidem

kom (2016) 0587 - Ingen titel

An internal estimate carried out by DG CONNECT in 2014 on the basis of the Analysys

Mason study of 2013 showed that Europe needed an additional €34 billion in investment to

reach the objective of 100% coverage at 30 Mbps, and an additional €92 billion to credibly

enable reaching the 50% take-up objective at 100 Mbps

. These figures already take account

of the amount that the private sector could be expected to invest

and would leave part of the

network below the performance levels required to serve a Gigabit society if substantial

copper-based parts of the networks (e.g. beyond the cabinets) were to be durably maintained

thereafter.

The

financial resources available

at the European level are certainly not sufficient to meet

the challenge presented above and need to be focused on mobilising more private and

national public investments.

he allocation of

European Structural and Investment Funds

(ESIF) for high speed broadband networks experienced a sharp increase from €2.7 billion in

2007-2013 to around €6 billion for 2014-2020 (about €5.1 billion ERDF and an estimated

€0.9 billion EAFRD)

. Most of this investment is expected to be made in the form of grants.

However, the Communication on the Investment Plan for Europe called for a doubling in the

use of financial instruments under ESIF

, including an indicative target of 10% of support in

the field of ICT. Overall, with the leverage effect of ERDF and EAFRD on public (national

and/or regional co-funding) and private co-funding, it is expected €9-10 billion will be

invested in broadband during the 2014-2020 programing period, with the target to provide

high speed connectivity to an additional 14.6 million households and make 33% of rural

population benefit from new or improved ICT services or infrastructures.

The Connecting Europe Facility (CEF) in the digital area is endowed with a limited budget of

€150 million to support deployment of state-of-the-art broadband infrastructure, based on the

provision of financial instruments via the European Investment Bank (EIB). The broadband

component of CEF is expected to mobilise around €1 billion

Finally, the European Fund for Strategic Investment (EFSI) does not have sectorial

earmarking, hence it is difficult to anticipate how much broadband infrastructure investment

will be facilitated by it.

Based on a 75% coverage assumption.

According to the Digital Agenda Scoreboard, telecom (including fixed, integrated and mobile-only) CAPEX in Europe

was € 43 billion in 2013. CAPEX figures remained relatively stable over the 2011-2014 years despite the fact that in the

same period NGA coverage increased from 29% to 68%. In 2014, Mobile CAPEX spending represented 59% of total

spending. However, this CAPEX is not only directed at modernising the network, so the part which is spent on increasing

coverage in the coming years might be subject to the operators' strategic priorities being it, among others, network

expansion, modernization or cost optimization. On top of it, operator expenses depend on the evolution of the market

trends and the investment climate.

An estimate as the Commission cannot differentiate between allocations foreseen in EAFRD for ICT and Broadband as

this type of information is not requested by the regulation. However, additional information is requested and will be

provided in the context of monitoring activities (in particular, monitoring will be done for ''N° of operations", "Population

benefiting from new or improved IT infrastructure" differentiating here between "Broadband" and "Other than

broadband"). See for more information the

ICT Monitoring Tool

http://s3platform.jrc.ec.europa.eu/ict-monitoring

See COM(2014) 903 final, p.10

https://cohesiondata.ec.europa.eu/themes/2

Under the pilot phase of the Europe 2020 Project Bond Initiative, the EIB and the Commission closed in July 2014 the

first deal on a broadband project bond (in France – Axione is the beneficiary). The leverage factor foreseen for the

broadband part of CEF is around 7 times, so it is expected to mobilise around €1 billion. This leverage was exceeded by

the Axione deal which had a leverage factor of 14 times.

kom (2016) 0587 - Ingen titel

Analysys Mason's most recent study for the European Commission (SMART 2015/0068)

"Costing the new potential connectivity needs" analysed 6 scenarios, the outcomes of which

and the necessary investments are presented in table 4.

Table 4: Description of costing scenarios with their price tags in the recent Analysys Mason

study

Gigabit connectivity for socioeconomic drivers

Scenario A

Most important e-health, e-

education and e-

government players as well

as medium sized SMEs

486 thousand medium sized SMEs (50-249

employees),

210 thousand local authority buildings,

110 thousand hospital and doctor's surgeries;

210 thousand primary and secondary

schools; half of the existing s SMEs,

hospitals and secondary school is assumed to

have already a leased line connection

All from scenario A plus:

31.5 million small and microenterprises,

10 million teleworkers and freelancers

265 thousand libraries, museums, and sites

of cultural interest

134 thousand post offices and police stations

Ubiquitous mobility

Scenario F1

Scenario F2

Scenario F3

Railways

Railways and motorways

Railways, motorways and

state roads

Railways, motorways, state

and provincial roads

Currently 75% covered

Railways, motorways (75% coverage), state

roads (50% coverage)

Railways, motorways (75% coverage), state

and provincial roads (50% coverage)

€5.2 billion

€6.7 billion

€28 billion

€46 billion

Scenario D

All big and small

socioeconomics drivers

such as teleworkers,

professionals, small and

micro enterprises and most

important cultural sites

€149 billion

Scenario F4

€103 billion

Improved connectivity for rural areas

Scenario B

Residential coverage for all

the population with

wireless connectivity above

100 Mbps (macro cells)

Residential coverage for all

the population with 1 Gbps

wireless connectivity

Base station radius: 4km;

active equipment cost €20 000 per site for

interim upgrade and EUR40 000 per site for

full upgrade

Base station radius: 200 m; active equipment

cost €1000 per cell

€79 billion

Scenario C

€143 billion

For medium SMEs, hospitals and secondary schools only half are supposed to need connection as the other half are

assumed to be using leased lines.

For the Scenario F all the calculations are based on the price per length of the transport corridor calculated by reference to

the amount of connected base stations required.

kom (2016) 0587 - Ingen titel

(micro cells)

Scenario E

Residential coverage for all

the population with 1 Gbps

fixed connectivity

1Gbps access available to all residential

customers; 81% of them are assumed to take

up services

€183 billion

Figure 20 shows the total standalone costs for each of the above-mentioned scenarios.

Figure 20: Summary of the scenarios: Total costs split by technology and commercial

viability).

300

Total investment cost (EUR billion)

250

200

150

119

100

B, Macro

mobile

197

249

172

140

103

E, Residential F, Transport

coverage

links

Total

A, Large

SEDPs

C, Small cell

mobile

D, Small

SEDPs

Fibre commercial

Fibre remainder

Wireless

For scenarios B and C costs reflect 95% population coverage. Source: Analysys Mason, 2016

Despite the investments already undertaken, there is still a long way to go to complete any of

the scenarios as to date at most roughly 25% of the total sum has been invested, as shown in

Figure 21.

kom (2016) 0587 - Ingen titel

Figure 21: Summary of the scenarios: Investment already undertaken.

For scenarios B and C costs for 100% coverage. Source: Analysys Mason, 2016

3.1.2

The trajectory of private infrastructure investments

The Commission estimates, on the basis of trends throughout the decade to date, an annual

investment by operators of €40 billion that should reach a cumulative investment of €360

billion by 2025. This baseline estimate is based on the projected overall capex for the period,

based on continuation of most recent observable trends, and corrected on the basis of

assumptions on non-network-related capital expenditure (e.g. terminal equipment such as

routers /set-top boxes).

