Celtic project CP5-013
MARCH - Multilink architecture for multiplay services
Deliverable D3.4
Market development up to 2015
Issue date:
02.12.2010
Version:
1.0
Security group:
Public
Author:
Kjell Stordahl, Telenor
MARCH D3.4
Market development up to 2015
Executive summary
The deliverable gives an overview of the broadband situation especially in Western Europe.
Western Europe is one of the most advanced telecommunication areas in the world.
The market evolution since introduction of fixed broadband in 1999 and mobile broadband
since introduction in 2006 have been analysed.
Specific attention is shown on the traffic evolution.
Multilink is a promising technology. The technology can be used to increase the traffic
capacity on specific network routes, on specific bottleneck sites, on minor areas where traffic
congestion are identified.
In fairly large fixed and mobile broadband networks there will be much maintenance and
engineering activities. Different type of failures will occur. In addition planning,
dimensioning, network extensions and network upgrading are carried out. The network
extensions will be a trade off between investments giving overcapacity for a period and too
often usage of manpower to do the capacity upgrading. In this situation the multilink
technology could be a relevant alternative.
The objective with the deliverable is to describe, analyse and forecast the future traffic growth
in the fixed and mobile broadband networks.
To be able to make good traffic forecasts, it has been necessary to make forecasts for
subscription penetration and then traffic forecasts per subscription in the busy hour. These
forecasts are put together for making total traffic forecasts during the busy hour which is the
key information for planning and dimensioning of the networks.
Short and long-term traffic forecasts are crucial for network planning, dimensioning and for
long-term strategies. However, there are definite uncertainties in the traffic forecasts which in
turn may create traffic bottlenecks and congestions in the networks.
In addition there are already traffic bottlenecks in the broadband mobile network. Especially
limited sector capacity at the base stations and limited backhaul capacity creates traffic
problems in the busy hour. So far the problems are partly solved by introduction of traffic
volume cap of 5 -10 GigaByte per month for a subscription.
The long-term traffic forecasts both for the fixed broadband network and for the mobile
broadband network in the busy hour show an explosive growth.
For instance the mobile broadband traffic in 2015 is forecasted to increase 36 times the 2008
traffic level and 16 times the 2009 traffic level.
In addition the fixed broadband traffic is estimated to be 19 times larger than the mobile
broadband in 2009, while the proportion is reduced to 6 times larger in 2015.
However, it is important to underline that there are significant uncertainties in these forecasts.
The continuous evolution of mobile broadband technologies may elevate the traffic forecasts.
1
MARCH D3.4
Market development up to 2015
Abstract
The deliverable gives a description of the evolution of the fixed and mobile broadband
market. Specific attention is put on subscription and traffic demand.
The historic evolution of fixed and mobile penetration is shown since the introduction
respectively in 1999 and 2006. Mean values for Western European countries are presented.
The deliverable gives an overview of the important fixed and mobile broadband technologies.
Without such knowledge it is impossible to make good long-term forecasts.
Separate long-term forecasting penetration model for fixed and mobile broadband have been
developed based on four parameter Logistic models.
The long-term penetration forecasts for fixed broadband show that the Western European
market within some years is close to saturation.
The market share evolution of DSL, HFC(cable modem), fiber and fixed wireless access is
described and analysed. There are significant substitution effects between technologies.
Long-term forecasting models for market share evolution of these technologies have been
developed. The long-term forecasts show that especially fiber and HFC catches significant
market shares while DSL continues to lose rather large market share.
Western Europe has been an advanced telecommunication area. However, the fiber
deployment is lagging behind compared with US, Japan and South Korea. So far EU is much
delayed in finalizing recommendation on fiber regulation.
Long-term forecasting models are also developed for mobile broadband penetration. The
market is dived in consumer and business market. Like for fixed broadband, four parameter
Logistic models are applied for the long-term forecasts.
In contrast to subscription statistics, very limited traffic statistics are available. Some traffic
statistics from the Scandinavian market are used.
The direction of the broadband traffic seems to change from peer to peer and file sharing to
entertainment/video as the most important traffic source in the future.
The mobile broadband network has definite traffic bottlenecks. The capacity of the radio part
of the base station and the backhaul from the base station to access network infrastructure
limits the traffic demand. Description is given how the new systems and additional capacity
upgrades will enhance the traffic situation.
Long-term traffic forecasts for the fixed and mobile broadband have been developed. The
traffic forecasts for the fixed network are given as busy hour traffic, both as traffic per
subscription and as total traffic.
Long-term traffic forecasts for mobile broadband have also been developed for monthly
traffic volume and traffic in the busy hour, both as traffic per subscription and as total traffic.
The traffic forecasts both for fixed and mobile broadband show an explosive traffic evolution
which need continuously upgrading of capacity and introduction of higher capacity systems.
2
MARCH D3.4
Market development up to 2015
Table of Contents
Executive summary .................................................................................................................. 1
Abstract 2
1
Background .............................................................................................................. 5
2
2.1
2.2
2.3
Broadband overview ................................................................................................ 6
Broadband .................................................................................................................. 6
Fixed broadband ........................................................................................................ 6
Mobile broadband ...................................................................................................... 8
3
3.1
3.2
3.3
Long-term forecasting models .............................................................................. 10
Forecasting methodologies for fixed and mobile broadband subscriptions ............ 10
Logistic four parameter model ................................................................................. 10
Parameter estimation................................................................................................ 10
Part A Fixed broadband ........................................................................................................ 11
4
4.1
4.2
4.3
4.4
4.5
Fixed broadband access penetration .................................................................... 12
DSL – Digital Subscriber Line technology.............................................................. 12
HFC/Cable modem – Hybrid Fiber Coax ................................................................ 16
Fiber access .............................................................................................................. 19
Fixed Wireless Accesses .......................................................................................... 22
The fixed broadband market situation ..................................................................... 22
5
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
Long-term broadband access penetration forecasts ........................................... 27
Drivers for additional broadband subscriptions ....................................................... 27
Long-term penetration forecasts for the residential market ..................................... 27
Long-term HFC market share forecasts ................................................................... 29
Long-term fiber market share forecasts ................................................................... 30
Long-term FWA market share forecasts .................................................................. 31
Long-term DSL market share forecasts ................................................................... 31
Long-term penetration forecasts for DSL, HFC, fiber and FWA ............................ 33
Broadband evolution ................................................................................................ 33
6
6.1
6.2
6.3
6.4
6.5
Long-term broadband traffic forecasts for fixed broadband ............................ 35
Drivers for the traffic evolution ............................................................................... 35
Busy hour traffic and dimensioning......................................................................... 37
Busy hour traffic evolution per subscriber .............................................................. 38
Busy hour traffic forecasts per subscriber ............................................................... 38
Total traffic forecasts in the busy hour .................................................................... 39
7
Fixed broadband tariffs......................................................................................... 42
Part B Mobile broadband ...................................................................................................... 44
8
8.1
8.2
8.3
8.4
Mobile broadband.................................................................................................. 45
The Mobile broadband market ................................................................................. 45
Mobile broadband technologies ............................................................................... 45
Mobile device and application evolution ................................................................. 47
Subscriptions and mobile broadband portfolio ........................................................ 48
3
MARCH D3.4
Market development up to 2015
9
9.1
9.2
Mobile broadband penetration ............................................................................. 50
Mobile broadband penetration in Western European countries............................... 50
Mobile broadband market potential ......................................................................... 51
10
10.1
10.2
10.3
10.4
Long-term mobile broadband penetration forecasts .......................................... 53
Forecast input ........................................................................................................... 53
Consumer market ..................................................................................................... 53
Business market ....................................................................................................... 54
Large and small screen subscriptions ...................................................................... 55
11
11.1
11.2
11.3
11.4
11.5
Mobile broadband traffic evolution ...................................................................... 58
Mobile data traffic divers ......................................................................................... 58
Mobile data traffic ................................................................................................... 59
Mobile broadband traffic volume forecasts ............................................................. 60
Mobile broadband traffic forecasts in the busy hour ............................................... 62
Uncertainties ............................................................................................................ 65
12
Mobile broadband tariffs ...................................................................................... 67
13
Conclusions ............................................................................................................. 68
References ............................................................................................................................... 70
4
MARCH D3.4
Market development up to 2015
1 Background
The deliverable aims to give a description of the evolution of the fixed broadband and mobile
market. There has been a continuous explosive traffic increase in both markets since the
introduction.
Fixed broadband was mainly introduced in 1999 – 2001 in most European countries, while
mobile broadband was mainly introduced 2006-2007.
Since 2004 the broadband traffic per subscription has increased exponentially. In addition
there has been significant subscription growth, but during the last years the fixed broadband
subscription growth is reduced and the penetration in some countries limits the saturation
level. Even though, the total traffic increases exponentially.
Mobile broadband traffic is increasing significantly. The main drivers for the mobile traffic
growth are: traffic evolution per subscription, number of subscriptions and the device- and
application evolution. Especially, the number of mobile broadband subscription is still in a
rapid growth phase which influences the traffic growth significantly.
The mobile and broadband networks are expanded to carry the increasing traffic. However,
the networks are very large and from time to time there will be situations with bottlenecks in
parts and specific places in the network. Network planners who make the plans for network
roll out and expansion have to take into account surplus capacity for a period and stretch the
period until additional capacity is needed. Optimization of the capacity utilization and the
traffic forecasts and the related uncertainties may induce traffic bottlenecks.
Also uncertain plans for introduction of new technology may create congestion in the network
- especially, when the technology or delivery of the components are delayed. The broadband
and mobile broadband networks are large and complex. New customer movements, new
habits, new applications may influence the traffic and create problems for the network
capacities in some parts of the network.
We see more dynamic changes and faster changes among the users now, compared with
earlier. Therefore, it is important to have possibilities to handle traffic congestion, bottlenecks
and the stochastic traffic variation in a more flexible and faster way.
Multilink equipments have these possibilities. An interesting aspect is to include multilink in
the planning procedures for optimizing the traffic flow and the investments.
This deliverable shows the traffic evolution in the international market, shows the traffic and
the subscriptions forecasts. The deliverable also gives an overview of the situation in different
countries. The enormous traffic growth in these areas clearly shows that there are a lot of
challenges for roll out the fixed and mobile broadband network.
5
MARCH D3.4
Market development up to 2015
2 Broadband overview
2.1
Broadband
In the 90-th the telecommunication market was classified in three speed parts:
-
Narrowband (Up to 128 kbps)
-
Wideband (128 kbps – 2 Mbps)
-
Broadband ( More than 2 Mbps)
The first class covered speed like ISDN including two B-channels. The classification was
relevant also some years in the beginning of the new decade. However, the concept wideband
has disappeared and broadband is now defined as capacities larger than ISDN.
The speed on the mobile systems has during the last years increased significantly and it has
been necessary to introduce the concept mobile broadband. To distinguish between broadband
in the fixed network and the mobile network the concept fixed broadband has been
introduced.
2.2 Fixed broadband
The main broadband technologies are:
-
DSL (Digital Subscriber Line)
-
HFC (Hybrid Fiber Coax)
-
Fiber (FTTH, FTTB/LAN)
-
Fixed Wireless Access (WLAN, WiFi, WiMAX)
For the business market also IPVPN should be included. Other broadband technologies with
only limited market share are satellite, local multipoint distribution system and power line
communication. Figure 2.1 gives a brief view of broadband network structures.
Figure 2.1 Physical network structures for some important broadband technologies
6
MARCH D3.4
Market development up to 2015
The figure also shows TV distribution from satellite to the DTT (Digital Terrestrial Television
network, to DTH (Direct To the Home) and to the head end (Broadband access point) for
Cable Television and HFC. In addition the network structure for ADSL and VDSL on copper
lines is illustrated and wireless and also FTTB (Fiber To The Building).
Digital Subscriber Line (DSL) is the most important broadband technology. The technology is
utilising the traditional twisted copper pair network structure. DSL is the broadband
technology with the highest market share on the worldwide basis. There is a set of different
DSL technologies which continuously are evolving. The broadband access could be
symmetric (SHDSL) or asymmetric (ADSL). The access capacity decreases with the length of
the copper line and the diameter on the copper line. Rate adaptive DSL (RADSL) extends the
reach of the copper lines. VDSL2 extends the speed. For short copper lines 0.5 km or less, the
possible speed offered may be more than 50 Mbps. There are methodologies under way to
increase the DSL speed by utilising additional copper pairs and reduce the signal to noise
proportion.
The Hybrid Fibre Coax (HFC) system also denoted as Cable modem was the first real
broadband technology to be established in the residential market. The technology is still the
dominating one in North America, but has lost part of its position to DSL technology in
Europe. The system is based on fiber connections on the top of the network from head end
and uses parts of the traditional cable television network by splitting the network into separate
small islands with cable structure, which are connected with a fiber droop from the head end
to the separate coax islands. In addition a return channel is established. The DOCSIS is
implemented for giving possibilities for high speed. 4 Channels DOCSIS 3.0 makes it
possible to offer more than 100 Mbps.
