2009 03 09 Operator Strategies

Research report

Operator strategies for network evolution: the road to LTE

Helen Karapandžić and Terry Norman

March 2009

Published by Analysys Mason Limited • St Giles Court • 24 Castle Street • Cambridge • CB3 0AJ • UK Tel: +44 (0)1223 460600 • Fax: +44 (0)1223 452800 • Email: [email protected] •

www.analysysmason.com/research Registered in England No. 5177472

© Analysys Mason Limited 2009

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means – electronic, mechanical, photocopying, recording or otherwise – without the prior written permission of the publisher. ISBN 1 905495 86 2

© Analysys Mason Limited 2009 Contents

Contents

About Analysys Mason vi

0 Summary vii

1 Wireless data growth will drive network evolution 1 1.1 Network evolution will be largely driven by consumer demand for services

delivered over high-performance networks 1 1.2 Wireless network traffic from voice and data services will increase dramatically

by 2015 2 1.3 The volume of traffic will erode revenue 4

2 Operators need to understand, and make, a complex array of technology choices 5

2.1 LTE offers a compelling proposition 5 2.2 There are other technology choices, including CDMA, TD-SCDMA and

WiMAX, but most operators will migrate to LTE 8 2.3 Existing assets and legacy services will influence future deployment plans 11 2.4 LTE will not be deployed as a standalone layer, but as a complement to a

dedicated indoor coverage mechanism 14

3 Spectrum issues will affect MNOs’ LTE deployment strategies 18 3.1 Operators will deploy LTE in the IMT-2000 expansion band 18 3.2 The timing of GSM 900MHz (and GSM 1800MHz) spectrum refarming is

critical 19 3.3 Digital dividend spectrum may be used for mobile communications 20 3.4 The US regulator has made spectrum available for Advanced Wireless Services 21

4 Device availability will be key to realising data growth 22 4.1 LTE-compatible devices will be available from 2H 2009 22 4.2 The first LTE-compatible devices are likely to support four frequency bands 23 4.3 Vendors have agreed a standardised royalty framework for LTE handsets 24

5 There are a number of pathways to LTE 25 5.1 The combined GSM and UMTS operator 25 5.2 The UMTS-only operator 26

iv

Contents © Analysys Mason Limited 2009

5.3 GSM-only operators 26 5.4 CDMA operators 27 Key to acronyms 30 Research from Analysys Mason 31 Consulting from Analysys Mason 32

List of Figures and Tables

Figure 0.1: Cornerstones of an operator’s strategy for deploying a data-optimised network vii

Table 1.1: Factors that are driving the increase in wireless network data traffic 1 Table 1.2: Key forecasts for mobile traffic in developed and developing regions 3 Figure 1.1: Global wireless network traffic growth, 2008–2015 3 Figure 1.2: The cost of data traffic using legacy networks, and revenue 4 Figure 2.1: Typical mobile network costs per subscriber per cellular sector 5 Figure 2.2: LTE can deliver profit by significantly reducing opex and capex 7 Figure 2.3: All technology roadmaps can lead to LTE 9 Table 2.1: SWOT analysis of migration from HSPA directly to LTE 12 Table 2.2: SWOT analysis of migration from HSPA to HSPA+ (Release 7) 13 Figure 2.4: Network traffic generated by an example mobile service mix, split between

indoor and outdoor usage, 2008–2012 15 Figure 3.1: Timeline for 2500–2690MHz spectrum auctions 18 Figure 4.1: Expected LTE device availability 22 Figure 5.1: LTE by geography/operator 28 Figure 5.2: Deployment strategy guidelines 29

Authors

Helen Karapandžić (Analyst) joined Analysys Mason in April 2007. Her research interests include wireless access, mobile broadband and mobile payment. Her most recent work includes several reports on the subjects of mobile payment, near field communications, WiMAX access technology and networks, radio access network evolution, mobile broadband and service bundling. Prior to joining Analysys Mason, Helen was Business Development Manager for an international asset management company, where she was responsible for Europe, the Middle East and Africa (EMEA). She has also held positions with Cambridge Investment Research (an independent, Cambridge-based strategy consultancy) and eBay AG. Helen has an MA in Modern Languages from the University of Cambridge.