This calculation is in line with the growing interest from the private telecom sector in

investing in the network. In November 2014, 10 CEOs of the main telecom and

manufacturing companies committed to invest €150 billion in CAPEX in Europe over five

years, to play a major role in the Commission's agenda for investment, growth and

employment

This estimate also takes into account that a share of the investment required will be common

to all proposed options and that certain synergies can be expected. For instance backhaul

high-speed networks can be used across scenarios, and high–speed Internet access to

socioeconomic drivers resulting in fibre deepening diminishes also the cost of connecting

households in the same area. The exact amount of the spending saved through synergies

depends on the scenarios selected and increases with the amount of investment.

3.2 The importance of defining objectives for broadband

In 2010 the Digital Agenda for Europe (DAE) defined objectives for connectivity by 2020: it

introduced objectives of universal availability at 30 Mbps to ensure territorial cohesion in

Europe and a penetration objective of 100 Mbps (subscription by at least 50% of European

households) to anticipate future competitiveness needs.

"Make the Net Work" initiative of Alcatel-Lucent, Deutsch Telekom, Ericsson, Liberty Global, Orange, Telecom Italia,

Telefónica, Telenor, TeliaSonera, Vodafone.

kom (2016) 0587 - Ingen titel

These objectives have progressively become a reference for public policy. In retrospect, the

DAE objectives represented a sea change for European digital policy. Despite a lot of early

scepticism, in particular from the private sector, the 2010 objectives succeeded in formulating

the level of ambition and providing direction, consolidating first public and then increasingly

private investment plans around the path towards 2020. The objectives were taken up as a

reference point under the rules and guidelines in both the European Structural and Investment

Funds and the Connecting Europe Facility (CEF Broadband), as well as under the Broadband

State Aid guidelines.

Private sector investment plans are often adjusted to the objectives as well, and the expressed

research and innovation ambition with regard to improvement of current technologies often

refers to DAE objectives. At national level, setting objectives has become the cornerstone of

broadband deployment public policy. Every Member State has today established broadband

targets but also, on this basis and as shown in a recent study commissioned by the European

Commission on National Broadband Plans in the EU

, adopted or planned a set of funding

and regulatory measures aiming at reaching them.

The impact of objectives, in terms of setting direction and a European common approach

based on common minimum standards, is well reflected in the overview of the NGN plans

adopted by Member States. Many Member States have indeed aligned their national or

regional NGN plans to the DAE speed categories.

AteneKom: National Broadband Plans in the EU -28 Smart 2014/007.

kom (2016) 0587 - Ingen titel

Figure 22: Overview of the degree of alignment of NBP with DAE objectives: Map

kom (2016) 0587 - Ingen titel

Figure 23: Overview of the degree of alignment of NBP with DAE objectives: Table

Member State

NBP-objectives

National vs

Coverage

objective

N.A.

National vs

take up

objective

N.A.

Austria

Belgium

Bulgaria

Croatia

Cyprus

Czech Republic

Denmark

Estonia

Finland

France

Greece

Germany

Hungary

Ireland

Italy

Latvia

Lithuania

Luxembourg

Malta

Netherlands

Poland

99% coverage with 100 Mbps by 2020

50% connections with 1 Gbps by 2020

100% coverage with 30 Mbps by 2020

50% of households and 80% of businesses

subscribing > 100 Mbps by 2020

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 MBps service by

2020

100% coverage with 100 Mbps download and 30

Mbps upload by 2020

100% coverage with 30 Mbps by 2020

60% coverage with 100 Mbps by 2020

99% of all permanent residences and offices

should be located within 2km of an optic fibre

network or cable network that enables connections

of 100 Mbps

100%coverage with 100 Mbps by 2022

100% coverage with 30 Mbps by 2020

50% coverage with 100 Mbps by 2020

100% coverage with 50 Mbps by 2018

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

100% coverage with 30Mbps by 2020

50% HH penetration with 100Mbps service by

2020

100% coverage with 30 Mbps by 2020

85% HH coverage to reach 50% penetration of

100 Mbps services by 2020

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

100% coverage with 30 Mbps by 2020

50% coverage with 100 Mbps by 2020

100% coverage with 1 Gbps by 2020

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

100% coverage with 30 Mbps by 2020

N.A

N.A.

kom (2016) 0587 - Ingen titel

Portugal

Romania

Slovakia

Slovenia

Spain

Sweden

United

Kingdom

50% HH penetration with 100 Mbps service by

2020

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

80% coverage with 30 Mbps by 2020

45% HH penetration with 100 Mbps service by

2020

100% coverage with 30 Mbps by 2020.

96% coverage with 100 Mbps

4% coverage 30 Mbps by 2020.

100% coverage with 30 Mbps by 2020

50% HH penetration with 100 Mbps service by

2020

90% coverage with 100 Mbps by 2020

95% coverage with superfast broadband by 2017

N.A.

Source: National Broadband Plans in the EU -28 Smart 2014/007; AteneKom

kom (2016) 0587 - Ingen titel

4. Strategic objectives for 2025

4.1 The importance of setting realistic political guiding objectives

The Europe 2020 Strategy

set five ambitious objectives - on employment, innovation,

education, social inclusion and climate/energy - to be reached by 2020. Each Member State

has adopted its own national targets in each of these areas, underpinned by concrete actions at

EU and national levels.

Political objectives – also sometimes called targets - are recognised as a clear signal of

priorities to be achieved and as a way of influencing decision-making – by both public and

private players – on a continuous and homogeneous basis. By themselves they cannot resolve

the challenges related to achieving them; additional instruments – policy, funding and

legislation, have to be used so that they are fully effective. Well defined objectives can

nonetheless have an early effect on private investor decision making if there is a clear public

policy commitment to achieve them, by mobilising public support through available

instruments, and avoiding public policy interventions which would be counter-effective.

An increasing number of EU Member States is already focusing on new objectives beyond

those set in the Europe 2020 Strategy, notably with a view to providing businesses and the

public sector with Gigabit connectivity. In Bulgaria, structural funds help connecting

community centres. In Poland, all schools will be connected with at least 100 Mbps, and

some bigger ones even with 1 Gbps by 2017. Similarly, in Ireland, 750 schools have been

connected with 100 Mbps in 2014 and the program is continuing.

Outside the EU, in South Korea, the Giga Korea project has the ultimate goal of building a

digital information distribution infrastructure for a Gbps fixed-line and mobile connection

network covering all people by 2020

. The US have the ambition that every American

community should have affordable access to at least 1 Gbps broadband service to anchor

institutions such as schools, hospitals and government buildings by 2020. Russia plans to

make 100 Mbps available to 80% of Russian residents by 2018.

Against this background the European Commission has identified 3 strategic objectives for

2025 that will complement those set up for 2020:

Gigabit connectivity for all main socio-economic drivers such as schools, transport

hubs and main providers of public services

as well as digitally intensive enterprises.

All urban areas

and all major terrestrial transport paths

to have uninterrupted 5G

coverage.