Fiber is deployed directly to the home (FTTH) or as fiber to the building with a local area
network inside the building (FTTB/LAN). The technology is expensive to deploy. The
technology is based on a fiber structure down to a fiber node and then dedicated fiber to the
customers or shared fibers by using passive optical PON/GPON solutions. Europe is
staggering behind in the fiber deployment especially compared with countries like Japan,
South Korea and US. The main problem for the European market is uncertainties regarding
possible future regulations of the fiber market. Few operators are willing to do major fiber
investments if competitors get rights to rent fiber capacity. Hence, relevant fiber regulations
must be in place, before significant rollouts in the European market take place.
Wireless broadband technologies are important in rural areas and also in hotspots. The
technologies are used in DSL-residual market for covering the additional part of the
broadband market (WiMAX). The wireless LAN like WiFi is used in hotpots like fuel
stations, hotels, airports, railway stations etc for nomadic communication. Now, WiFi is also
used for mobile broadband as a backbone network for users to handle heavy traffic volumes.
Also larger areas could be covered by WiFi hotspots in competition with other broadband
technologies.
Since fixed broadband was introduced, different type of statistics has been developed. In this
deliverable available statistics from EU, Eurodata, OECD, Strategy Analytics, Ovum,
Forrester and Point topics have been used. In case there are minor differences between the
broadband observations, then mean values have been used. In some cases only one source has
been used, and then the name of the source is stated.
7
MARCH D3.4
2.3
Market development up to 2015
Mobile broadband
The NMT (Nordic Mobile Telephony) was the first fully automatic cellular phone system. It
was specified by Nordic telecommunications administrations, and was opened for service in
1981. 10 years later in 1991 the first GSM network was opened. NMT and GSM could only
transfer voice.
GPRS was the first mobile technology developed for data communication. GPRS is based on
circuit-switched
(CSD) and high speed circuit-switched technologies (HSCDS), starting with offering of 9.6
and then 28.8 kbps which were fairly small capacities. GPRS was extended to offer 60/20 and
80/40 kbps download/upload speed.
EDGE is an enhanced technology carrying a bandwidth up to 237 kbps (4 timeslots in a
packet mode) and 474kbps (8 timeslots). EDGE handles four times as much traffic as the
standard GPRS.
UMTS (Universal Mobile Telecommunications System) is one of the third-generation (3G)
mobile telecommunications technologies, which is also being developed into a 4G
technology. The first deployment of the UMTS is the release99 (R99) architecture. It is
specified by 3GPP and is part of the global ITU IMT-2000 standard. The most common form
of UMTS uses W-CDMA technology.
Unlike EDGE and CDMA2000 (IMT Multi-Carrier), UMTS requires new base stations and
new frequency allocations. However, it is closely related to GSM/EDGE as it borrows and
builds upon concepts from GSM. Further, most UMTS handsets also support GSM, allowing
seamless dual-mode operation. Therefore, UMTS is sometimes marketed as 3GS,
emphasizing the close relationship with GSM and differentiating it from competing
technologies.
HSPA Basic generations offer the speed alternatives: 1.8, 3.6, 7.2, 10.8 and14.4 Mbps. HSPA
will probably be upgraded to 20, 28, 42, 80 and 160 Mbps by the technologies 64-QAM,
MIMO and channel combining.
LTE – Long Term Evolution will offer speeds at the upper level of HSPA and higher.
The available speed is a crucial factor for the mobile traffic forecasts.
The Norwegian regulator has decided that the UMTS and LTE technologies are defined as
mobile broadband. For instance, EDGE is not a mobile broadband technology. The speed
offered is too low.
The regulator has also decided that only subscriptions for large screen devices are broadband,
while subscriptions for small screen – handsets etc is not mobile broadband.
However in this deliverable the data traffic from small and large screens will be analysed.
8
MARCH D3.4
Market development up to 2015
There are limited international traffic data available and the mobile subscription broadband
statistics are also difficult to interpretate. This creates uncertainties in the analysis and in the
forecasts.
9
MARCH D3.4
Market development up to 2015
3 Long-term forecasting models
3.1 Forecasting methodologies for fixed and mobile broadband subscriptions
Diffusion models are often developed for long-term subscription forecasts. The models
describe S–shaped evolution with a long-term saturation level. Some relevant models are
Logistic models, Bass models, Gompertz model, Fisher-Pry and Richard’s model. The two
last one are special variants of the Logistic models. In [1] the Fisher-Pry models are applied
for long-term forecasting of many different telecommunication services. In [2] and [3] the
Logistic models are applied for long-term broadband forecasts.
The models are suitable for long-term forecasting and forecasting when the number of
observations is limited. Mobile broadband was introduced in the European market 2006/2007.
Hence, number of yearly observations is limited. A natural choice has been to apply diffusion
models. A suggestion here will be a four parameter Logistic model. One of the parameters in
the model is the market potential.
The models are also well suited when the accumulated demand starts to approach the
saturation level. This is the situation for fixed broadband. Hence, the models are used for
long-term penetration forecasts in this deliverable. References to similar work on this field are
given in [4-7].
3.2 Logistic four parameter model
A relevant long-term forecasting model is the extension of the Logistic model with three
parameters model to four parameters also by including the  parameter for modelling the
asymmetry in the demand. The model is defined by the following expression:
Yt = M / ( 1 + exp (t))
where the variables are defined as follows:
Yt :
M:
t :
:
Accumulated demand at time t
Saturation level
Time
Parameters
are growth parameters. The variable t is time in years. t = 0 is the introduction year.
Without the parameter  (, the demand curve will be symmetric.
3.3 Parameter estimation
The parameters cannot be estimated simultaneously by ordinary least squares
regression since the model is nonlinear in the parameters. The main objective in the fitting is
to get the best overall fit. Therefore, the parameters in the model are estimated by ordinary
least squares regression (OLS) for different values of The OLS estimation is based on the
following transformation:
ln ((M/Yt)1/ – 1) = t
The saturation level M and are fixed values in the estimation process. M is usually found by
market research, while is estimated by systematic calculations of RMSE (root mean square
error) for a set of different values or by using a nonlinear estimation procedure.
In [8] it is shown that transformation of the above mentioned diffusion models give a linear
expression in two of the parameters in the model. Then nonlinear estimation procedures can
easily be used to estimate the optimal parameters in the model.
10
MARCH D3.4
Market development up to 2015
Part A Fixed broadband
11
MARCH D3.4
Market development up to 2015
4 Fixed broadband access penetration
4.1 DSL – Digital Subscriber Line technology
The DSL dominating position in Western Europe is explained by cost effective rollout of
DSLAM/eDSLAM together with expansion of capacity in parts of the access network. Many
European countries opened their access networks early by introducing Local Loop
Unbundling (LLU), which in turn increased the demand for DSL.
ADSL has been dominating so far. The offered speed has gradually increased. Now, ADSL2+
is offered and in some countries also VDSL2. The coverage of ADSL for most counties in
Europe is about 94 % to nearly 100 % of the population because of the copper loop length.
The area which cannot be covered by DSL is called the DSL residual broadband market. Long
range ADSL can improve the coverage to a certain extent. VDSL2 997/998 and ADSL2+
have much less coverage, 1 – 1.5 km, with an acceptable speed. To increase the coverage the
access network has to be expanded with fibre and fibre nodes (FTTN).
IPTV is a driver for ADSL2+ and VDSL2. The incumbent operators and LLU operators in
some countries also have to meet competition from fiber operators and cable operators with
upgraded HFC networks. Hence, the incumbents and the LLU operators have to expand their
access capacities by ADSL2+ and VDSL2. OECD reported in October 2009, [9], the
following figures for maximum DSL speeds offered:
Figure 4.1 Maximum speed offered using DSL in OECD countries October 2009, [9]
Japan and Korea have since 2006/2007 offered extremely high DSL speed (102 Mbps). Now,
Portugal has started to offer the same speed same places. The high speed is caused by fiber
deployment to buildings and utilization of short copper lines inside the buildings. In Hungary,
Sweden and Denmark also very high speed is offered on copper. The figure shows that VDSL
is offered in Turkey, Netherlands, Austria, France, Germany, New Zealand, Greece, Finland
and Australia – among the OECD countries. However, the volume of high speed accesses in
Western Europe is so far very limited compared with the ADSL access volumes.
The following figure gives a picture of the distribution of speed classes offered in different
EU countries. The majority of broadband accesses are DSL. Hence, the figure gives an
indirect picture of the DSL also.
12
MARCH D3.4
Market development up to 2015
Figure 4.2 Distribution of fixed broadband accesses by speed, EU27, January 2010, [10]
Almost two thirds of the fixed broadband accesses in the EU offered speeds between 2 and 10
Mbps. Now, only 16% of the download rates are low speed accesses.
The figure shows that the countries with lowest broadband penetration rate have highest
volume of high speed accesses. Especially Portugal, Romania and Bulgaria have high
percentage of broadband accesses with speed of 10 Mbps or more. Also Belgium, Denmark
and Sweden have a high proportion of high speed accesses.
Austria, Cypros, Polen and Slovakia have a large part of low speed accesses left. Obviously,
this situation explains why the mobile broadband penetration in Austria is one of the highest
in the world.
The signal on the copper lines is affected of noise and cross talk on the lines in the bundle.
Hence, available speed is high for short copper lines, but the speed is reduced nearly
exponentially as a function of loop length. Figure 4.3 shows the speed for ADSL2+ and
VDSL2 as a function of the loop length. The ADSL speed for long loop length is comparable
with the one for ADSL2+, but the reach is longer especially when RADSL - rate adaptive
DSL is used.
13
MARCH D3.4
Market development up to 2015
Figure 4.3 ADSL2+ and VDSL2 speed as a function of loop length, [11]
An important long term target for EU and specific countries has been “Broadband to all. The
key technology for the broadband roll out has been DSL. The technology is rather cheap
because the copper lines are already in place. However, when more sparsely populated areas
are left, the economic value of the DSL roll out is more questionable. In [12] it is shown that
the operator who first enters these small areas will, in practice, get a broadband monopoly,
which improves the business case. The first mover advantage is also demonstrated in [13].
The part of the area which will not be covered by DSL is called the DSL residual area. Figure
4.4 shows the DSL national and rural area coverage among EU-nation members.
Figure 4.4 DSL National coverage and rural coverage in Europe, [14]
The DSL mean national coverage increased from 92.7 % to 94 % from December 2008 to
December 2009, while the rural coverage reached 80 % at the end of 2009.
14
MARCH D3.4
Market development up to 2015
The possibility for the DSL technology to maintain its very high broadband market shares is
continuous introduction of higher up and down speed and further rollout in rural areas. The
economic value (Net present value) for DSL rollout starts to decrease because the rollout
areas are smaller [13].
The last years have shown that the DSL meets competition from Fibre and HFC operators in
high capacity markets/areas and fixed wireless access (FWA) like WiMAX and also mobile
broadband in the low capacity markets/areas.
Based on statistics from EU, Eurodata, OECD, Strategy Analytics, Ovum, Forrester and Point
topics the demand data for broadband has been structured from 1999 to 2009 for the Western
European market. The countries are EU15 and Iceland, Norway and Switzerland. During
these 10 years the population has increased from 390 million to 405 million, while the number
of households has grown from 157 million to 175 million.
DSL has been a very attractive broadband technology during these years. Figure 4.5 shows
the DSL penetration in the residential market.
Figure 4.5 DSL penetration, Western Europe in the residential market, 1999 - 2009
The figure shows a clear S-curve form which is very good fitted by diffusion models.
The forecasting models also utilize the market share evolution. Figure 4.6 shows the evolution
of DSL market share. The total market consists of DSL, HFC, Fiber and FWA. In the fixed
wireless access (FWA) category there are broadband wireless accesses like WiMAX and a
small rest group of other technologies.
15
MARCH D3.4
Market development up to 2015
Figure 4.6 DSL market share, Western Europe 1999 – 2009
DSL has continuously captured market share until 2008, but it is interesting to see that in
2009 the technology is losing market share.
4.2 HFC/Cable modem – Hybrid Fiber Coax
CATV networks started to be deployed for more than 40 years ago. The first upgraded CATV
networks with two ways broadband connection appeared for nearly 13 years ago. Figure 4.7
shows that the market share for cable modem accesses in some countries continuously are
losing market share, while Figure 4.8 shows how the technology in other European countries
are catching market share. However, the situation is different in various Western European
countries.