Terry Norman leads Analysys Mason’s Wireless Networks research and analysis programme, which focuses on the development of wireless access technologies and their commercial applications. Terry has more than 30 years’ experience in the radio communications industry in radio planning, access network design, operations and latterly ten years consulting. He has authored numerous reports on the

Operator strategies for network evolution: the road to LTE v

© Analysys Mason Limited 2009 Contents

subject of wireless access technologies. He has been quoted in the broadsheet press and regularly contributes to media analysis. Terry holds a doctorate in radio propagation, and is a Chartered Physicist and a member of the Institute of Physics and the Institute of Technology.

Acknowledgements

The authors would like to thank all those who assisted in the preparation of this report: Anthony Berkeley (Director, LTE Strategy, Alcatel-Lucent), Dan Warren (Director of Technology, GSMA), Mats Blumenberg (LTE Product Line Management, Huawei), Stephane Daeuble (Senior Manager, Global Marketing, Motorola), Manash Goswami (Director of Business Development, Platforms and Mobile Devices, Motorola), Sami Jockinen (Senior Manager, Wireless Access Technology Marketing, Nokia), Peter Carson (Senior Director, Product Management, Qualcomm), Hervé Dubreil (Strategic Advisor, CTO Office, Orange Group), Frédéric Gastaldo (Head of Development, IT Services, Swisscom), Rico Schwendener (Head, Network Innovation Centre, Swisscom) and Jaime Lluch (Radio Access Technology Manager, Telefónica S.A.). The authors would also like to thank Analysys Mason colleagues Matt Hatton and Andrew Parkin-White for their input, and Sarah Peake and Claire Varley for editorial assistance.

Disclaimer

Figures and projections contained in this report are based on publicly available information only and are produced by the Research Division of Analysys Mason Limited independently of any client-specific work within Analysys Mason Limited. The opinions expressed are those of the stated authors only.

Analysys Mason Limited recognises that many terms appearing in this report are proprietary; all such trademarks are acknowledged and every effort has been made to indicate them by the normal UK publishing practice of capitalisation. However, the presence of a term, in whatever form, does not affect its legal status as a trademark.

Analysys Mason Limited maintains that all reasonable care and skill have been used in the compilation of this publication. However, Analysys Mason Limited shall not be under any liability for loss or damage (including consequential loss) whatsoever or howsoever arising as a result of the use of this publication by the customer, his servants, agents or any third party.

About Analysys Mason © Analysys Mason Limited 2009

About Analysys Mason

Analysys Mason is the preferred partner to telecoms, IT and media organisations worldwide. Through our global presence, we deliver strategy advice, operations support and market intelligence to leading commercial and public sector organisations in more than 80 countries. Our clients rely on our consulting and research services to make better business decisions.

The intellectual rigour, operational experience and insight of our people have helped our clients to meet some of the toughest challenges they face within the industry. We have consistently delivered significant and sustainable business benefits to our clients during the past twenty years on issues ranging from advising on operator strategy and tactics, development of national sector regulation, through execution of major financial transactions, to the deployment of public and private network infrastructure.

We are respected worldwide for the exceptional quality of our work, our independence and the flexibility of our teams in responding to client needs. We are passionate about what we do and are committed to delivering excellence to our clients.

Analysys Mason offers two types of service:

Consulting

We work with our clients to provide customised advice and support throughout the business cycle, helping major players to set strategy, plan for change and implement that change.

We have worked on a wide range of projects that support our partners and help them to increase revenue, reduce costs, plan for the future and minimise risk.

Research

We offer research programmes and custom research to help our clients to identify key strategic issues, formulate strategies, identify trends and opportunities, and measure performance.

The research programmes are:

• Industry Strategy • Global Growth Markets • Fixed Broadband • Mobile Broadband • Voice • Consumer Content and Applications • Telecoms Software • Fixed Networks • Wireless Networks • Core Data and Forecasts

For more information, visit our website at www.analysysmason.com.