As an Intermediate objective for 2020, 5G connectivity to be available as a fully-

fledged commercial service in at least one major city in each Member State, building

on commercial introduction in 2018.

Communication from the Commission EUROPE 2020; A strategy for smart, sustainable and inclusive growth,

COM(2010) 2020 final

Electronics

and

Telecommunications

Research

Institute

(ETRI)

President

Kim

Heung-nam;

http://newsworld.co.kr/detail.htm?no=436,

Covering: e.g. primary and secondary schools, train stations, ports and airports, local authority buildings, universities,

research centres, doctors' surgeries, hospitals and stadiums.

As per definition:

http://ec.europa.eu/eurostat/statistics-explained/index.php/European_cities_%E2%80%93_the_EU-

OECD_functional_urban_area_definition.

Motorways, national roads and railways, in line with the definition of Trans-European Transport Networks.

kom (2016) 0587 - Ingen titel

All European households, rural or urban, to have access to Internet connectivity

offering a downlink of at least 100 Mbps, upgradable to Gigabit speed.

4.2 Leveraging mutually reinforcing objectives and existing infrastructures

Demand developments (see section 2.2.) confirm the virtuous dynamic between availability

and take-up of VHC broadband services. Given the very slow start of take-up for 100 Mbps

services in Europe, the strategic objectives for 2025 need to a) maximise the number of

Europeans who have access to VHC networks for a given cost as well as b) enable the most

innovative and beneficial connected services to be used throughout the Digital Single Market.

The new strategic objectives proposed for 2025 are therefore mutually reinforcing, focusing

on entities which are best placed to further stimulate the demand for and use of online

services and applications and on encouraging synergies in network expansion. They all

require the deployment of VHC networks.

The deployment of VHC fixed networks will also contribute to the backhaul needs for the

upcoming dense 5G wireless network deployments as close as possible to the end-user, thus

strengthening the convergence between mobile and fixed networks.

The new objectives for 2025 build on the DAE objectives, in particular on the 100 Mbps

objective. Indeed, the efforts to reach the 100 Mbps DAE objective will contribute to the

achievement of the objectives for 2025.

The new strategic objective to reach all European households with 100 Mbps builds on and

reinforces the 100 Mbps DAE objective (50% of households having 100 Mbps subscriptions

or higher). It is very likely that particularly rural area networks built to reach the 30 Mbps

DAE objectives in 2020, will be able to deliver 100 Mbps in 2025, as it is economically

advantageous for mobile operators to expand networks coverage into rural areas (e.g. by

serving 5G base stations with the same backhaul).

In addition, reaching the DAE 100 Mbps take-up objective can only be done with far-

reaching deployment of VHC networks. Reaching the first and second strategic objectives,

can thus benefit from using such deployments, bearing in mind that at least 75-80% of the

population has to be covered with such networks for at least 50% of European households to

actually subscribe to 100 Mbps services.

4.3 Benefits of the

strategic objectives for 2025

4.3.1

Gigabit connectivity for all main socio-economic drivers such as schools,

transport hubs and main providers of public services as well as digitally

intensive enterprises

This objective will optimise investment in VHC networks by connecting with priority socio-

economic drivers, i.e. physical places or online hubs where people gather or which they visit

to learn, to work and to access public services and where a single connection provides

Internet to multiple users. Providing European socio-economic drivers with VHC

connectivity by 2025 will have multiple benefits for the DSM, in particular by:

Encouraging such institutions to subscribe to very high-capacity connectivity (Gigabit,

symmetrical, low latency, etc.) that will enable the use of the best products, services and

applications and to provide the best service to European citizens. This in turn, creates a

market for such online services. The Commission estimates that almost 100 million pupils

kom (2016) 0587 - Ingen titel

and students, more than 70 million workers as well as almost 2 million doctors and more

than 2.5 million patients in hospitals across Europe will benefit directly

Contributing to and stimulating the extension of very high capacity networks and the

densification of mobile coverage, including for 5G backhaul needs, would benefit from

proximity of very high capacity interconnection and backhaul opportunities. This will in

turn increase the chances for advanced users, including in particular small and micro

businesses and home office users, to be served by VHC connections.

Creating a spill-over effect on demand for better connectivity in the rest of the economy

(business ecosystems) and in the population at large (households), since users who

experience very high capacity networks are more likely to subscribe to such services

when they become available

. Higher take-up will in turn improve the economic case for

further investment, as the higher penetration of services results in more users sharing the

same connection and the cost of the investment, leading to a decrease of the cost per

user..

The Commission services have considered, on the basis of the findings of Study SMART

2015/0068

, two main options for this strategic objective:

A. Providing all the significant socio-economic drivers (in particular schools, transport hubs

digitally intensive enterprises, local authority buildings

) with Gigabit connectivity.

According to the cost estimates, the investment needs would be approximately €46

billion.

B. Providing all the socio-economic drivers without exception (in particular, in addition to

the significant socio-economic drivers: micro and small enterprises, libraries and

museums) with Gigabit connectivity. According to estimates, the investment needs

amount to €149 billion

Option A has been chosen, primarily on cost-effectiveness grounds. By bringing Gigabit

connectivity to e.g. over 200.000 schools, over 200.000 public authorities buildings and

roughly half a million digitally intensive enterprises

, it is already fully relevant to stimulate

the extension of further fixed and mobile networks and to create additional appetite for better

connectivity.

It is also worth noting that these efforts will complement the growing investment in the

digitalization of industry and other sectors. Digitized products and services generate

Estimate based on the findings of the Study "Costing the new potential connectivity needs" (Smart 2015/0068) by

Analysys Mason.

The socio-economic return on Stockholm municipality's Stokab investment in fibre infrastructure is estimated to be over

16 billion SEK (approx. €1.6 billion), almost three times the investment in 2013. See Figure 2.20 of the Annex in Study

SMART 2015/0002.

Study "Costing the new potential connectivity needs" (Smart 2015/0068) by Analysys Mason

The full list includes primary schools, secondary schools, universities, local authority buildings, digitally intensive

enterprises, hospitals, doctor's surgeries, research centres, business parks, airports and stadiums.

The full list includes, in addition to the large socio-economic drivers: micro enterprises, small enterprises, post offices,

police stations, libraries, community centres, home workers, farmers markets, museums and galleries.

The number of digitally intensive enterprises has been estimated by the Commission by cross referencing the estimates

from the study by Analysys Mason "Costing the new potential connectivity needs" (Smart 2015/0068) and the assumptions

underpinning the analysis of Integration of Digital Technology by enterprises in the DESI Index.

kom (2016) 0587 - Ingen titel

approximately €110 billion of additional revenues per year for the European industry. It is

estimated that by 2020, European industrial companies will invest €140 billion annually in

industrial internet applications

. The European Commission Communication on "Digitising

European Industry - Reaping the full benefits of a Digital Single Market"

proposes actions

that are expected to mobilise close to €50 billion of public and private investment in the next

5 years, but which are largely dependent on underlying connectivity provision.