Market share evolution, Cable, selected countries,
2004 - 2007
60 %
50 %
Austria
Iceland
40 %
Luxembourg
Spain
Sweden
Switzerland
30 %
20 %
UK
10 %
20
07
Q
4
20
07
Q
2
20
06
Q
4
20
06
Q
2
20
05
Q
4
20
05
Q
2
20
04
Q
4
20
04
Q
2
0%
Figure 4.7 Countries in Western Europe with continuous decreasing cable modem market
share (Point topic 2008)
16
MARCH D3.4
Market development up to 2015
Market share evolution, Cable, selected countries,
2004 - 2007
60 %
50 %
Belgium
Denmark
Finland
France
40 %
30 %
20 %
Germany
Netherlands
10 %
Norway
Portugal
20
07
Q
4
20
07
Q
2
20
06
Q
4
20
06
Q
2
20
05
Q
4
20
05
Q
2
20
04
Q
4
20
04
Q
2
0%
Figure 4.8 Countries in Western Europe where the cable modem market share has started to
increase slowly (Point topic 2008)
All countries except for Italy and Greece have reported cable modem/HFC networks. In
Iceland there is a modest market share: 0.2%. The broadband market share of the cable
modem technology is 38.8% in Netherlands, 38.7% in Portugal and 36.1% in Belgium
followed by 31.1% in Switzerland 31.0% in Austria and 28.5% in Denmark. UK has a market
share of 21.5%, Sweden 19.8%, Spain 19.4% and Norway 16.6%. In the large countries
Germany and France the cable modem market share is 5.7% and 5.1%. The largest increase in
number of accesses, from 2006 to 2007, has been in Germany, UK, Netherlands and Spain
with 590’, 360’, 250’ and 200’ accesses respectively. In spite of that, cable modem is still
losing market share in UK and Spain.
The figures show that cable modem market share in Austria, Luxembourg, Spain, Sweden,
Switzerland and UK is continuously reduced, while the cable modem market share in the
group of countries: Belgium, Denmark, Finland, France, Germany, Netherlands, Norway and
Portugal has started to increase slowly. Summing up the access volume in the two groups of
countries, the proportion of cable modem accesses is about 50/50.
In upgraded areas, the growth of new customers is limited. The cable operators’ possibilities
to consolidate the situation are to upgrade parts of CATV networks which so far have not
been upgraded. The main upgrade is installing of cable modem/HFC. The other upgrade is
expansion from DOCSIS 1.0/1.1 to 2.0 and 3.0 together with a possible restructuring of the
network to maintain increased traffic. DOCSIS 3.0 is now installed in the European market.
For 4 Channel DOCSIS 3.0 the maximum speed is 171/123 Mbps, while the speed for 8
channels is 343/123 Mbps. The EURO DOCSIS standard gives slightly higher capacity.
However, it is important to note that the capacity is shared among accesses in the last part of
the network. The following figure made by OECD shows the maximum cable modem
capacity offered in OECD countries:
17
MARCH D3.4
Market development up to 2015
Figure 4.9 Maximum capacity offered using cable modem in OECD countries (OECD
October 2009), [9]
The figure shows that in most European countries the maximum speed offered is comparable
with VDSL2 or higher.
Many European cable operators compete with the incumbent and offer broadband telephony
in addition to TV broadcast and broadband accesses (Triple play). These possibilities make
reasonable business cases for upgrading of the traditional cable TV network.
Figure 4.10 HFC/Cable modem penetration in Western Europe, residential market 1999 2009
It is interesting to see that the HFC penetration growth was reduced 2005 -2007. This is also
documented in figure 4.7. However, during the last years the cable operators have intensified
the CATV upgrading. This is also shown in figure 4.11 which shows that HFC for the first
18
MARCH D3.4
Market development up to 2015
time increases the market share. There is a modest market share increase from 15.3 % to
15.7%.
Figure 4.11 HFC/Cable modem market share evolution, Western Europe, residential market,
1999 -2009.
4.3 Fiber access
Fiber access is denoted as fiber to the home (FTTH) or fiber to the building (FTTB) or simply
FTTx. In broadband statistics fiber to the building with a local area network (FTTB/LAN) is
also included in this category. The FTTx technology is based on a fiber structure down to a
fiber node and then dedicated fibers to the customers or shared fibers by using passive optical
PON/GPON solutions.
Fiber access makes it possible to use very high capacities. In a way FTTx is the ultimate
telecommunication solution. The demand for FTTx will be driven by IPTV, interactive
broadband applications, HDTV and entertainment demand.
In Europe the investments for establishment of FTTx infrastructure in the access network
have been considered to be too high. The strategy so far has been to utilize other broadband
technologies, especially ADSL2+ and then VDSL2. HFC with DOCSIS 3.0 is a very relevant
competing alternative for cable operators.
One way to establish fiber accesses is to deploy fiber in new buildings – Greenfield. Another
way is to restructure and deploy fiber in areas with a lot of failures. A third possibility is to
pull fiber in ducts where there is enough space. Fiber deployment starts to be expensive when
parts of the deployment have to be based on digging and ducting. In Japan the poles are also
used to stretch the fibers. Figure 4.12 shows the limited deployment of fiber accesses in
Western Europe compared with Japan, South Korea and USA.
19
MARCH D3.4
Market development up to 2015
Figure 4.12 Evolution of number of fiber accesses in Japan, South Korea, USA and EU, [10]
The populations in Western Europe, USA, Japan and South Korea are 400 million, 291
million, 127 million and 47 million respectively. The figure shows that Japan is the leading
country having about 17 millions fiber accesses followed by South Korea with about 8 million
fiber accesses. The yearly increase is very significant for both countries. Since 2004 the DSL
market share has been reduced from 69 % to 32 % in Japan and from 57 % to 20 % in South
Korea. However, the market share of the cable modem in the two countries has remained
fairly constant during these years. Hence, the cable modem technology so far has been able to
compete with fiber accesses. Outside the OECD countries only Hong Kong has fiber market
share comparable with Japan and Korea.
The figure shows that USA has started and passed the fiber deployment in Western Europe.
Especially Verizon has aggressive rollout plans for the next five years. According the OECD
statistics the fiber market share in US was 4.9 % in Q4 2009. In Western Europe the fiber
market share was 2.0 % at the same time. Especially Sweden, Norway, Denmark and Italy
have the most significant fiber deployment in Western Europe regarding fiber access volume
and also penetration. New updated fiber statistics was released by the FTTH council in
September 2010. It is important to note that the FTTH council statistics showed in the
following figure is relative to number of households (not number of inhabitants).
20
MARCH D3.4
Market development up to 2015
Figure 4.13 FTTH and FTTB/LAN penetration relative to number of households. September
2010, [15]
Figure 4.13 shows that only few Western European counties are on the top of the ranking.
However, Eastern European countries as Lithuania, Slovenia, Bulgaria, Estonia, Slovakia,
Latvia, Russia and Romania are on the list. This indicates that the countries are building
network with new technology.
Figure 4.14 illustrate the difficult situation for Western European countries.
Figure 4.14 Fiber penetration, Western Europe, residential market 1999 – 2009.
Possible FTTx regulation in Europe is a huge topic. There are several unanswered questions
and uncertainties. Passive optical networks are relevant for fiber accesses. The technology
makes it difficult for wholesale. Will it be possible in practice to open the ducts for LLU
21
MARCH D3.4
Market development up to 2015
operators? Where will the natural connection points be? It is reasonable to believe that fiber
deployment will evolve faster when most of the regulations problems are solved?
Western Europe has been an advanced telecommunication area. However, the fiber
deployment is lagging behind compared with US, Japan and South Korea. So far EU is much
delayed in finalizing recommendation on fiber regulation. The operator are so far not willing
to do huge investments in fiber rollout and afterword get EU recommendations for sharing
deployed fibers with their competitors without controlled revenue.
4.4 Fixed Wireless Accesses
The fixed access wireless technologies are mainly WiMAX 802.16e, WiMAX 802.16 and
wireless LAN. These technologies can also be denoted as fixed wireless access and nomadic
access respectively.
Wireless LAN like WiFi covers the nomadic broadband market. So far international
broadband statistics give not a complete picture of the nomadic broadband market.
Fixed wireless access (FWA) statistics are available from most countries and are a natural part
of the fixed broadband statistics. WiMAX is the main technology. Even if the mobile
broadband technologies, CDMA, WCDMA, HSDPA, HSUPA are not fixed broadband
connections and not a part of the fixed broadband statistics, the evolution, deployment and
coverage will affect the fixed broadband demand. Hence, it is important to take the fast
increasing coverage of mobile broadband into account when fixed broadband forecasts are
developed.
The wireless technologies are shared resources with limited capacities. The available capacity
for the customer is dependent of the number of customers in the area, which are using the
service at the same time. For WiMAX the shared capacity is 14.4 Mbps which probably gives
a few Mbps per customer. New generations of WiMax may give higher capacity.
The demand for fixed access wireless technologies has been modest during the last 10 years.
The FWA penetration in Western Europe is increasing very slowly from 0.1% in 2002 till
0.6% in 2009.In Norway substantial means are transferred from the Government and the
municipalities for rolling out broadband – especially in DSL-residual market. Here FWA
solutions have been attractive solutions for covering the residual market. This roll out is not
finished and will still be a driver for FWA.
The fixed wireless access will meet competition in rural areas. HSPA and later LTE will be
rolled out in some part of these areas. The opening of the digital dividend, 800 MHz band,
will make possibilities for longer reach for the mobile technologies.
4.5
The fixed broadband market situation
It is difficult to concentrate analyses on specific countries. Western Europe is an advanced
telecommunication area. In this deliverable parts of the overviews and forecasts are
concentrated on Western Europe.
Figure 4.15 and 4.16 are summing up the residential broadband technology demand in
Western Europe from 1999 to 2009.
22
MARCH D3.4
Market development up to 2015
Figure 4.15 Penetration different technologies, Western Europe, residential market 1999 2009
Figure 4.16 Market share different technologies, Western Europe, residential market 1999 2009
The OECD statistics give a more detailed picture of the broadband situation among different
countries.
23
MARCH D3.4
Market development up to 2015
Table 4.1 OECD countries: Fixed broadband penetration. Q4 2009. Fixed broadband
accesses per 100 inhabitants by technology. Sum residential and business market, [9]
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Netherlands
Denmark
Sw itzerland
Norw ay
Korea
Iceland
Sw eden
Luxembourg
France
Germany
Canada
United Kingdom
Belgium
Finland
United States
Japan
Australia
New Zealand
Austria
Spain
Italy
Ireland
Portugal
Hungary
Greece
Czech Republic
Poland
Slovak Republic
Chile
Mexico
Turkey
OECD
DSL
22,1
22,4
25,1
20,9
6,6
30,7
17,8
26,5
28,7
27,4
13,2
23,3
16,5
22,2
10,7
7,9
19,0
21,7
15,1
17,1
19,9
16,0
10,4
8,2
17,0
7,4
8,1
6,8
9,6
6,8
8,6
13,9
Fibre/LAN
Cable
0,8
14,2
4,2
10,1
0,3
10,0
4,3
8,6
16,4
10,6
2,2
0,0
7,4
6,2
0,1
5,2
0,1
1,6
0,2
2,8
0,0
16,4
0,0
6,2
0,2
12,3
0,2
4,2
1,3
14,1
13,5
3,4
0,0
4,2
0,0
1,5
0,1
6,8
0,1
4,0
0,6
0,0
0,1
3,4
0,3
7,2
1,1
8,4
0,0
0,0
1,3
4,2
0,1
3,9
3,3
1,5
0,0
0,0
0,0
2,0
0,1
0,2
2,6
6,7
Other
0,0
0,3
0,3
0,1
0,0
0,0
1,0
0,1
0,0
0,0
0,0
0,0
0,1
0,1
0,3
0,0
0,0
0,0
0,1
0,0
0,0
0,0
0,0
0,1
0,0
0,0
0,0
0,0
0,0
0,5
0,1
0,1
Total
37,1
37,1
35,6
33,9
33,5
32,8
32,4
31,9
30,4
30,3
29,6
29,5
29,0
26,7
26,4
24,8
23,3
23,2
22,1
21,2
20,5
19,5
17,9
17,8
17,0
12,9
12,1
11,6
9,6
9,2
9,0
23,3
Source: OECD December 2009.
The table shows that 4 of the Nordic countries Denmark, Norway, Iceland and Sweden are
among the top 7 on the ranking. Finland had a higher ranking earlier, but has due to mobile
broadband substitution lost its position.
This OECD statistics [9] cover both the residential and the business market. Usually the
business market is 10 – 15 % of the total fixed broadband market.
Figure 4.17 shows the fixed broadband penetration in Western European countries. As
pointed out the fixed broadband evolution in Finland has decreased since the end of 2007.
24
MARCH D3.4
Market development up to 2015
Figure 4.17 Number of broadband accesses per 100 inhabitants (sum residential and business
market) by country, 1997 – 2009Q2
Figure 4.18 Number of broadband accesses per 100 inhabitants (sum residential and business
market) for OECD countries and Western European countries.
The OECD had a faster broadband penetration the first years after introduction, while during
the last years the penetration in Western Europe has grown faster and is significantly higher.
Looking on the Nordic countries, figure 4.19 shows fixed broadband penetration between 30
and 37 accesses per 100 inhabitants for the countries. The penetration is close to saturation
and as mentioned in the penetration in Finland is now decreasing. The penetrations in the
25
MARCH D3.4
Market development up to 2015
Nordic countries are much higher than the group means for OECD counties and the Western
European countries
Figure 4.19 Number of broadband accesses per 100 inhabitants (sum residential and business
market) for the Nordic countries
26
MARCH D3.4
Market development up to 2015
5 Long-term broadband access penetration forecasts
5.1
Drivers for additional broadband subscriptions
The fixed broadband evolution started mainly in 1999 – 2001. The subscription growth has
followed an S-shaped curve, with an initial exponential growth followed by an approximate
linear growth and with more limited growth the last years as the aggregated demand starts to
come closer to saturation.