© Analysys Mason Limited 2009 0: Summary

0 Summary

LTE is the only technology that can meet the requirements for the anticipated dramatic growth in wireless data traffic

Wireless data traffic is forecast to increase tenfold in developed countries between 2008 and 2015, and sevenfold in emerging markets over the same period. It is clear that mobile networks will reach their limits in terms of capacity and throughput within a few years. At the same time, revenue per megabyte continues to fall, driven down by increasing competition and the introduction of flat-rate pricing. Therefore, the decision to deploy a data-optimised radio network appears inevitable for most operators. There is, however, a complex interplay of choices, which operators will need to understand and sequence if they are to achieve an optimal position to deal with data demand and pressures on their revenue.

In this report, we evaluate the four main factors (besides the individual operator’s business objectives) that influence mobile operators’ strategies, timing and costs of deploying a data-optimised RAN architecture, as shown in Figure 0.1.

Figure 0.1: Cornerstones of an operator’s strategy for deploying a data-optimised network [Source: Analysys Mason, 2009]

Operatorstrategies forRAN evolution

Device availability

Traffic growth Technology mix

Spectrum availability


Device availability




Device availability




Device availability



A summary of our key findings in these four areas is as follows.

Traffic growth. Total wireless network traffic generated from voice and data services will increase dramatically by 2015. Traffic per cellular user per month in developed regions will

viii Operator strategies for network evolution: the road to LTE

0: Summary © Analysys Mason Limited 2009

rise from an average of 56MB in 2008 to 455MB. Over the same period, the combination of a substantial rise in wireless network traffic per user and a relatively modest increase in ARPU will result in a significant decline in revenue per megabyte: this will fall from USD0.86 to USD0.12 in developed regions and USD0.57 to USD0.15 in emerging markets over the same period. The costs of the existing network will soon begin to erode profit margins and will eventually outstrip revenue. As a consequence, the operator must:

• deploy a data-optimised network architecture • reduce network carriage cost per megabyte • ensure a level of indoor service that is comparable to alternatives, like Wi-Fi.

Device availability. The introduction of new and improved wireless devices (most notably USB modems and datacards) enabling a greater range of services has been one of the major drivers of mobile data usage and will be a key enabler of LTE. Major handset vendors operate under the assumption that LTE access infrastructure will be available from late 2009, with commercial services from 2010. Devices will be introduced in phases during 2009–2010. Vendors expect it to take 1–2 years thereafter for devices to become widely available.

Technology mix. Compared with standard UMTS, LTE offers significantly increased capacity for a given bandwidth (spectral efficiency), and other important capex and opex savings. Therefore, operators will deploy LTE at some point to meet the rising demand for data and falling revenue (per megabyte). However, as operators evolve their networks towards LTE, they may wish to retain some access technologies and replace others, resulting in a technology mix. This mix depends upon the operators’ legacy infrastructure and business objectives:

• LTE will not be deployed as a standalone solution. Instead, it will be used in conjunction with GSM, UMTS and a separate indoor solution. It is likely to be deployed as a capacity overlay on existing UMTS and GSM services, and a dedicated indoor coverage solution is needed in order to meet the demand for data from the home or office. For example, Telefónica plans to deploy femtocells, while Orange has opted for a UMA/Wi-Fi solution.

• Many operators will retain GSM900/1800 to support circuit-switched voice and legacy services, and to protect incoming revenue from roaming services for several years. The earliest GSM switch-off is likely to occur in 2015, the majority of GSM networks will remain until 2020.

• Operators are currently divided on HSPA+. Some operators, like T-Mobile, (typically those that have already deployed HSDPA7.2/HSUPA across a large proportion of their networks) will carry out software upgrades to HSPA and HSPA+ without deploying MIMO. Others, like Orange, plan to deploy HSPA+ as soon as the technology becomes available in order to provide capacity relief in high-density areas.