In this respect, ETNO has recently underlined

that digitization is important in all public

service domains and should include: i) 100% EU schools and universities connected to at

least 1Gbps by 2030, ii) 100% administrative procedures digitised by 2025 and iii) 100% IDs

in the EU to be turned into electronic IDs or mobile IDs by 2025 (all the IDs in EU to be

digitized and e-administration enabled).

4.3.2

High performance 5G connectivity: by 2020 a fully-fledged commercial service in

at least one major city in each of the 28 Member States and by 2025

uninterrupted 5G coverage of all urban areas and major terrestrial transport

paths

The 5G objective aims at addressing Europe's future competitiveness in wireless technology.

5G services will be crucial to cope with increasingly demanding connectivity. Media

applications, professional-grade communications in various industrial and service sectors

such as automotive, transport, manufacturing, health as well as next generation safety and

emergency services will have to rely on a seamless, shared, fixed and wireless infrastructure

which offers various pre-determined levels of reliability and quality of service tailored to

specific business needs.

In Japan, 5G networks have received substantial policy attention, and there are objectives to

introduce a system trial in Tokyo in 2017, with implementation in time for the 2020

Olympics and Paralympics. In Europe Teliasonera and Ericsson have already announced that

they will launch 5G services in Stockholm and Tallinn in 2018. Building on this experience, a

fully-fledged commercial service in at least one major city in each Member State by 2020 is a

necessary but reasonable objective and a stepping stone in the process of wider deployment

of 5G.

In addition, as part of the 5G connectivity objective, the Commission aims at achieving by

2025 a 5G coverage of major terrestrial transport paths (defined as motorways, national

roads, and railways).

5G connectivity on railways is needed to address in the long term the train passengers

connectivity needs on board. According to estimates, in order to provide railways with 5G

connectivity investment needs exceed €5 billion.

In addition, 5G coverage of road transport routes will be the basis for the development of

innovative applications for connected cars. The Commission services have considered the

following scenarios

http://www.strategyand.pwc.com/media/file/Industry-4-0.pdf

COM(2016) 180 final; 19 April 2016.

new

digital

Union

What’s

for

citizens

with

IoT

and

5G",

https://www.etno.eu/datas/publications/studies/Narrative_Final_2016.pdf

Based on the findings of the Study "Costing the new potential connectivity needs" (Smart 2015/0068) by Analysys Mason.

kom (2016) 0587 - Ingen titel

A. Providing motorways with 5G connectivity. According to estimates, the investment needs

would be €1.5 billion.

B. Providing motorways and national/state roads with 5G connectivity. The investment

needs would be close to €23 billion.

C. Providing motorways, national/state roads and provincial roads with 5G connectivity. In

this case the investment needs would amount to €98 billion.

Option B

has been chosen because it would allow an EU-wide infrastructure to be

developed which is capable of supporting solutions delivering information and entertainment

on board and the progressive evolution towards fully autonomous driving, including the

delivery of security functions reducing dramatically road casualties in Europe. Option A is

considered too restrictive to achieve these goals. As to Option C (including provincial roads),

it seems at this stage too costly but could represent a natural future evolution.

4.3.3

All European households, rural or urban, to have access to Internet connectivity

offering a download speed of at least 100 Mbps, upgradable to Gigabit speed.

The challenge of comprehensive coverage objectives that concern all households lies in the

last percentages of the population, those living in the most remote areas which are the most

difficult and costly to reach. The lack of connectivity has an extremely high cost, not only by

forgoing benefits of the digital single market, but also in terms of digital divide,

depopulation, lack of cohesion etc.

Figure 24: Total cost (cumulative) of connecting countries with FTTP/H depending on the

population covered per country (Analysys Mason

)

Under this option,

or terrestrial transport paths, and depending on the considered transport service, account will be taken

of ongoing investments in C-ITS technologies while ensuring coordination with relevant stakeholders, Action 4 of the 5G

Action Plan

Study SMART 2015/0068, page 22.

kom (2016) 0587 - Ingen titel

Cumulative total investment cost (EUR billion)

50%

Population coverage

Germany

50%

Population coverage

Poland

Sweden

100%

France

Spain

Italy

Belgium

Slovenia

Romania

This objective aims at making sure that no one is left behind in the digital world. It is to be

seen against a wider ambition that there should be access to mobile data connectivity

throughout the territory, in all places where people live, work, travel and gather.

In most rural and remote areas connectivity plays an essential role in rural development and

agriculture, as well as in preventing digital divide, isolation and depopulation. It also enables

remote or rural businesses to reduce the costs of commercial activities through video-

conferencing, access to online administration, e-commerce, or data storage in the cloud.

European households, rural or urban, will require a minimum level of fixed or mobile

connectivity in 2025 much higher than it is today due to the growing needs for connectivity.

The investment needs for 100% household coverage of at least 100 Mbps will depend on the

type(s) of infrastructure deployed. According to estimates

The investment needs to provide 100% of the households with a (macro cell) wireless

connectivity offering at least a download speed of 100 Mbps amount to €79 billion.

The investment needs to provide 100% of the households with a (micro cell) wireless

connectivity offering at least a download speed of 1 Gbps would cost €143 billion.

The investment needs to provide 100% of the households with a fixed (FTTH)

symmetrical Gigabit connectivity offering at least a download speed of 1 Gbps amount to

€183 billion.

On the assumption that each type of infrastructure will account for an equal part of the future

deployment (macro and micro cells for wireless and FTTH for fixed connectivity, mostly in

urban areas), the investment needs for improved rural connectivity are estimated to amount to

€127 billion; including €24 billion to cover the last 5% (around 11 million households) with

wireless connectivity offering at least a download speed of 100 Mbps.

Based on the findings of Study SMART 2015/0068.

kom (2016) 0587 - Ingen titel

4.4 The need to prioritise investments

The proposed additional objectives for 2025 are based on a focused approach choosing

objectives that are designed to maximise societal and economic benefits relative to costs, but

continue to ensure adequate connectivity even in rural areas and at the same time widen the

approach to take into account the growing importance of the mobile/wireless dimension of

broadband. This can unlock significant additional benefits, including for the equipment

industry and for various "vertical" sectors (automotive, health, audiovisual) in which

European leadership in the connected transformation is attainable.

The Commission services have therefore examined strategic needs and corresponding

objectives of three types:



for leader institutions, with significant user numbers and an important place in

national and local economies and societies – the socio-economic drivers – which can

be expected to have especially important connectivity needs;

for wireless connectivity, by reference to the next (5G) technology generation;

for universal coverage, including of rural households.

The Commission has sought expert advice to cost some specific scenarios for 2025 defined

on the basis of the criteria described above. This includes the already mentioned study

"Costing the new potential connectivity needs"

and has been complemented by further work

by its own services.

The total estimated cost for the fulfilment of the three proposed objectives by 2025 is c. €515

billion. This entails an additional investment of €155 billion, to complement the investment

of €360 billion that can already be expected in a "business as usual" or baseline scenario from

telecommunications operators over the 2016 to 2025 period (see section 4.4).