Still there are areas which are not covered by fixed broadband networks. Hence, one
important driver is rollout in rural, sparely dens areas. Also the government and the
municipalities in various countries have declared support to broadband rollout in such areas.
The objective for the support is to have the same educational standard, to have possibility to
maintain working places and living standard also in these rural areas. A driving force is
“Broadband to all”.
Still there are specific user segments where the willingness to pay for broadband is low.
Especially older people, who are not familiar with PC usage, generate low demand. Also
young people for instance students and persons without own family generate low demand.
This user group prefer mainly mobile broadband subscriptions.
In addition the population is growing, the number of divorces is increasing, and the living age
is increasing and also number of immigrants. Hence the number of households is growing
which generate new potentials for broadband subscriptions.
Also new broadband applications and more extensive usage of existing applications,
broadband user groups, social networks, gaming and entertainment applications generates
new broadband subscriptions. At the same time there is a transition from the traditional fixed
broadband market to the high speed broadband market with access capacities from 20 Mbps
and upwards. The interest for the high speed, individual video and entertainment also create
demand for broadband accesses.
5.2
Long-term penetration forecasts for the residential market
Chapter 4 showed fixed broadband demand from 1999 to 2009. Here, these data is used as
input to the forecasting model described in chapter 3. The logistic four parameter model is
used to the long-term forecasting.
Usually long-term forecasts are made for a period of 5 years. Even for this period there will
be reasonably large uncertainties. Here, it is decided to make forecasts until 2018. There will
of course be very large forecast uncertainties in forecasts for such a long period. Especially in
an area which is significantly influenced of technology evolution and even technology shift.
The first step in the forecasting procedure is to make long-term forecast of the broadband
penetration and then split the forecast in separate forecasts for DSL, HFC, Fiber and FWA.
Here, we concentrate on the residential fixed broadband market, which is 8-9 times larger
than the business market.
27
MARCH D3.4
Market development up to 2015
The fixed broadband demand for the Western European market is collected carefully since
1999. From 1999 and until 2009 the following sources have been used: EU, Eurodata,
OECD, Strategy Analytics, Ovum, Forrester and Point topics. Figure 5.1 shows the fixed
broadband penetration.
Figure 5.1 Fixed broadband penetration, Western Europe, residential market, 1999 -2009
In 2009 the number of households in Western Europe was about 175 million. The penetration
was about 56.5% and the number of broadband household subscribers was nearly 99 million.
The figure shows that the yearly growth during the last years is reduced.
Figure 5.2 shows the yearly growth in the broadband penetration. In addition the figure shows
the fitting by using the Logistic four parameter model. The two curves are remarkable similar,
which indicates an extraordinary good fitting. Similar analysis and forecasting has been
carried out for sum of the residential and business market [16].
Figure 5.2 Yearly growth of fixed broadband penetration and the fitting by Logistic four
parameter model, Western European residential market, 2000 – 2009.
The accumulated penetration and the forecasts are shown in the next figure.
28
MARCH D3.4
Market development up to 2015
Figure 5.3 Broadband penetration forecasts, Western Europe, residential market 2010 – 2018
The saturation is estimated to 67.2 % and the forecasts are losing to the saturation in 2018. In
2009 the penetration is 56.4 %. So far the estimation has been very good.
However, there are significant uncertainties which may affect the long-term forecasts. In
Finland the penetration has started to decrease especially because of churn to mobile
broadband. In the coming period it must be expected that there could be substitution effects
which is not included in the development of the demand curve so far. Hence, it is important to
follow the demand and examine the evolution.
On long-term it is expected that entertainment applications and other very heavy traffic
consuming habits will attract the broadband customer and churn many mobile broadband
customers without fixed broadband subscription back again.
In the following, separate long-term forecasts for the market share between DSL, HFC, Fiber
and FWA are shown. The sum of the market shares of the four technologies are 100 %. The
market share forecasts of HFC, Fiber and FWA are made individually, while DSL is the
residual.
5.3
Long-term HFC market share forecasts
The HFC market share evolution is interesting. See chapter 4.2. As mentioned, HFC has
continuously lost market share until 2007. Then the market share has started to increase very
slowly. It is important to note that HFC is partly a fiber network with ability to offer very high
capacities when it is fully upgraded. Investments and upgrading to DOCSIS 3.0 is necessary.
But the first step of course is the two ways broadband upgrading itself. The demand is based
on selling HFC broadband subscriptions to homes passed and to sell to households in newly
broadband upgraded areas. This procedure has been carried out for a long time. Some
HFC/cable operators have nearly reached their upgrading possibilities. Some others have
larger potential network areas left for upgrading. The operators in main part of European
countries are in first mover advantage position because the offer very high broadband speed
before the fiber is introduced. VDSL2 has also been delayed in many countries.
29
MARCH D3.4
Market development up to 2015
It is interesting to see that HFC has retained its market share in Japan and South Korea in
spite of an extreme rollout of fiber accesses in the two countries.
Hence, it is assumed that the HFC/cable operators are able to increase their market share from
about 15 % to nearly 20% during the next years because of delay in fiber rollout in Western
Europe, because of to slow instruction of VDSL2 in the market and because the possibility to
utilize First Mover Advantage which means that many customer already have chosen HFC
before fiber and VDSL2 is offered in the area.
The long-term HFC penetration forecasts are shown in figure 5.4
Figure 5.4 HFC long-term market share forecast, Western Europe, 2010 – 2018
5.4
Long-term fiber market share forecasts
The fiber access forecasts are very difficult to develop. There are many significant
uncertainties. The largest uncertainties are related to EU’s possible regulation of the fiber
market. So far no recommendation has appeared. The recommendation has been delayed for
nearly two years. The consequences are that only few operators who take chance to roll out
fiber. The risk is too high to make major fiber roll out investments and get competitors who
could claim to use the same fibers without digging and ducting.
30
MARCH D3.4
Market development up to 2015
Figure 5.5 Fiber access market share forecasts, Western Europe, residential market, 2003 2018
Probably EU will recommend LLU and wholesale on fiber like on copper. Then the important
question is: what is the price for leasing fiber capacity going to be. These discussions have to
be settled before the operators will go for huge fiber investments in the access network. Only
in countries like Norway rich power utility companies are able to roll out fiber now. It is also
important to note that fiber because nearly unlimited capacity is the ultimate solution, but it
creates also huge investments for the fiber roll out.
5.5
Long-term FWA market share forecasts
FWA – fixed wireless access is a minor technology. Probably the market share will increase
to some degree because of coverage of the DSL-residual market. On the other hand FWA will
also be attacked by mobile broadband. Therefore, a modest market share forecast have been
developed which indicate a smoothly increase from about 1 % in 2009 to 1.5 % in 2018.
5.6
Long-term DSL market share forecasts
The resulting market share forecasts is DSL which is shown in figure 5.6
31
MARCH D3.4
Market development up to 2015
Figure 5.6 DSL market share forecasts, Western Europe, residential market: 2009 -2018
DSL started to lose market share from 2007. There are uncertainties in these forecasts. A key
point is how fast the various operators in the Western European countries are able to roll out
VDSL2 to compete with HFC and fiber. In Norway the market share evolution is more
dramatic than the forecasts shown in the figure. In 2005 DSL started to lose market share. In
Q3 2009 the DSL market share is about 53%. This is caused by extensive fiber and HFC
rollout, while VDSL2 so far is not introduced.
To be able to maintain significant DSL market share, it is important to roll out VDSL2.
Figure 5.7 gives an overview of the market share evolution of all four technologies. The
figure shows that fiber, HFC and FWA is increasing, while DSL is the big loser.
Figure 5.7 Market share forecasts for DSL, HFC, Fiber and FWA, Western European,
residential market, 2010 -2018
32
MARCH D3.4
5.7
Market development up to 2015
Long-term penetration forecasts for DSL, HFC, fiber and FWA
Then each of the market share forecast are multiplied with the overall penetration forecasts
which results in penetration forecasts for each technology. Figure 5.8 shows the forecasts.
Figure 5.8 Penetration forecasts DSL, HFC, Fiber and FWA, Western Europe, residential
market, 2010 -2018
The subscription forecasts are found by multiplying the penetration forecast by forecast for
number of households.
For network dimensioning purposes, the traffic generated from the different fixed network
technologies are important.
5.8
Broadband evolution
Figure 5.3 shows that the fixed broadband penetration is reaching a saturation level. During
the period 2010 – 2018 there will be churn effect caused by mobile broadband substitution.
The reasons are especially cheaper prices for mobile broadband compared with broadband,
but limitations of traffic volume usage.
It is expected that churn effect during some years in most countries will stabilise and there
will be some user segments which will have mobile broadband only. But most households
will have both fixed and a mobile broadband subscriptions.
Especially the entertainment applications will play an import role in the future. The traffic
generated by these applications demand will soon be much higher than peer to peer traffic.
And the traffic volume per month will be so high that mobile broadband will not be able to
33
MARCH D3.4
Market development up to 2015
transport the traffic. Hence, there will substitutions both ways between fixed and mobile
broadband.
There are significant uncertainties which may affect the long-term penetration forecasts. In
the coming period it must be expected that there could be substitution effects which is not
included in the development of the demand curve so far. But as mentioned there will also be
substitutions from mobile broadband to fixed broadband because need for downloading and
uploading of high traffic volumes. Hence, it is important to follow the demand and examine
the evolution.
In this deliverable the penetration forecasts for the fixed market have been developed for the
residential or the consumer market. The residential market is the dominating one and is 8-9
times larger than the business market regarding number of broadband subscriptions. The
evolution also shows that the residential market share continuously is increasing. Therefore,
the residential market gives a reasonable representative picture of the fixed broadband
evolution.
34
MARCH D3.4
Market development up to 2015
6 Long-term broadband traffic forecasts for fixed broadband
6.1
Drivers for the traffic evolution
The number of fixed broadband subscribers and the future evolution is an important driver. In
addition the population growth is a driver for the traffic growth.
The fixed broadband market changes from a low speed market to a high speed market. There
is a continuous upgrading of access speed. The operators offer higher speed. DSL is upgraded
to ADSL, ADSL2+ and VDSL2 and also internally in the different categories. HFC is
upgrading its protocol DOCSIS 1.0 to 1.1, 2.0 and 3.0. Higher speeds will also be offered by
fiber access. Fixed wireless access will do the same.
Another very crucial point is change of broadband technology. To get higher speed many
users leave DSL and order fiber or HFC technology.
The inheritance from internet has influenced the tariff structure for fixed broadband. The
traffic usage is supposed to be free – no charge for traffic volume usage. The principle is
called flat rate. The consequence is no higher price to pay for a high traffic user than for users
with low or extremely low usage. This is not a fair tariff system. The low traffic users, and
there are many of them, are financing the high traffic users.
The traffic per subscriber started very modest when fixed broadband was introduced. It is
interesting to note that the traffic has increased nearly exponentially for many years. The first
years after introduction, the growth was especially caused by the significant increase in
number of subscriptions, but growth was also caused by traffic growth per subscriber.
Surfing was the main driver the first years. Then file sharing and peer to peer (P2P) has been
very important. File sharing is distributing or providing access to digitally stored information,
such as computer programs, multi-media (audio, video), documents, or electronic books. It
may be implemented through a variety of storage, transmission, and distribution models.
Common methods of file sharing incorporate manual sharing using removable media,
centralized computer file server installations on computer networks, World Wide Web-based
hyperlinked documents. Peer-to-peer computing or networking is a distributed application
architecture that partitions the tasks or workloads between the users. The most important P2P
applications are BitTorrent, Limewar, Shareza, Kazaa, iMesh and eDonkey.
File sharing and peer to peer is widely used on film/movie file exchange. During the last years
it is identified that less than 10 % of the customers generate more than 90 % of the traffic.
Still peer to peer is the most dominant driver. However, the picture is changing. Cisco [17]
forecast that global video communication, especially internet video to PC and internet video
to TV, will dominate the market in the future and already at the end of 2010 generate more
traffic than P2P.
Important broadband applications are P2P, two ways video communication, video on demand,
Internet video to PC, Internet video to TV, advanced Internet video (3D and HD), video
conferencing, online gaming, Web and data transfer.
35
MARCH D3.4
Market development up to 2015
Multitasking and passive networking is characteristic for many broadband users. Their work
processes are continuously going on without an active participation all the time. Still much
traffic is generated.
Hyperconnectivity has emerged as an important dynamic concept. Important drivers for
hyperconnectivity are: Growing of high speed penetration, expansion of digital screen surface
area and resolution, evolution of network enabled devices, increase in power and speed of
computing devices.
Figure 6.1 Average global broadband speed forecasts 2009 – 2014 (Source: CISCO [17])
Figure 6.1 shows CISCO’s global forecasts of available access speed for the next years.
Obviously this development which also is indicated in figure 7.1 and 4.2 is a driver for
increased broadband traffic.
Figure 6.2 shows the CISCO’s screen surface predictions.
Figure 6.2 Installed screen surface areas inn billion squares feet. Forecast 2009 -2014.