In more detail, the individual amounts necessary to achieve the three objectives, not counting

synergies between them, are estimated as follows, based on plausible assumptions about the

proportion of underlying infrastructure needs that would in each area already be met by

"business as usual" market activity (e.g. fibre along motorways):



Gigabit connectivity for socio-economic drivers: Providing all main European socio-

economic drivers with gigabit connectivity by 2025 will require an additional €46 billion

investment on very high capacity fixed networks to connect primary and secondary

schools, local authority buildings, digitally intensive enterprises, business parks,

universities, research centres, doctors' surgeries, hospitals, stadiums, train stations, ports

and airports. This will amount

on average

to an estimated maximum additional

connectivity cost for each socio-economic driver institution of roughly €3500-4000

annually. In the case of a school of 20 classes with 25 students this would translate into an

additional annual connectivity cost of €7-8 per student,

High performance 5G connectivity along main transport corridors including railways,

motorways and national roads will require another €28 billion investment in mixed fix

and wireless networks: €5.2 billion to connect railways, €1.5 billion to connect

See Study SMART 2015/0068.

The costs are estimated assuming 15 years return on investment period, 20% profitability margin for the operator and

operational cost that are spread over all other users in a larger territory who will be served via the backhaul network that

serves the relevant socio-economic driver institution (e.g. nearby households and businesses, etc.).



kom (2016) 0587 - Ingen titel



motorways, and €21 billion to connect national roads. In comparison with the scale of the

automotive industry providing jobs for 12 million people and accounting for 4% of the

EU’s GDP (more than €400 billion) the investment is not huge and would be a logical

step in the process of innovation of the transport sector.

Improved connectivity in rural areas will require a €127 billion investment in a

combination of fixed, wireless large and small cell-based networks:

The cost per user grows exponentially as the network reaches less and less

populated areas, where the number of users sharing it is smaller. While the first

50% population coverage requires 20-25% of the total investment the last 5% are

the most costly and difficult to connect. For this reason, the last 5% of households,

roughly 11 million households, are assumed to be covered by macro cell

infrastructure only (with very high capacity backhaul). Deployment of wireless

connectivity offering the last 5% at least a download speed of 100 Mbps is

estimated to cost over €24 billion. It is clear that this investment will require

public intervention – public investment or publicly enabled investment.

For the other 95% of households, we expect a variety of approaches, between

macro and micro cells for wireless and FTTH/B for fixed connectivity. On the

assumption that each type of infrastructure will account for an equal part of the

future deployment, the investment needs would amount to €103 billion,

The total of €201 billion, being the sum of all three options, does not take into consideration

economies of scale and the synergies which could be used in combining more than one

scenario. Based on the study by Analysys Mason we assume in particular that the re-use of

the fibre networks rollout for reaching the first objective could produce up to € 46 billion

savings in the third one. Therefore the final additional investment needed would be

approximately € 155 billion.

kom (2016) 0587 - Ingen titel

Annex:

Technological developments

Which technology?

Each year more and more fibre is being laid in Europe increasing the capacity, reliability and

speed of the network. The deepening of the fibre (bringing it closer to the user) is taking

place in a growing number of densely populated areas. However, the picture is not

homogeneous. Even in more prosperous Member States, there are wide variations –

attributable in part to differences in costs, in availability of duct infrastructure, in digital

literacy and demand, but also no doubt to differences in commercial strategy – while some

poorer Member States are leap-frogging.

A key question for policy-makers as well as investors is how will technological developments

impact network performance and, in particular, i) whether future bandwidth requirements can

be met through incremental upgrades of existing twisted pair copper and coax networks or

require new investment notably in fibre networks reaching the user (FTTH/B); and ii)

whether evolution of mobile technologies may enable some degree of substitution with fixed

networks.

While in new greenfield deployments fibre will be the technology of choice, in the presence

of a network constituted of twisted copper pairs, operators must decide whether to prolong

the life of the existing network through new technologies, often replacing a limited part of the

network with fibre, or roll out an entirely new fibre network.

Twisted copper pairs may be able to cope with downlink demand for video until 2024, when

8KTV (the current highest ultra-high definition television resolution in digital television and

digital cinematography) will become available

. However the result is different in a demand

scenario taking account of cloud and small business and home office usage where customers

also have increased requirements for traffic symmetry (similar upload and download

capabilities). In this scenario, very high capacity (VHC) technologies (G.Fast very close to

the end user/FTTB, DOCSIS 3.1 and FTTH) would be needed at an earlier stage to meet the

challenge.

In due course and as an alternative to the fixed last mile, it will be most likely possible to

connect the customer with wireless high capacity networks, including wireless fixed links

which focus on serving specific premises rather than supporting mobile use throughout an

area. However, this will also require fibre connectivity to the base station, which will

translate into a few hundred meters from the user depending on the spectrum availability.

This having been said, it should be noted that no data exists that would allow the estimation

of future bandwidth needs of those sections of today's society that can be considered "digital

natives " (below 25 years old) and which are driving data consumption already today.

Overall, fully or almost fully fibre-based networks are clearly in a better position to handle

the challenges of improve connectivity parameters than their VDSL or cable competitors,

although technological evolution such as the advent of DOCSIS 3.1 may alleviate several of

the latter's constraints, though probably not offering all the benefits and characteristics

fully fibre-based networks (see section IV below). It must also be acknowledged that the

See WIK-Deloitte-IDATE Study "Access regimes for network investment and business models in Europe" SMART

2015/0002

kom (2016) 0587 - Ingen titel

quality of a fully fibred network can also vary depending on design choices of the operator,

including as regards active equipment. There is an emerging consensus among industry

players and investors that, in the medium and long run, connectivity providers, both fixed and

mobile, will have to rely on ubiquitous fibre infrastructures deployed very close to end-users'

premises (buildings, wireless small cells) to support their business.

There are also differences of view as to what the timeline should be, which may be founded

on genuinely different assessments as regards the pick-up in demand, the materialization of

new services and applications, and the progress that can be made through intermediate

technical solutions linked to legacy infrastructures – but can also be influenced by strategic

considerations of operators who currently own such legacy infrastructures and may consider

it to be profit-maximizing to delay such a transition, even at the cost of beneficial

externalities for society as a whole.

II.

What will technology deliver

The next figure illustrates that physical capabilities of the technologies or transmission media

set barriers to certain applications – some of which derive directly from the laws of physics

(the speed of light relative to the transmission of electrons). For example, the last arrow,

showing the efficiency range for technologies, illustrates that the high-speed copper solutions

give rapidly declining performance at more than 250 metres from the end-user, thus requiring

fibre to be deployed to a drop point situated at that maximum distance from the end user

premises to be effective.

Figure 25: Physical capabilities of the technologies or transmission media

kom (2016) 0587 - Ingen titel

III.

Copper based technologies

Twisted pair copper cables have been widely deployed in the past to provide telephony). In

western European countries they have reached almost 100% of homes, although in central

European countries the deployment has peaked at approx. 60% as mobile wireless access has

filled the gaps.

Internet access over twisted copper pairs has evolved over time. New technologies such as

ADSL have made possible to provide higher data rates over copper networks. Figure 26

shows the progress in the theoretical data rates of the various DSL access technologies. It is

notable that the introduction of VDSL2 vectoring is likely to raise download speed

capabilities to 100 Mbps, while G.Fast could reach 1Gbps in a 2020 timeframe. However, the

faster the DSL technology, the shorter is the distance from the drop point (where the fibre

component of the network ends) to the customer that can still be covered on the basis of an

upgraded copper network.