(Source: CISCO [17])
The increased screen surface area is a factor which generates additional traffic. The installed
screen surface is dominated by TV and PC screen which have got a significant growth the last
years.
36
MARCH D3.4
Market development up to 2015
Morse law for increased computing power is a fundament for increased traffic. The PC’s
ability to produce more structured information cause additional traffic.
In general a digital video ecosystem compounding of consumer electronic, consumer
behaviour, the media and entertainment sector and the service providers and operators is
evolving.
Significant number of video devices like game consoles, dedicated DMA, Set Top box/DVR,
DVD/Blue ray and Flat panel TV are now connected to fixed broadband.
Because of analogue switch-off, global digital TV homes are growing strongly. DTT (Digital
Terrestrial Television network) and DTH (Direct to The Home satellite) will increasingly
deploy hybrid solutions using the fixed broadband access for video on demand. All platforms
will increasingly offer Web video options for access TV at any time and also accessing TV
anywhere.
IPTV is offered by several European DSL and fiber operators. The TV channels are
transported by the core network, and then by multicast in the access network. The multicast
technology saves significant traffic by not transferring TV channels which are not demanded
and not more than one dedicated TV channel to customers who demand the channel. Multicast
is based on broadcasting of the TV channels in real time.
However, the real explosion in the traffic will occur when the customers individually generate
demand for downloading TV films, TV news, TV programs, etc not in real time, but at time it
suites the customer either on PC or TV. Instead of broadcast by the multicast technology,
there will be individually TV demand transferred on demand basis by the broadband
customers. This will create an explosive evolution of the broadband traffic.
The next step is increasing of the traffic by going from TV to HDTV and 3DTV. The question
is the ability to transfer such information on individually demand basis. In parallel optimal
coding methodologies are developed to reduce the amount of data transfer, but it will only
reduce the data streams marginally.
There will be limits of how much traffic which can be transported on the fixed broadband
network without additional payments. Will the flat rate tariff principle survive the steadily
increasing broadband traffic or will it be necessary some time to implement other business
model than to day. Probably there will be limits, as we now see it in the mobile broadband
network, for the operators to extend the network capacity without getting paid for it.
6.2
Busy hour traffic and dimensioning
The multilink system is valuable equipment for enhancing traffic flow in networks. As
mentioned there will always be traffic bottleneck in the fixed broadband and mobile network.
If not the networks will be extended based on fixed overcapacity which gives not optimal
investments.
The applied network dimensioning is based on busy hour traffic data. It is interesting to know
that the busy hour traffic is much lower than the available capacity to the mean user. See
figure 6.1.
There are not much available official traffic statistics for the busy hour. One reason is that
busy hour traffic for the fixed network is not charged and is therefore not interesting for
revenue controlling purposes. However, the traffic measured in the busy hour is important for
37
MARCH D3.4
Market development up to 2015
evaluation of rollout strategies, upgrading and continuous expansion of the fixed broadband
network.
The busy hour traffic per customer is defined as the sum of all traffic generated by a given
number of customers in the area divided by the number of customers. It is important to note
that many customers are not using the broadband access in the busy hour. Hence, the
customers who are using the broadband at that time will have higher speed than the mean
busy hour speed.
6.3
Busy hour traffic evolution per subscriber
In this deliverable the evolution of DSL traffic per subscriber in the busy hour from 2004 until
2009 is stipulated for the Western European countries. The basis for the stipulation is
Telenor’s DSL busy hour and traffic statistics from CISCO ([17]). Telenor’s broadband traffic
is not published, but the stipulated broadband traffic for Western Europe will be a reasonable
good basis for the analyses and the forecasts. The traffic statistics is measured on routers at
the edge of the core network.
The traffic is based on number of subscriptions multiplied with the traffic per subscription for
the various technologies. The total traffic is found by summarising the traffic for each
technology.
Figure 6.3 shows the traffic per subscription for the DSL traffic.
Figure 6.3 Busy hour traffic in kbps per DSL subscription 2006 -2009
The traffic evolution per DSL subscription these four years show increasing growth year for
year. The yearly growth has been more than 30% and continues to increase with the same
rate. The busy hour traffic per subscriber is estimated to be 123 kbps in 2009.
6.4 Busy hour traffic forecasts per subscriber
Based on the traffic evolution 2006 -2009 forecasts have been developed for the period 2010
– 2018. The period is very long and there are definite uncertainties connected to the very long
forecasting period.
38
MARCH D3.4
Market development up to 2015
Figure 6.4 Busy hour traffic forecasts in kbps per DSL subscription
Observation of the busy hour traffic in Telenor’s network show that the HFC traffic per
customer is on the same level as the DSL traffic. However, the HFC broadband traffic will
due to capacity upgrading offer higher speed than DSL even when DSL subscriptions get
higher speed because of higher speed offerings and transitions from ADSL to ADSL2+ and
VDSL2. It is assumed that a mean HFC customer in the long run will generate 30% higher
traffic per access than a DSL customer.
Measurements show that fiber accesses have generated about doubles as much broadband
traffic as DSL customers. There are significant uncertainties about the future evolution.
However, the same proportion is used in the traffic forecasts per customer.
Fixed wireless access systems like WiMax have capacity on 14.4 Mbps. It is a shared medium
and many subscribers have to use the system at the same time in the busy hour. Such a system
may have 500 -750 subscribers per site. It depends of course on how the operators are rolling
out the sites for the system. Here, it is assumed that the busy hour traffic per subscriber is 30
kbps. As mentioned speed for a customer using WiMAX in the busy hour, will be much
higher than 30 kbps simply because not all customers are using the WiMAX at that time. The
busy hour traffic is a mean traffic for all subscribers connected to the specific WiMAX site.
It is difficult to forecast the busy hour traffic per subscriber for fixed wireless systems because
the future portfolio is not very well described. Therefore, it is assumed that the traffic increase
per subscriber follows the DSL increase proportionally based on the relative level between the
two access technologies. There may be some danger for overestimation of the FWA busy hour
traffic.
6.5
Total traffic forecasts in the busy hour
The forecasting assumptions can be summarised in the following figure:
39
MARCH D3.4
Market development up to 2015
Figure 6.5 Busy hour traffic forecast per subscriber in kbps for the technologies DSL, HFC,
Fiber and FWA 2010 – 2018.
Figure 6.6 Busy hour forecast per subscriber in kbps and the contribution from the other
broadband technologies DSL, HFC, Fiber and FWA 2010 - 2018
Then the busy hour traffic forecasts per subscriber are found by multiplication of the
penetration forecasts for each of the broadband technologies.
The busy hour traffic in the figure is the traffic bases on the real penetration in the given
years. The busy hour traffic for a mean subscriber is larger. The busy hour traffic in figure 6.6
is calculated based the proportion of the subscribers who have fixed broadband access.
The figure shows that DSL gives the largest contribution to the busy hour traffic. The reason
is much higher penetration for DSL compared with the penetration of the other technologies.
40
MARCH D3.4
Market development up to 2015
However, DSL will not be dominating closer to 2018 because higher fiber and HFC
penetration and the fact that the traffic per subscriber for these technologies is larger.
IPTV is not included in traffic forecasts. The IPTV traffic will be traffic generated by
transferring a set of TV channels down to the DSLAMs by multicasting. Some additional
traffic is handled by zapping between different TV channels. The IPTV traffic is minor
compared with the other part of the fixed broadband traffic.
Figure 6.6 shows an increase in the busy hour traffic from about 100 kbps to 600 kbps in the
period taking into account the penetration rate.
Now, the total busy hour traffic for Western Europe with 175 million households can be
calculated. The number of household is multiplied with the penetration forecasts and then the
busy hour traffic forecasts per subscriber.
The busy hour traffic for Western Europe is shown in figure 6.7.
Figure 6.7 Busy hour traffic forecasts in Terabps (Tbps) for Western Europe 2010 - 2018
The figure shows that the busy hour traffic for Western Europe have an exponential growth
until the last years in the period where the subscription penetration approaches the saturation.
Then the traffic per subscription is the dominating driver.
41
MARCH D3.4
Market development up to 2015
7 Fixed broadband tariffs
To get broadband access, an installation fee has to be paid and a monthly subscription fee.
There is no charge for traffic usage. The traffic costs for carrying the individual subscribers’
traffic is included in the fixed monthly fee. The tariff principle is called flat rate – an
inheritance from internet.
The flat rate principle is not a good system. It would be fair to pay for the usage. Now, low
traffic users are paying for high traffic users.
Also the network operators get problems. They have to expand the network without having
incitements for the expansion. The best solution also for the operator would have been to
charge for traffic usage, especially in the busy hour since they are dimensioning the network
based on busy hour traffic.
Fixed broadband is a relative mature area. Most of the technologies are mature and there is
significant or mass production of network components. The broadband prices have decreased
significantly during the last years. In addition the speed offered to the subscribers is
increasing.
Figure 7.1 shows how the fixed broadband prices for different speeds in EU have decreased
2007 - 2009.
Figure 7.1 Broadband subscribers download speed rates and monthly medium prices in the
EU, 2007 – 2009 [10]
The figure shows the download speed distribution 0.5 Mbps to 100 Mbps from 2007 – 2009.
It is seen that the speed is increasing significantly during the period. At the same time the
figure shows that the monthly prices are decreasing. The prices on low speed are not
relatively reduced as much as for the high speed rates.
42
MARCH D3.4
Market development up to 2015
EU defined the broadband performance index (BPI) in 2008. BPI is used to measure the
relative performance of countries in what they call wide economy. That’s to identify relative
weaknesses and strengths of individual countries for making strategies and fine-tune policies
to progress the broadband economy. The BPI index takes into account the following
dimensions: broadband rural coverage, degree of competition, broadband speeds broadband
prices, take up of advanced services and socio-economic context.
Figure 7.2 shows the BPI for the European countries. The broadband price varies. It is seen
that countries like Eastland, Slovakia, Lithuania, Norway, Finland and Iceland have relatively
high broadband prices.
Figure 7.2 Broadband performance index, European countries [14].
The figure also shows a general broadband ranking. Here, Sweden, Netherlands and Denmark
get the highest BPI. These countries also have the highest broadband penetration. The second
group consists of UK, France, Norway and Belgium with high score except for one or two
factors. For instance prices are relatively high in Norway and average speed are particularly
too low in UK and Norway.
43
MARCH D3.4
Market development up to 2015
Part B Mobile broadband
44
MARCH D3.4
Market development up to 2015
8 Mobile broadband
8.1 The Mobile broadband market
The broadband technology evolution and future coverage have significant impact on the
market potentials for MBB subscriptions and MBB traffic. Very well developed strategies,
planning and timing for rollout of new technologies is crucial important for cost optimised
mobile broadband network. The investments and running costs influence the prices for MBB
portfolio of products offered to the market.
Therefore, it is important to be able to set the right market prices, to have good quality and
performance to catch significant market share. Also competition presses the prices and creates
additional demand.
In countries with a poor twisted pair network and limited HFC and fiber networks, the
broadband communication infrastructure needs heavy investments which may open the
possibility for MBB to catch market share from the fixed broadband market – not only
subscriptions, but also traffic volume.
The broadband penetration is important and especially the demand distribution between
different broadband technologies and the related speed. The fixed broadband traffic depends
on the mixed coverage of broadband technologies: DSL, HFC, Fiber, FWA(fixed wireless
access) which are available in specific geographic segments. For example: in sparsely
populated areas DSL or FWA may be available or none broadband technology at all.
This deliverable concentrates on Western European countries with a very well communication
infrastructure. Hence, it is not assumed significant substitution effects because of poor
network quality. The long-term forecasting models for fixed broadband in Western Europe are
previously described. However there are areas for instance in immature market, where the
fixed and mobile broadband situation is different.
The demand for MBB is generated in a competitive arena. We see that there is a churn
process especially among young people from fixed broadband to mobile broadband. In areas
with no broadband rollout, part of the broadband penetration potential can be taken by MBB.
But it depends on how fast MBB is rolled out with reasonable speed in the sparsely populated
areas. There will be a competition between Fixed Wireless Access like WiMAX and MBB.
It is important to distinguish between traditional MBB subscriptions and M2M subscriptions
in the statistics. MBB is the important one because M2M generates low traffic volume per
subscription. In this analysis M2M subscriptions are excluded.
8.2
Mobile broadband technologies
To be able to make reasonable good forecasts it is important to understand the technological
evolution. The available capacity and reach of the mobile systems and the how the systems
are dimensioned are a crucial points.
The generations of wireless communication systems are [18]:
-
1G:
NMT
2G:
GSM, GPRS
2.75G: EDGE/EDGE/EDPRS, CDMA2000
45
MARCH D3.4
-
3G:
3.5:
3.75:
4G:
Market development up to 2015
UMTS (W-CDMA), CDMA2000, FOMA
UMTS (HSDPA, HSUPA), CDMA2000(EV-DO Rev A)
UMTS (HSPA+), CDMA2000(EV-DO revB/3xRTT)
3GPP LTE, WiMAX, Flash-OFDM)
The first system for data communication was GPRS with very limited capacity. EDGE is an
enhancement of GPRS with maximum speed 474 kbps.