Figure 26: Evolution of DSL data rates, 1995-2020

Source: FttH Council Europe (2014). G.Fast. Brussels: FttH Council Europe

The practical consequence is that operators which are relying on improvements in copper

technology to drive higher speeds must ultimately deploy fibre and locate associated active

equipment ever closer to the customer

See Study "Access regimes for network investment and business models in Europe" SMART 2015/0002, Section 1.2.

Ibidem, Section 5.1.

kom (2016) 0587 - Ingen titel

Figure 27: Copper access line use migration: Higher bandwidth over shorter sub-loops

(source WIK)

Figure 28 shows the historic technological developments and speed ranges in Belgium from

Proximus and Telenet to the period 2015 and a forecast for subsequent years.

The forecast is based on the extrapolation of the 16% bandwidth growth rate observed over recent years as measured by

Akamai, this being combined with the forecast on technological capabilities According to Akamai, the average connection

speed is low because of: (1) parallel requests, whereby an average webpage generates 90 requests for content; i.e.

involving relatively small files as many components make up a webpage; each session being too short to ramp up to

maximum speed; and (2) IP address sharing, whereby multiple devices use an internet connection with an unique IP

address, with simultaneous requests sharing the available bandwidth. The average peak connection speed reflects the

highest connection speeds from each unique IP address. Thereby it is representative of internet connection capacity. It

reflects larger files, such as software updates occurring late at night. Source: Akamai (2015). State of the Internet, 4Q-

2014. Cambridge, MA.

kom (2016) 0587 - Ingen titel

Figure 28: The useful life of copper, 2025 perspective

Source: Wolter Lemstra

This scenario projects an average data rate of 50 Mbps and an average peak rate of approx.

220 Mbps downstream for the year 2025. The chart reflects the early and wide deployment of

vectoring by Proximus. With many G.Fast trials underway and the first commercial

deployments being announced, broad deployment of G.Fast could start in 2018. The

predicted capacity of FttDP

+G.fast+G.vector on copper loops of 100 meters or less is given

as at least 250 Mbps symmetrical. Under this scenario the useful life of twisted pair copper in

the cable drop and in-house cable segment would be extended well beyond 2025. However, it

should be noted that G.Fast requires the deployment of fibre very close to the end-user

(FttDP), as well as continued OPEX heavy management of the energy and maintenance needs

of active equipment at many points close to the edge of the network. It is thus not a costless

solution.

While DOCSIS 3.0

could meet projected demand up to 2020, this scenario suggests an

average peak connection rate in excess of 100 Mbps for 2020. However, within the next 5

years period Telenet will be in a position to upgrade from DOCSIS 3.0 to DOCSIS 3.1 and

hence will be able to meet – and generate - growing demand under this scenario.

Fibre-to-the-Distribution-Point – i.e. somewhere between Fibre-to-the-Node (FTTN) and Fibre-to-the-Home (FTTH).

Standard employed by many cable television operators to provide Internet access over their existing hybrid fibre-coaxial

(HFC) infrastructure.

See Study "Access regimes for network investment and business models in Europe" Smart 2015/0002, Section 1.2.

kom (2016) 0587 - Ingen titel

In conclusion, this case analysis suggests the possibility of an ‘extended life’ for copper-

based access technologies towards the 2025 horizon, but only as the "last mile", which in

practical terms means few hundred meters or less.

However, the above discussion concerns the demand on the downlink. Demand on the uplink

has traditionally been much smaller than on the down link, but a shift towards more balanced

use is expected as a result of an increase in the use of cloud services, social media and more

business usage from the customer premises. G.Fast is said to support 250 Mbps symmetrical,

while DOCSIS3.1 supports 10G/1G in a shared configuration and the ratios can be adapted to

market demand. Hence, both technologies appear to have ‘head room’ on the uplink for a

"business as usual" evolution.

In addition, beyond the traditional focus on speed, quality (in particular jitter and packet loss)

as well as responsiveness (in particular latency) of the network are increasingly relevant.

Figure 25 shows that while incremental upgrades of copper networks through technologies

such as vectoring and G.Fast can deliver higher speeds in particular in short distances (e.g. G-

fast up to 250m), fibre networks seem to be in a better position than copper-enhanced

networks to handle challenges such as symmetry between up- and download, low latency,

jitter and packet loss due to the physical parameters of the medium (e.g. electric vs optical

signal).

IV.

Optical fibre – a new generation of networks

In order to meet growing demand from the users, access with new technologies will be

necessary and the most appropriate technology mix would be a combination of the fixed and

wireless networks proving the most efficient mix at a reasonable price. Whatever technology

is selected as the access to the customer in the final mile, the core network and backhaul will

require optical fibre. Similarly, fibre deepening to a large degree is a prerequisite to all the

other technologies, which deliver high connectivity.

Optical fibre-based transmission is probably the most robust high-throughput technology

available today. It has seen wide deployment in long-distance networks following the

liberalization in the late 1980s, in support of the growing needs from the booming Internet

and mobile communication services.

Apart from output fibre networks offer quite a few beneficial characteristics:



Low latency:

the ability to support instantaneous connections and transit data without

almost any ‘delay’, measured in just Milliseconds over distances greater than

1000km, and only microseconds at distances in the 100’s of KM

Availability:

these networks are inherently stable, offering extremely high

availability

Security:

such networks, given the physical and often buried, nature of the cable, are

harder to interfere with than wireless networks (or even ‘radiating’ copper based

cables)

See for a discussion of pan-European network deployments during the euphoric period and the consolidation in the

aftermath of the crash Lemstra, W. (2006). Dissertation: The Internet bubble and the impact on the development path of

the telecommunication sector. Department Technology, Policy and Management. Delft, the Netherlands: TU Delft.

kom (2016) 0587 - Ingen titel



Packet loss & low jitter:

near zero packet loss and variance in packet delay, i.e.

“smooth signals”

Distance agnostic:

grade of service is essentially the same almost regardless of

distance

Dynamic symmetry:

such networks can be configured to allow flexibility in

assigning upload and download throughput for each connection, depending on the use

case

Low maintenance:

fibre is inherently reliable, hence little maintenance is needed

Future proof:

transmitting equipment can be easily replaced and raw medium is

future proof

No radio frequency interference:

signals travelling over fibre are not subject to

radio frequency interference versus copper which is susceptible.

Figure 29: What is throughput? – measured as “time to download*”

Digital item (examples)

Average Kindle eBook

CT scan (sent across hospitals)

Virtual reality game

Blu-ray movie

Galaxy S7 storage

4K movie

Hard disc of a PC

Medium sized corporate server restore

Human genome (uncompressed)

Typical size

2.6 megabyte

2 gigabyte

5 gigabyte

25 gigabyte

32 gigabyte

100 gigabyte

240 gigabyte

6 terabytes

7 terabytes

Legacy network

1 second

14 minutes

34 minutes

2.8 hours

3.6 hours

11 hours

27 hours

28 days

33 days

FTTH network

50 ms

40 seconds

1.7 minutes

8 minutes

11 minutes

33 minutes

1.3 hours

33 hours

39 hours

* Today’s actual effective download speeds illustrated a 20Mbps for a typical European ‘legacy

network’ (usually ADSL2+ type) and 0.4Gbps for FTTH network

Source: Creating a Gigabit Society; Arthur D. Little for Vodafone

a. Fibre-to-the-Business

As business users have higher demand than consumers and are typically located in business

districts or business parks, the business case for providing a new network technology by

digging up the streets has largely been positive. In many cases we can observe competing

fibre-based services being offered to business users. This includes the incumbents, locally as

well as from abroad, but also new entrant providers specializing in servicing the business

community.

b. Fibre-to-the-Home

Two standards are prevailing in the market place in Europe: the GPON standard providing

2.5 Gbps downstream and 1.25 Gbps upstream, shared by a maximum of 64 users; and the

XG-PON

providing 10 Gbps downstream and 2.5 Gbps upstream, i.e., a fourfold increase

http://www.vodafone.com/content/dam/group/policy/downloads/Vodafone_Group_Call_for_the_Gigabit_SocietyFV.pdf

Serving 128 customers per splitter.

kom (2016) 0587 - Ingen titel

compared to GPON. This latter arrangement provides a minimum of 78/19 Mbps per end-

user, depending on how much capacity is made available per user and reflecting shared

medium constraints in case of uniformly massive demand growth.

For point-to-point networks the ‘Ethernet in the First Mile’ (EFM) standard, used over single

mode fibre, is defined for a span of nominal 10 km. State of the art technology is able to

bridge higher distances up to 80 km.

c. Outlook for the 2025 horizon

For the PON architectures, a further upgrade is foreseen under the name of NG-PON2 with a

capacity of at least 40/10 Gbps (download/upload), for a minimum span of 20 km. Extensions

to 80/20 Gbps and 60 km are being addressed. This standard makes use of four wavelengths

and eight in the extended version. The minimum end-user data rates are 625/156 at a ratio of

1:64, respectively 312/78 Mbps at 1:256.

A combination of TDM-PON and TWDM-PON is also foreseen running at 100 Gbps

reaching 100 km with a split ratio of 1:1024.

The evolution in user data rates using fibre is shown in figure 30.

Figure 30: Evolution of FttH data rates, 1995-2025

Source: FttH Council Europe (2014). FTTH Handbook: FTTH Council Europe

d. Investing in FTTH

While DSL and DOCSIS technologies have mostly been deployed as upgrades to legacy

networks by historic incumbent and cable operators, FTTH is characterised by a more diverse

investor-base.

kom (2016) 0587 - Ingen titel

In Sweden

, which has a very low population density – 23 persons per square kilometre,

municipalities have taken the lead in the deployment of open-access fibre networks,

encouraged into that role by the central government.

The business case for these largely

rural area deployments turns positive through a combination of factors: (1) the costs for

procurement of communication services for the local government in all its facets turns from

external to internal; (2) the community network provides lower cost bundled access for

service providers; (3) at the same time providing for increased services competition; (4) the

use of civil infrastructure owned by the municipality; (5) shared interest in the execution of

the works; and (6) willingness of prospective subscribers to pay an upfront fee .

Commercial non-telecom infrastructure investors have also in some cases seen a profitable

business case in building an FTTH access network in competition with the existing copper

network(s). These cases are typically built around demand aggregation (i.e. activating

projects in an area only when a minimum threshold of interested future subscribers has been

passed), often in combination with payment of upfront fees by prospective end-users, and

novel low-cost techniques of laying the fibres. These networks are often open-access, with

competition on the services layer. A typical example is the case of Reggefiber in the

Netherlands.

The lack of broadband provision by traditional players in central Europe has led to grass-

roots initiatives by local entrepreneurs in deploying fibre to provide Ethernet-based

connectivity in apartment buildings and city neighbourhoods.

This has propelled some

central European countries, for instance Latvia and Bulgaria, to the top of the broadband

league tables in terms of peak connection data rates.

Higher deployment of FTTH can also be observed in circumstances where the costs of

deployment are lower (for instance thanks to duct availability) and in countries with

competitive pressure from alternative FTTH investors. Incumbents may also be expected to

deploy FTTH in greenfield situations, as it provides for higher data rates, is more future proof

and comes with lower operational costs.

e. Will fibre-based technologies be able to meet growing demand?

Fibre has been and still is the technology providing the highest possible data rates with the

highest and most consistent quality of service. Since the early 2000s fibre-based Internet

access has been offered at 50 Mbps and 100 Mbps symmetrical to residential customers and

into the Gbps range to business customers. If demand grows the split ratio in fibre PON

networks can be reduced. If demand would grow much further, wavelength division

multiplexing (WDM) can be introduced which makes it possible to use multiple colours of

For a description of the Stockholm municipality fibre network Stokab, the role of housing corporations, initiatives in

Swedish municipalities including rural communities and an example of regional coordination, see Section 1.3 of the Annex

in Study SMART 2015/0002, see footnote 12.

See Chapter 7

Sweden

by Forzati and Mattson in Lemstra, W. & W. H. Melody (Eds). The dynamics of broadband

markets in Europe: Realizing the 2020 Digital Agenda. Cambridge: Cambridge University Press.

See Chapter 4

The Netherlands

by Lemstra in Lemstra, W. & W.H. Melody (2015). “The dynamics of broadband markets

in Europe – Realizing the 2020 Digital Agenda”. Cambridge University Press.

See Chapter 15

Latvia

by Virtmanis and Karnitis in Lemstra, W. & W.H. Melody (2015). “The dynamics of broadband

markets in Europe – Realizing the 2020 Digital Agenda”. Cambridge University Press.) As well as. Rood, H. (2010). Very

high speed broadband deployment in Europe: The Netherlands and Bulgaria compared. Telecom Policy Research

Conference, Arlington, VA: TPRC.

kom (2016) 0587 - Ingen titel

light to expand the capacity of a single (existing) fibre. This can be accomplished by an

upgrade of the electronics at either end of the fibre.

Recent technological developments suggest that a combination of time and wavelength

division multiplexing (TWDM) can be deployed over passive optical networks (PONs) such

that the (physical) point-to-multipoint architecture can be turned into a (logical) point-to-

point architecture, combining the benefits of both architectures.

According to many sources

fibre can stay ahead of even the most optimistic forecasts in

terms of end-user demand for at least next 25 years.

Wireless technological developments

Although Moore’s Law does not directly apply to progress in performance of radio frequency

chips

, exponential growth is very clearly reflected at the systems level when performance is

expressed in peak data rates, in particular as achieved in mobile communications, Figure 31

represents a data rate increase by a factor of approx. 1.8 every year. This is essentially the

result of more capable digital signal processing. We may expect this trend to continue since

the succession of generations of cellular mobile networks shows a high degree of regularity.

Figure 31: Peak data rates in mobile, 1990-2020

Source: Niemegeers, I. & S. Heemstra de Groot (2015). Cognitive Radio+ for 5G and beyond. Eindhoven:

Technical University Eindhoven

Research done by e.g. Ericsson, FTTH Council and ITU.