Figure 8.1 Principal forecasts for number of subscriptions by GSM(GPRS)/EDGE,
UMTS/HSPA and LTE
The figure describes roughly evolution of main mobile broadband technologies over a 30
years period. For advanced telecom countries a new generation of mobile systems will be
probably be on the market long time before 2030.
The capacities to the different technologies are shown in figure 8.2.
Figure 8.2 Expected capacity upgrade in Mbps for HSPA and LTE systems
46
MARCH D3.4
Market development up to 2015
The denoted speed in the figure is per sector in the base station.
It is of course difficult to predict when the mobile operators in different countries upgrade
their mobile networks. So far the operators have mainly installed Basic HSPA. But there are
some operators who have installed LTE already. An important incitement for the upgrade is
the customers’ increasing demand for higher traffic volume during a month.
The operators utilize the following possibilities for offering capacity to their customers:
-
Apply one or more carrier (5 -10 MHz bandwidths)
-
Apply three sectors
-
Increase the system speed
-
Rollout of new base stations (condensing)
-
Increase capacity of the backhaul
The operators should utilize the possibilities in the best economic way for offering additional
speed to their customers. Also the competition situation in the various countries influences on
the decisions.
8.3 Mobile device and application evolution
There development in mobile data communication is very fast. More and more handsets and
different devices will have connecting possibilities to the mobile networks. One classification
of the devices could be: Devices for the pocket, for the home, for job and for the car. In
addition different devices will be used for surveillance, measurements and registrations. This
is denoted as M2M equipment. Examples of portable devices are:
-
Digital cameras
-
Portable media player
-
Ebook (Dedicated reader device)
-
Portable game consol
-
Smartbook (browsing and cloud based)
-
Tablets or iPAD (PCs with touch screen, browsing, email, video, imaging, E-book)
-
Netbook (PC based, browsing, email, music, video)
-
Notebook (PC based, thin and light)
-
Mobile Internet Device (MID – mobile multimedia adapted equipment)
In addition new generations of handset will be produced including enormous growing number
of applications. The replacement time of the devices have been 1.5 years creates an
innovation running race between the manufacturers. Strategy Analytics has started to group
the handsets in:
47
MARCH D3.4
-
Feature phones
-
Smartphone (iPhone type)
-
Superphone
Market development up to 2015
The last one is an advanced PC with 4 inches screen and 16Hz processor. It is expected that
iPhone with Apple operation system and Android will catch increasing market share.
8.4 Subscriptions and mobile broadband portfolio
There have been substantial problems with the mobile broadband statistics in some countries.
Precise definitions and classification is important. In addition the market consultant
companies have unfortunately divergent statistics. The mobile broadband statistics from
Strategy Analytics, Analysys and Informa have been used for the analysis.
In the deliverable the term subscription and not subscriber is used. Usually a subscriber has
only one subscription, but there are significant possibilities for a person to have both a
consumer and business mobile subscription. Different coverage, prices, speed offering etc
may also give reasons for more than one subscription.
Computing devices can be connected to the mobile networks. The computing devices are PCs,
notebooks, netbooks, tablets and external modems with express cards, PCMCIA card, and
USB modem dongle USB stick.
The computing devices represent Large screen which have much higher traffic potential than
other type of connecting devices.
Small screen represents handsets (feature phones), smartphones and other type of connecting
devices.
It is important to distinguish between the different ways to use mobile data communication.
The four main alternatives are:
-
Large screen (Computing device) contract
-
Small screen contract (Different type of surfing)
-
Smartphone contracts
-
Mobile broadband data communication without contract (Pay as you go)
This deliverable uses three alternatives since small screen and smartphone contracts are
merged to the category small screen contracts. The important for the classification is either to
have a contract or not. For instance a smartphone owner without contract is a Pay as you go
user.
The mobile world seems to use the following terminologies in the future:
-
Mobile broadband (Large screen/PC based)
-
Other mobile data communication (“Other contracts” and Pay as you go)
In this deliverable we also distinguish between “other contracts” and Pay as go. It is also
interesting to follow traffic evolutions generated by the smartphones.
48
MARCH D3.4
Market development up to 2015
In addition the market is segmented in business and consumer subscriptions.
Chapter 9 and the first part of chapter 10 examine the large screen/PC based market. Later in
the deliverable forecasts are also developed for small screen and for Pay as you go.
The mobile broadband forecasts are dependent on available broadband and mobile broadband
infrastructure and available terminals and handset and the future evolution of mobile
technologies CDMA, HSPA and LTE.
49
MARCH D3.4
Market development up to 2015
9 Mobile broadband penetration
9.1 Mobile broadband penetration in Western European countries
So far the realisation of the CDMA technology is limited in Western Europe. Introduction of
LTE is in an introduction phase. The GPRS and EDGE offer limited capacity and are not
classified as mobile broadband technologies. Hence statistics of the HSPA evolution so far
reflects reasonable well the MBB situation in Western Europe.
The analyses are based on collected mobile broadband (MBB) demand data from Western
Europe. The long-term forecasts reflect a situation where mobile operators face reasonable
well developed broadband networks and broadband infrastructures. In many countries outside
Western Europe the MBB traffic may have less competition because of a poor cooper and
broadband infrastructure. However, the MBB growth in Western Europe has been significant
compared with other markets, especially emerging markets.
The MBB subscription penetration is shown for Western European countries in figure 9.1.
Austria has been in front until the end of 2008. Now, the Nordic counties Finland and Sweden
are on the top followed by Portugal and Austria. It is also interesting to know that the
broadband penetration (fixed network) in Finland as the first OECD country Figure 4.17 is
decreasing.
25 %
Finland
Sweden
Portugal
20 %
Austria
Denmark
15 %
Switzerland
Spain
UK
10 %
Ireland
Norway
Italy
5%
Germany
France
0%
Netherlands
Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2006 2006 2007 2007 2007 2007 2008 2008 2008 2008 2009 2009 2009 2009
Belgium
Greece
Figure 9.1 Mobile broadband penetrations Q3 2006 – Q4 2009. Countries in Western Europe
(Copyright: Informa UK limited)
Figure 9.1 shows a significant variation in the MBB growth the first years. Especially Greece
and Belgium have until now moderate growth. But also largest countries Italy, Germany and
France together with Netherlands have modest mobile broadband penetration development.
As mentioned, the penetration in emerging markets will be delayed compared with the
penetration in Western European countries and also the growth intensity [19] gives a good
introduction to MBB rollout and the forecasting procedure.
50
MARCH D3.4
Market development up to 2015
9.2 Mobile broadband market potential
This MBB analysis covers all different Western European countries. A mean value for
Western European countries is established. Based on the mean values the forecasts for the
evolution in Western Europe will be outlined.
In the first part, only mobile broadband subscriptions will be analysed. During the first years
there are only statistics of large screen connections which also is the basis for mobile
broadband.
The number of inhabitants in Western Europe was about 406 million in 2009. The number of
households was about 175 million and the number of employees is estimated to about 187
million.
The market consists of the consumer market and the business market which is analysed
separately. The methodology for estimating the market potential for mobile broadband
subscriptions is performed in two steps. The first step is to estimate the market potential for
ordinary mobile subscriptions. The next is to estimate a long-term proportion of ordinary
mobile subscriptions who will order a MBB subscription.
The methodology for estimation of the market potential for this medium size country can be
illustrated by using some data from the Norwegian market. In Norway the penetration of the
ordinary mobile phone was about 80% in 2008 in the consumer market. For the business
market the penetration of employees who have an ordinary mobile business subscription was
about 54% in 2008.
The mobile penetration at the end of 2008 was 110% and in 2009 111.7% relative to the
Norwegian inhabitants. The penetration tells that many Norwegian inhabitants have more than
one ordinary mobile subscription. Some of them have both a consumer and a business
subscription. In many Western European countries the total mobile penetration (sum
consumer and business) is larger than 100%.
Now, suppose that the saturation levels for the ordinary mobile subscription for Western
European country are 85% and 60% respectively for the consumer market and business
market. Then the total mobile subscription saturation level will be 113% which means that the
long-term demand for mobile subscriptions is a little bit more than the penetration in Norway
reached in 2009.
Table 9.1 shows how the long-term saturation level for mobile broadband subscription is
estimated for the consumer and business market respectively.
It is assumed that 40% of persons in the consumer market who has an ordinary mobile
subscription will in the long run order a MBB subscription. In the business market, it is
assumed that 50% of the employees who already have a mobile subscription in the long run
will order a mobile broadband subscription.
Table 9.1 Saturation level for mean penetration for Western European countries
Population
Ordinary
mobile
Share
MBB
Potential
MBB
Potential
MBB
Inhabitants
406.000.000
85 %
40 %
34.0 %
138.040.000
Employees
187.000.000
60 %
50 %
30.0 %
56.100.000
Total/mean
593.000.000
77%
43.2 %
32.7 %
194.140.000
51
MARCH D3.4
Market development up to 2015
Hence the market potentials for MBB consumer subscriptions will be 34% of the population
and the market potential for MBB business subscriptions will be 30% of the employees.
These rough assumptions may change if the development of mobile broadband contracts
stimulates the users to have more than one contract in the different markets. This is a crucial
point since the future market potential depends on the policy for offering various subscription
contracts. The methodology and the related forecasts are described in [20].
52
MARCH D3.4
Market development up to 2015
10 Long-term mobile broadband penetration forecasts
10.1 Forecast input
The four parameter Logistic model is used for the forecasting .The estimated saturation level
for consumer and business market respectively are input to the forecasting models.
In addition the mobile broadband demand data for the first years 2007 - 2009 are used as input
to the forecasting model. The penetrations for the Western European countries are shown in
figure 8.1
Based on statistics from Strategy Analytics, Informa and Analysys, the mean mobile
broadband penetration for the years 2006 – 2009 is estimated to: 0.4 %, 1.8 %, 4.5 % and 7.1
%.
The penetration proportion between the consumer market and the business market for 2008
and 2009 is 2:3, which is close to what has been observed in the Norwegian market. In
addition it is observed that the penetration in the business market was much higher than the
penetration in the consumer market in 2006 and 2007 simply because the business market
adopted mobile broadband faster than the consumer market the first two years.
Based on these assumptions for the mean MBB penetration for Western European countries,
for the years 2006 – 2009 are given in table 10.1:
Table 10.1 MBB consumer and business penetration 2006 - 2009
Market
2006
2007
2008
2009
Consumer
0.1 %
0.7 %
2.4 %
4.1 %
Business
0.6 %
2.3 %
4.5 %
6.4 %
The MBB penetration in the business market is relative to number of employees in the
business market (187 million) and not to the population.
10.2 Consumer market
The market potential for mobile broadband subscription in the consumer market is estimated
to 34% of the population. Four yearly observations are used as input to the model for
estimating the parameters.
Figure 10.1 shows the long-term MMB penetration forecasts.
53
MARCH D3.4
Market development up to 2015
Figure 10.1 Consumer market: Mobile broadband penetration forecasts (Large screen),
Western Europe, 2010 – 2015
During the first years mainly large screen subscriptions were offered in the Western European
MBB market. This also reflects the penetration statistics in figure 10.5, since small screen
subscriptions are introduced in 2009.
10.3 Business market
The market potential for MBB subscription in the business market is estimated to 30% of the
employees. Four yearly observations are used as input to the model for estimating the
parameters.
The initial penetrations 2006 – 2009 indicate a more symmetric evolution of MBB in the
business market. Figure 10.2 shows the long-term forecasts.
Figure 10.2 Business market: Mobile broadband penetration forecasts (Large screen),
Western Europe, 2010 – 2015
Figure 10.3 shows that the evolution of the MBB consumer penetration is more symmetric
than for the business which are caused by lower penetration increase during the first years.
The figure also shows that the MBB in the business market will reach the saturation faster
than MBB in the consumer market.
54
MARCH D3.4
Market development up to 2015
Figure 10.3 Consumer and business market: Mobile broadband penetration forecasts (Large
screen) Western Europe, 2010 – 2015
The forecasts of number of subscriptions are shown in figure 10.4. The forecasts reflect the
different potentials. The Western European population is about 406 million inhabitants, while
number of employees is about 187 million.
Figure 10.4 Forecasts of number of consumer and business MBB subscriptions (Large
screen), Western Europe, 2010 – 2015
10.4 Large and small screen subscriptions
The main part of the mobile broadband traffic is generated by users who have subscription
contracts. The subscriptions are divided in two main segments: Large screen and small screen.
The portfolio of mobile broadband products changes continuously in Western Europe within
the two main segments. This is caused by hard competitions among the market players. Most
55
MARCH D3.4
Market development up to 2015
of the products have volume cap mainly to limit carried traffic volume for saving investments
and offer good performance also in peak hour.
It is possible to use mobile broadband without having a contract. The traffic generated by
these users is called Pay as you go traffic. Roughly about 5 - 10% of the total MBB traffic in
the Norwegian market comes from users without contracts. However, there is no guarantee
that the situation will be so one in the future. The market share between users with and
without contracts depends strongly on the price and the possibilities by different contracts
offered by the operators.