As Golio observed: “In

digital systems, the fundamental unit of electrical design is the bit. Any bit will work as long as the

difference between the ‘1’state and the ‘0’ state can be distinguished as scaling reduces the size of devices, no

fundamental change in electronic design is required. In contrast, the fundamental unit of RF/wireless systems design is the

Watt. A specific amount of power is required from the transmitter in order to achieve link margin at the receiver the

required power level is not nearly as scalable as the bit.”

Source: Golio, M. (2015). "Fifty years of Moore's Law."

Proceedings of the IEEE

103(10):

1932-1937.

kom (2016) 0587 - Ingen titel

Since data communication was introduced through GPRS the data rates have increased

through different technologies. LTE Advanced was first introduced in 2013 and provides a

peak cell capacity of 1.2 Gbps, which is shared among the users. The design vision for 5G

aims for 1000 times higher overall capacity and for 10 to 100 times higher end-user data

rates. This is expected to result in guaranteed user rates of over 50 Mbit/s. For more

information on the future capabilities of 5G systems, please consult the Staff Working

Document on "5G Global developments"

accompanying the Communication "5G for

Europe: An Action Plan"

Figure 32: Features of connectivity medium and technologies

Down

upstre

Rate

)

Medi

Technol

ogies

Efficie

ncy

range

)

Typi

cal

laten

(5)

Shar

medi

for

last

mile

Freque

ncy

bandw

idth

(6)

Infrastructure

architecture

Suitability

Future

technology

the

Wired

Broadba

Technol

ogies

ADSL,

ADSL2,

ADSL2+

VDSL,

VDSL2,

Vectorin

24/1

Mbps

15-

0,0022

GHz

· internet access by

transmitting

digital

data over the wires of

a local telephone

network copper line

terminates

telephone exchange

(ADSL) or street

cabinet

(VDSL)

Vectoring:

Elimination of cross

talks

for

higher

bandwidths

· G.Fast: Frequency

increase up to 212

MHz

achieve

higher bandwidth

· coaxial cable in the

streets and buildings;

fibre at the feeder

segments

· network extensions

to provide backward

channel functionality

5 km

100

/40

Mbps

1 km

15-

0,017

GHz

copp

G.Fast

500/5

Mbps

250 m

15-

0,212

GHz

· use of existing

telephone

infrastructure

· fast to install

· small efficiency

range due to the

line resistance of

copper connection

lines

· further speed and

range improvements

by enhancing and

combining

new

DSL-based

technologies

(phantom

mode,

bonding, vectoring)

· bridge technology

towards

complete

fibre optic cable

infrastructure

CATV

200/1

Mbps

(4)

2-100

(2)

15-

yes

1 GHz

· use of existing

cable

television

infrastructure

· fast to install

· high transmission

rates

Further

implementation of

new

standards

(DOCSIS 3.1) will

allow to provide

higher bandwidth to

end-users

fibre

Optical

Broadba

Technol

ogies

Commission Staff Working Document 5G Global Developments COM(2016)XXXX final.

Communication from the Commission to the European Parliament, the Council, the European Economic and Social

Committee and the Committee of the Regions, COM(2016) XXX final, 5G for Europe: An Action Plan.

kom (2016) 0587 - Ingen titel

p2p

· signal transmission

via

fibre

distribution

signals by electrically

powered

network

equipment

unpowered

optical

splitters

p2mp

1/1

Gbps

(and

more)

10-60

0.3

ms (5

µs

per

km)

50000

GHz

yes

· highest bandwidth

capacities

· high efficiency

range

· high investment

costs

bandwidth

depends on the

transformation of

the optical into

electronic signals at

the cabinet (FTTC),

building (FTTB) or

home (FTTH)

· next generation

technology to meet

future

bandwidth

demands

Wireless

Broadba

Technol

ogies

100/3

(1000

/30)

Mbps

(3)

LTE

(Advanc

ed)

3-6

5-10

yes

0.1

GHz

HSPA

42,2 /

5,76

Mbps

3 km

30-

yes

0.005

GHz

air

20/6

Mbps

500-

700

GHz

Satellite

High

yes

Wi-Fi

300/3

Mbps

300 m

100 -

1000

yes

0.005-

0.160

GHz(7

)

· mobile devices send

and receive radio

signals with any

number of cell site

base stations fitted

with

microwave

antennas

· sites connected to a

cabled

communication

network

and

switching system

· highly suitable for

coverage of remote

areas (esp. 700 and

800

MHz)

· quickly and easily

implementable

· shared medium

limited

frequencies

commercial

deployment of new

standards

with

additional features

(5G) and provision

of more frequency

spectrum

blocks

(490 - 700 MHz)

· meets future needs

of mobility and

bandwidth accessing

NGA-Services

· highly suitable for

coverage of remote

areas

· quickly and easily

implementable

· run time latency

· asymmetrically

· 30 Mbps by 2020

based

generation of high-

throughput satellites

WiMAX

4/4

Mbps

60 km

yes

0.01

GHz

· inexpensive and

proven

· quickly and easily

implementable

· small efficiency

range

· shared medium

· increased use of

hotspots at central

places

· gets continually

replaced by Wi-Fi

and LTE

kom (2016) 0587 - Ingen titel

Glossary



FTTB – fibre to the building. Used also for buildings inhabited by more than one

tenant.

FTTC – fibre to the cabinet. Fibre to a street cabinet close to the end-user.

FTTH – fibre to the home. Typically used for residential homes.

FTTP – fibre to the premise = FTTB + FTTH.

High Capacity – capacity relates to the bandwidth of the connection and does not take

into account the distance and the time necessary to establish connection.

High Connections Density – is necessary when many devices try to connect with each

other.

High Download Speed – download speed is how fast one can pull data from others to

you. It is measured in megabits per second (Mbps). The majority of online activity,

like loading web pages or streaming videos, consists of downloads.

High Peak data rate – peak data rate is the fastest data transfer rate for a device,

typically available in short bursts during transfer activity, and not sustainable for long

periods of time

High Reliability – reliability is an attribute of any computer-related component

(software, or hardware, or a network, for example) that consistently performs

according to its specifications.

High Security – secure connection is difficult to intercept and decode without the

permission of the parties communicating, therefore the content is safe.

High Ubiquity – ubiquity means being connected everywhere, anytime, on any

device.

High Upload Speed – upload speed is how fast one sends data from you to others. It is

measured in megabits per second (Mbps). Uploading is necessary for sending files via

email, or in using video-chat to talk to someone else online (since you have to send

your video feed to them).

Low Jitter – jitter is the deviation from true periodicity of a presumed periodic signal,

often in relation to a reference clock source.

Low Latency – latency is the delay between the sender and the receiver decoding it,

this is mainly a function of the signals travel time, and processing time at any nodes

the information traverses.

Low Packet Loss – packet loss occurs when one or more packets of data travelling

across a computer network fail to reach their destination.

NGA – Next Generation Access with a minimum download speed of 30 Mbps.

Symmetry – when download bandwidth is the same as upload bandwidth.

VHC - Very high-capacity networks are networks with best-in-class performance in

terms of speed (i.e. significantly above 100.