Operators in many Western European countries have recently introduced MBB contracts for
mobile handset, which are attractive for mailing, streaming and surfing. However, it is
difficult for the moment to get a representative overview of the market share between large
and small screen simply because we have limited historical data. Therefore the forecasts for
small screen are very uncertain.
Here, it is assumed that the long run proportion between large and small screen will be 2:1
and 3:1 both for the consumer and the business market respectively. In the analysis it is
assumed that MBB handset subscriptions are introduced in 2009 in both markets. Figure 10.5
shows the penetration forecasts of small screen and large screen subscription in the consumer
market.
Figure 10.5 Consumer market: Mobile broadband small screen and large screen penetration
forecasts, Western Europe, 2010 -2015
The figure shows a small discontinuity year 2009 which is explained by introduction of small
screen subscription which catches part of the subscription market.
Similar forecasts are developed for the business market based on the MBB business
penetration showed in figure 10.6.
56
MARCH D3.4
Market development up to 2015
Figure 10.5 Business market: Mobile broadband small screen and large screen penetration
forecasts, Western Europe, 2010 -2015
For a five years period there are significant uncertainties in the forecasts. The portfolio of
mobile broadband products is important, the quality and performance is very important, the
speed offered, the possible volume cap and of course the prices.
Volume cap is the maximum volume GigaByte per month allowed. Especially the volume cap
and the price for mobile broadband products have significant influence on the market share
between the large screen (MBB) and small screen subscriptions and the size of mobile data
traffic generated without contract (Pay as you go).
57
MARCH D3.4
Market development up to 2015
11 Mobile broadband traffic evolution
11.1 Mobile data traffic divers
The traffic evolution depends on many factors. Since the radio part of the MBB network is
very expensive and also the capacity of the backhaul, the operators have started to limit the
traffic through the subscriptions contracts. Especially low price subscription contracts limit
the traffic over day and night and have in addition a monthly volume cap.
At the same time mobile operators expand their MBB coverage and the speed. They are
introducing new technologies UMTS/HSPA and later LTE. They are using different antenna
sectors, they are increasing capacities utilizing different spectrums and dedicated carriers and
they are condensing their base stations.
As opposed to MBB subscription, very limited public information is available on MBB
traffic. This deliverable uses mainly traffic information which is available from Scandinavian
MBB market as a basis for the traffic forecasts. The Scandinavian market may be a little bit
ahead of several Western European countries, but it is better to use these traffic estimates than
some lower estimates for a Western European mean which cannot be documented.
Like for fixed broadband the MBB traffic is for the moment dominated by peer to peer users.
The file sharing is described in chapter 6 and is widely used for film/movie exchange. Cisco
[17] forecasts that global video communication will dominate the market in the future. Then,
there will not be limited number of subscribers who generate a very significant part of the
traffic, but there will be definite part of the market. However, the situation is not so on for the
moment.
The situation in Sweden in 2009 is described in the following figure.
Figure 11.1 Distribution of subscriptions and the use of packet data at the end of 2009, [21]
The volume cap for one month in Sweden is mainly 10GByte. The figure shows that 3 % of
the Swedish MBB users generate 68% of the MBB traffic.
In Norway the MBB volume cap is mainly 5-6 Gbyte. Traffic measurements on HSPA 7.2
Mbps systems show that the traffic decreases significantly the last part of the month because
parts of the subscribers reach the volume cap before the end of the month. Information from
the Norwegian regulator PT shows that 10 % of the traffic the first half year 2010 was Pay as
you go traffic. [22].
58
MARCH D3.4
Market development up to 2015
So far the main drivers for mobile data traffic have been a limited number of subscribers who
probably apply file exchange/peer to peer. However, new services and applications are
developed and of course enhancements of the established. Chapter 6 described the expected
explosive traffic evolution especially because of entertainment and news demand, to watch
the TV programs when you want. MBB will in addition make it possible to watch program
where you want to watch them. Two ways video communication, video on demand, mobile
internet to video, mobile internet to TV, video conferencing, on line gaming, Web and data
transfer are heavy traffic applications.
For MBB, like fixed broadband, it is reasons to believe that in the future not only a limited
part of the subscribers will generate a significant part of the traffic, but a much larger part of
the subscribers.
Large screen generates the substantial part of the traffic compared to small screen including
iPhone and traffic generated without contracts.
During the last years there have been developed a huge number of applications to iPhone. The
applications create traffic and also for uploading of new versions of the applications. So far
the traffic monitored on iPhone is still much, much lower than traffic generated for large
screens.
Increased capacity upgrade is described in figure 8.2. The figure shows that the mobile base
station sector speed will increase from 1.8 Mbps to 14.4 Mbps for Basic HSPA. For HSPA+
the sector speed is expected to increase from 20 Mbps to 42 Mbps, and then to 80 -160 Mbps
for advanced HSPA and LTE.
The increased capacity upgrade is a very important driver for the MBB traffic. However, the
capacity number may create confusion. It is very important to note that even if the capacity is
upgraded, there will not be the same grade of upgrading in a substantial part of the sector.
Hence it may be necessary to roll out additional base stations for having the upgraded
capacity in the whole sector.
Another very important point is the backhaul capacity which easily can be the main traffic
bottleneck. The backhaul capacity from the base station to the access network structure has to
be upgraded either with higher capacity wireless systems or simply by fiber which are heavy
investments.
11.2 Mobile data traffic
The mobile data traffic is segmented in three groups: MBB traffic generated by large screen
subscription contracts, traffic generated by small screen (handset) contracts and mobile data
traffic generated by users without subscription contracts (Pay as you go). Forecasts are
developed for proportion of users with different contracts, while separate traffic volume
forecasts are developed for Pay as you go traffic.
The traffic volume forecasts per subscription per month takes into account historical traffic
demand data from some Western European countries. The long-term forecasts also reflect the
operator’s technology and capacity upgrades.
The traffic forecasts per subscription in the busy hour is calculated directly from the traffic
volume forecast per subscription per month given assumption of the proportion of traffic in
the busy hour relative to the 24-hours period. Based on traffic volume forecasts per
subscription for large and small screen and subscription forecasts for the consumer market
and the business market, long-term mobile broadband traffic forecasts have been developed.
59
MARCH D3.4
Market development up to 2015
To understand the total traffic evolution, it is important to study the traffic evolution per
subscription. The total traffic is found by multiplying the traffic evolution per subscription by
the number of subscriptions.
The mobile traffic evolution is based on statistics from Norway, Sweden and Denmark – the
Scandinavian market ([21] [22].and [23]). Mean estimate for GB per month per subscription
in the Scandinavian countries for the consumer and the business market is shown in table
11.1.
Table 11.1 Scandinavian consumer and business market: GB per month per subscription for
large screen (MBB) 2007 - 2009
Market
2007
2008
2009
Consumer
1.04
2.08
2.29
Business
0.35
0.58
0.71
The table shows significant traffic increase in GB per month per subscription in both markets.
The traffic per subscription is 3 -3.5 times higher in the consumer market compared with the
business market. The traffic is reduced because volume cap of the total monthly GB usage in
the different contracts. The operator’s practice of truncating the access speed when the
monthly maximum traffic volume (volume cap) limit is reached for subscribers is also crucial
for the traffic volumes.
The traffic volume statistics are important for revenue calculations and the revenue forecasts.
In addition it is important to examine the busy hour traffic. The busy hour forecasts are crucial
for dimensioning base stations and the network.
11.3 Mobile broadband traffic volume forecasts
Logistic four parameter models are used also to make the traffic volume forecasts per
subscription. For the subscription forecasts the market potential was estimated based on
service considerations and demographics (population etc). For the traffic, it is not possible to
find natural saturation levels. The long term saturation level for the traffic volume in busy
hour can always be discussed.
However, the model adjusts the fitting based on different assumptions on the saturation level
through estimation of the other parameters. Thus, there is important to use earlier history to
estimate the parameters.
Figure 11.2 and 11.3 show the forecasts 2010 – 2015 of the traffic volume per subscription
per month for large and small screen respectively and for the consumer market and the
business market.
60
MARCH D3.4
Market development up to 2015
Figure 11.2 MBB consumer market: Traffic volume forecasts per subscription in GB per
month for large and small screen respectively 2010 -2015
Figure 11.3 MBB business market: Traffic volume forecasts per subscription in GB per month
for small and large screen respectively 2010 -2015
The figures show significant yearly increase of the traffic for all groups. There are of course
large differences between large screen and small screen traffic. It is important to note that in
many countries subscriptions covering both large and small screen usage were introduced the
first years. Now, specific subscriptions are offered for the handsets – small screens. The
intention was mainly to use the handset for mailing, surfing and streaming. However, the
handset technology is evolving through iPhone, iPad, Android and new Smartphones. The
traffic increases because of new applications and new applied programs which have to be
uploaded and updated. The size of the handset is increasing, which also generate demand for
more traffic.
On the other hand, substantial work is carried out to reduce the mobile broadband traffic
through more effective browsing, signal coding etc. In addition most of the subscriptions have
volume cap per month. The traffic generated by some customer segments will be limited both
for large and small screen traffic. In the future there will be more speed and capacity
available. However, there will during a very long period be limited speed for some user
segments.
61
MARCH D3.4
Market development up to 2015
The figure 11.2 and 11.3 show the forecasts for small screen traffic are growing relatively
faster than for large screen. But, still it is a significant growth also for large screen. The
forecasts show that traffic is growing relatively faster for business subscriptions (%) than for
consumer subscriptions. However, the proportion between business and consumer traffic per
subscription is assumed to be 1:3 in 2009 (See table 11.1), while the forecasts roughly show
the proportion 1:2 in 2015. Probably the business subscribers will utilize their MBB
subscriptions more heavily to private purposes.
The monthly traffic volume forecasts for the Western Europe consumer market are found
adding the monthly traffic volume forecasts for the consumer market, for the business market
and the Pay as you go traffic. Here the Pay as you go traffic forecasts are assumed to be 7 %
of the total traffic.
The monthly traffic volume forecasts for Western Europe are the sum of the large screen and
the small screen traffic forecasts both for the consumer market and the business market.
The monthly large screen traffic volume forecasts are given as the population
(inhabitants/employees) multiplied with the subscription penetration forecasts and the
monthly traffic volume per subscription forecasts – and analogue for the small screen
forecasts.
The total monthly traffic volume forecasts in Western European countries are given in
TeraByte (TB) and shown in figure 11.4
The figure shows a strong exponential growth. The very significant growth is caused by the
subscription growth and at the same time a significant traffic volume growth per subscription.
Figure 11.4 Monthly traffic volume forecasts in TeraByte (TB) for Western European
countries, 2010 - 2015
11.4 Mobile broadband traffic forecasts in the busy hour
In [24] JPMorgan assumed that 15% of the mobile broadband traffic during 24 hours is
transferred in the busy hour. Here the traffic profile is assumed to be more uniform and the
busy hour traffic proportion during one day and night is set to 10%.
The calculation of the busy hour traffic based on traffic volume during one month in GB is as
follows: The GB per month is transformed to Kbps in busy hour given an assumption of
62
MARCH D3.4
Market development up to 2015
percentage of traffic during the busy hour compared with the total traffic during one day and
night (24 hour). The number of days in a month, the number of hours in a day and night,
number of seconds in an hour and number of bits in a Byte are taken into account.
Calculations show that 1GB per month corresponds to 7.4 Kbps when 10% the day and night
traffic of the mobile broadband traffic is carried during the busy hour. Hence the traffic
forecasts in GB per month in figure 11.2 and 11.3 multiplied by 7.4 give the traffic in Kbps in
the busy hour, which is shown in table.
Table 11.2 Traffic forecasts 2010 – 2015 in Kbps per consumer and business subscriptions
respectively in the busy hour for small and large screen
Market
Year
2007
2008
2009
2010
2011
2012
2013
2014
2015
Consumer market
Business market
Small screen Large screen Small screen Large screen
0
8,5
0
2,6
0
12,6
0
4,1
1,1
16,3
0,5
5,9
2,0
20,2
1,0
8,2
3,1
24,9
1,7
10,8
4,4
29,9
2,6
13,8
6,1
35,1
3,8
16,8
8,0
40,5
5,2
20,0
10,1
45,8
6,7
23,2
The total mobile broadband traffic in the network is generated of two groups of users: The
users who have a mobile broadband contract and users without a contract. As mentioned the
traffic generated by users with no contract is called “Pay as you go traffic”.
Statistics identifies number of consumer and business users, but not number of Pay as you go
users. The Pay as you go traffic depends of the mobile broadband portfolio and the
specification of each mobile broadband product. If the broadband products are attractive
enough compared with the prices for using the broadband network without contract, there will
be limited usage of Pay as you go. The operators define the market place also because of
internal competition. In some countries the Pay as you go traffic tariffs have decreased
significantly during the last years. Therefore, it is difficult to forecast the Pay as you go
traffic.
The total busy hour forecasts are found by multiplying the respectively subscriptions forecasts
(type of screen and type of market) with the busy hour traffic forecasts. In addition traffic
forecasts are made for Pay as you go traffic is added. Here, it is assumed that 7% of the total
traffic is generated by users without a mobile broadband contract.
Figure 11.5 shows the busy hour traffic forecasts in Gbps 2010 - 2015. The figure describes
an exponential traffic growth in the period 2007 - 2015. The growth is based on two main
factors: The growth of number of subscriptions and the growth of traffic per subscription.
63
MARCH D3.4
Market development up to 2015
Figure 11.5 Total mobile data traffic in Gbps in busy hour for the consumer market and
business markets Western Europe 2010 – 2015
The next figure shows traffic forecasts in busy hour for large screen, small screen and Pay as
you go.
During the Mobile World Congress in Barcelona 15 – 18th February 2010, Nokia Siemens
Network underlined the exponential MBB traffic growth and told that calculations they have
made, estimated 10.000% traffic growth until 2015 in all the world’s mobile networks.
At the same congress AT&T explained that they had underestimated the MBB traffic growth
and needed to revise the dimensioning procedures. Later AT&T as the first mobile operator
removed the flat rate principle from MBB.
Figure 11.6 Total Mobile data traffic forecasts in Gbps in busy hour for the Large screen,
Small screen and Pay as you go for Western Europe, 2010 - 2015
The figure shows the very significant difference between large screen traffic the small screen
traffic and pay as you go traffic. As mentioned the new statistics differentiate between large
screen usage and other type of activating the mobile broadband network. Large screen is
denoted as mobile broadband, while the small screen usage including smartphones and Pay as
64
MARCH D3.4
Market development up to 2015
you go is not. Large screen (mobile broadband) traffic is dominating small screen traffic
especially because the traffic per subscription is forecasted to be much higher. There will be
more Pay as you go users, but they will probably not generate very much traffic.
Table 11.3 shows the traffic forecast increase for a medium size European country based on
assumptions and market input described.
The first part of the table shows the amount of traffic 2009 – 2015 compared with the traffic
at the end of year 2008. The traffic in 2015 is 36 times as high (3647%) as in 2008. Nokia
Siemens Network estimated the traffic from 2009 to 2015 to 100 times. However, the
consideration was based on the whole world and not Western Europe consisting of fairly
advanced communication countries.
Table 11.3 Total MBB traffic increase in the busy hour
Increase I: Increase relatively to the traffic end of year 2008 in %
Increase II: Yearly traffic increase in %
Year
2008
2009
2010
2011
2012
2013
2014
2015
Increase I
Increase II
100 %
979 %
223 %
123 %
472 %
112 %
855 %
81 %
1386 %
62 %
2056 %
48 %
2818 %
37 %
3647 %
29 %
The second part of the table shows the traditional yearly percentage increase defined as
percent of traffic increase relative to the traffic that year. Because the traffic is low the first
years the yearly percentage increases are very high. The same pattern is usually seen when
new services are introduced.
11.5 Uncertainties
Still mobile broadband is a new service. The data extracted are roughly for a four years
period. The forecasts will be more precise when the observed period is longer.
The penetration and traffic data would have been more precise for Norway because of more
internal knowledge than for Western Europe. Hence, the estimated data for Western Europe
are more uncertain, especially the traffic data.
There are significant uncertainties in the traffic evolution because of introduction of monthly
volume cap. The limits which is 5 -10GB per month is implemented because of traffic
bottlenecks at the base station and the backhaul. Control routines are implemented for
truncating the traffic for subscription who reaches the volume cap before the end of the
month. The volume cap creates uncertainties in the traffic evolution because the real traffic
demand is reduced.
There are significant evolutions of applications and programs used on the mobile broadband
terminals. The mobile broadband portfolio strategy and the prices, volume caps connected to
the product contracts are important.
65
MARCH D3.4
Market development up to 2015
In this deliverable no evaluation of uncertainties in the forecasts has been performed. This is
of course an important point. It is difficult to estimate uncertainties in the composite models
analytically. One way to illustrate the uncertainties is to perform a sensitivity analysis and
study the variations in the forecasts. Another possibility is to define suitable probability
distributions for each forecast component and use Monte Carlo simulations to estimate the
total forecast uncertainty.
The methodology is described in [25].
66
MARCH D3.4
Market development up to 2015
12 Mobile broadband tariffs
The mobile broadband tariffs have been under pressure especially in Western Europe. Hard
competition has continuously pressed the prices.
The investment costs for rolling out the mobile broadband networks have been huge, but the
ARPU is at the same time low. Economical analyses have underlined the necessity of
additional business models to increase the revenue.
The consumers in Western Europe often use mobile broadband services as a complementary
second connection rather than as a primary connection. Therefore, it is important for the
providers to have lower mobile broadband prices than fixed broadband prices.
In Central and Eastern Europe, use of mobile broadband as primary access broadband access
is more common. It is reflected by higher mobile broadband prices than fixed broadband
prices. One reason can be bad quality of the twisted copper pair network in many areas in
Central and Eastern Europe.
In contrast to fixed broadband, the mobile broadband traffic is truncated through the
introduction of volume cap because of limited capacity in the network. In spite of capacity
upgrades of the base stations there will be bottlenecks in the mobile broadband network for a
long time. Then the question is how to handle the traffic in best possible way.
So far, most of the operators do not differentiate between their broadband users. When their
subscribers reach the volume cap, they get significant speed reductions for the rest of the
month. There are other types of solutions. AT&T has left the flat rate principle and charges
their mobile broadband users according to their traffic usage. Another possibility is to sell in
additional capacity in bulk when the volume cap is reached.
Utilisation of different principles will in the future generate additional revenue to the
operators.
67
MARCH D3.4
Market development up to 2015
13 Conclusions
Long-term penetration forecasts and busy hour traffic forecasts are crucial for dimensioning,
planning and roll out of fixed and mobile broadband network.
The fixed broadband will during some years (Western Europe) be approaching a saturation
level. Fixed wireless accesses will still be rolled out to cover areas without DSL coverage.
There are identified governmental initiatives to support broadband deployment in sparsely
populated areas with the “Broadband to All” intention.
Analysis of the fixed broadband market shows that the fight about broadband market is
turning from low speed market to a high speed market. The relevant high speed technologies
are VDSL2, Fiber and HFC. The losing technology is DSL which is losing market share in
urban areas to high speed technologies and to mobile broadband in rural areas.
Even when the long-term penetration is approaching saturation within some years, the fixed
broadband traffic will still evolve exponentially. The reason is the significant traffic increase
per subscription.
The main drivers for the traffic growth in the fixed and partly the mobile network are: flat
rate, higher speed, better computer and internet knowledge, multitasking, larger screens,
increased speed of the computers, increased storage capacity, increased number of devices
connected to internet, peer to peer and file sharing, two ways video communication, VoD,
Internet to PC, Internet directly to TV, advanced Internet video (HD and 3D), iTube, iTunes,
online gaming, Web and data transfer.
The short-term forecasts are important for identifying bottleneck in the fixed network and to
expand the network at the right time with an optimal capacity. There will be a trade off
between investments giving overcapacity for a period and too often usage of manpower to do
the capacity upgrading and vice versa.
The short term-forecasts are important for traffic carried of the radio part of the mobile
broadband systems. The long-term forecasts are also important for establishing bid strategy
for new mobile spectrum. Furthermore, the long-term subscription and traffic volume
forecasts per month are essential for revenue and business case analysis.
We see more dynamic changes and faster changes among the users now, compared with
earlier. Therefore, it is important to have possibilities to handle traffic congestion, bottlenecks
and the stochastic traffic variation in a more flexible and faster way.
Multilink equipments have these possibilities. An interesting aspect is to include multilink in
the planning procedures for optimizing the traffic flow and the investments.
The long-term forecasts are important for investment strategies, replacements and
introduction of new technologies and systems.
New methodology is described for forecasting mobile broadband subscription penetration by
using ordinary mobile potentials to estimate mobile broadband potentials based on inhabitants
and employees.
The traffic volume forecasts and the busy hour traffic forecasts are uncertain and depend on
the operators’ strategy for traffic reduction. The flat rate principle is stilled applied for fixed
broadband network. This is not possible for the MBB traffic especially because the radio part
of the network is very expensive. Therefore, the main part of the MBB products offered on
the market has a volume cap – usually a GB limit per month.
68
MARCH D3.4
Market development up to 2015
The main problem is the mobile broadband network capacity. Continuous upgrading of HSPA
(frequencies, antenna sectors, speed) and introduction of LTE will improve the capacity. In
addition the backhaul capacity is limited. Higher capacity wireless systems have to be
implemented. On the very long run probably fiber connections have to be rolled out to heavily
loaded base stations.
At the same time the mobile operators will restrict the traffic by using priorities, volume cap,
specific subscription contracts and probably by introduction of some traffic pricing.
The mobile broadband traffic in 2015 is forecasted to increase 36 times the 2008 traffic level
and 16 times the 2009 traffic level. There may be uncertainties in the forecasts, but in any
case there will be an explosive growth.
Figure 6.7 shows the busy hour forecasts in Terabps for fixed broadband traffic in Western
Europe, while figure 11.6 shows the busy hour forecasts in Gbps for the mobile broadband
traffic in Western Europe. The figures show that the fixed broadband traffic is estimated to be
35 times larger than the mobile broadband in 2009, while the proportion is reduced to 9 times
as large in 2015.
As pointed out in chapter 11 there will be uncertainties in the long-term forecasts, but the
direction of the forecasts is quite clear.
69
MARCH D3.4
Market development up to 2015
References
1. Vanston L. K, R. L. Hodges ”Forecasts for the US Telecommunication network”
ISSN 0085 – 7130 Telektronikk vol 104 3/4 – 2008
2. Stordahl K, “Long-Term Broadband Evolution – Forecasts and Impact of new
Technologies. ISSN 0085 – 7130 Telektronikk vol 104 3/4 – 2008
3. Stordahl K, “Long-term mobile broadband forecasts. Forecasts when number of
observations is limited”. In proceedings 29th International Symposium on Forecasting.
Hong Kong, China, June 22-25th, 2009
4. Stordahl, K, “Mobile broadband traffic and subscription forecasting for optimalisation
of network investments” International Institute of Research, Mobile Network
Optimalisation conference, Brussels November 9 -12th, 2009
5. Stordahl K, “Forecasting the long-term evolution of broadband accesses and the role
of new technologies”. International Institute of Research, Telecom Market Forecasting
conference, London, September 28 – October 1th, 2009
6. Stordahl K, “Broadband demand and the role of new technologies”. In proc Networks
2008, the 13th International Telecommunications Strategy and Planning Symposium,
Budapest Hungary, September 28 – October 2th, 2008
7. Stordahl K, “Long-term access broadband forecasting and the influence of new
technology”. 28th International Symposium on Forecasting. Nice, France,
June 22-25 th, 2008.
8. Riihimäki V, Managing Uncertainties in Broadband Investments – Case Studies of
Real Options for Rural Access Networks. Dr Thesis, Helsinki University of Science
and Technology, 2010
9. OECD Broadband Portal
http://www.oecd.org/document/54/0,3343,en_2649_34225_38690102_1_1_1_1,00.ht
ml 10. June 2010
10. Europe’s Digital Competitiveness report 2010.
http://ec.europa.eu/information_society/digital-agenda/index_en.htmI SBN 978-92-7915829-2
11. ASOTEL, 2010, http://www.asotel.net/news.php?news_id=140
12. Stordahl K, K. Elnegaard ”Broadband in the residual market: First mover’s
advantage”. In proc World telecommunication Congress. Budapest, Hungary, April 30
– May 3th, 2006.
13. Stordahl K, K Elnegaard “Adoption rate forecasts and roll out strategies” In proc 5th
Conference on Telecommunications Techno-economics. Athens, Greece, June 8 -9th
2006.
14. Europe’s Digital Competitiveness report Volume 1; i2010 - Annual Information
Society Report 2009.
15. FTTH council press release, September 15, 2010
16. Stordahl K, “Long-term broadband forecasts and the saturation level”. The 30th
Annual International Symposium on Forecasting, San Diego, USA, June 20 -23th,
2010
17. CISCO White paper: Cisco Visual Networking index: Forecast and Methodology,
2009 -2014. 2010.
18. Wireless system generations
http://no.wikipedia.org/wiki/Universal_Mobile_Telecommunications_System
19. Elnegaard N K, Stordahl K, Lydersen J, Eskedal T. G. “Mobile Broadband evolution
and the possibilities”. Telektronikk (104) no 3/4, 2008
70
MARCH D3.4
Market development up to 2015
20. Stordahl K, “Mobile broadband subscription and traffic evolution. International
Telecommunication Society. 18th Biennial Conference Tokyo, Japan, June 27 -30th,
2010
21. The Swedish Post and Telecom Agency “The Swedish Telecommunication Market
2009, Market Statistics, June 3th 2009
22. The Norwegian Post and Telecom Authority. Half year report in Excel, 2010
23. National IT and Telecom Agency, Denmark “Telecom Statistics 2009, Version April
29th 2009
24. JPMorgan “The Power of Mobile Broadband” May 2th 2008
25. Elnegaard N. K and Stordahl K. “Modelling uncertainties and risk in
telecommunication investment projects”. Telektronikk (104) no 3/4, 2008
71