Operator strategies for network evolution: the road to LTE ix

© Analysys Mason Limited 2009 0: Summary

• WiMAX is expected to have limited impact on LTE deployments, although some global operators with footprints across both developed and emerging markets may choose to deploy WiMAX opportunistically in areas where spectrum is constrained.

Operators do still have scope to maximise their existing network assets, but this may be a short-lived opportunity in view of the exponential growth in mobile data in many markets.

Spectrum availability. The success of wireless solutions relies on the availability of spectrum and bandwidths, and the timing of auctions. The ability to take advantage of new spectrum allocations and the opportunity to potentially refarm existing GSM spectrum are two important developments that will shape LTE deployment strategies. Depending on the country, operators have a choice of spectrum, including: 700–800MHz, 900MHz and possibly 1800MHz, as well as new spectrum in 2.5–2.6GHz. The choices that operators make will determine range, capacity and building penetration, which, in turn, will affect cell count, and therefore opex and capex. It is crucial that operators understand what spectrum is going to be available and when.

Collectively, technology, devices, spectrum and traffic growth, along with the operators’ business objectives, are the cornerstones of operators’ strategies for deploying data-optimised networks.

In developed markets, operators’ networks will typically reach full capacity within the next 2–3 years. We expect to see widespread deployment of LTE by 2014, with early adopters deploying by 2012. In emerging markets, operators typically have less-evolved legacy infrastructure. Data demand is not as great as in many developed markets, so we expect to see the first deployments of LTE nearer to 2015.

In both developed and developing regions, we expect that many operators will evolve through the various access technologies (GSM, GPRS, EDGE, W-CDMA, HSDPA, HSUPA). We believe that there are four basic operator LTE deployment strategies, based on four operator types: the operator with both GSM and UMTS access networks; the UMTS-only operator, the GSM-only and the CDMA-only.

There will be anomalous situations where early adopters in emerging markets leapfrog from GSM directly to LTE. We expect this to occur in certain South-East Asian markets. The same will also be true for a number of CDMA operators, including Verizon, which is expected to be one of the first operators to deploy LTE.

The time is ripe for operators to develop a coherent network evolution strategy, based on the four cornerstone drivers (traffic growth, device availability, technology mix and spectrum availability) and to design a phased deployment that stays one step ahead of rising demand.

© Analysys Mason Limited 2009 1: Wireless data growth will drive network evolution

1 Wireless data growth will drive network evolution

By the end of 2008, mobile accounted for 20–30% of broadband subscribers in the more-advanced Western European markets, such as Austria and Ireland. The rise in data traffic carried over wireless networks in developed markets has taken many network operators by surprise. At the same time, revenue per megabyte is continuing to fall, driven down by increasing competition and the introduction of flat-rate pricing.

1.1 Network evolution will be largely driven by consumer demand for services delivered over high-performance networks

A number of factors are driving wireless data, as shown in Table 1.1.

Table 1.1: Factors that are driving the increase in wireless network data traffic [Source: Analysys Mason, 2009]

Type Factor Description

Network Widespread deployment of 3G technologies that are supporting greater usage and range of services

3G enhancements (including HSPA and broadband EV-DO) offer faster throughput, greater capacity, better quality of service and lower costs than previous generations of mobile technology

Improved mobile devices enable a greater range of data services

Plug-and-play wireless USB modems and advanced smartphones (with improved displays, user interfaces, processing power, memory and batteries) have become more widely available

Increasing indoor usage of mobile devices

Heavy indoor usage of traffic-intensive data services, such as mobile broadband services and mobile TV, will drive the demand for increased capacity

Devices

Increasing mobile penetration

Mobile penetration has reached near-saturation in developing regions, but there are opportunities for users to adopt secondary devices, such as USB modems

2 Operator strategies for network evolution: the road to LTE

1: Wireless data growth will drive network evolution © Analysys Mason Limited 2009

Type Factor Description

Video and HD are here The amount of music and video traffic has grown significantly in the last couple of years as a result of popularity of YouTube, Napster and iTunes, the increased availability of streamed radio content and the burgeoning number of professional and amateur podcasts. Music and video streaming and downloads are some of the most intensive applications available on the Internet. It was reported in the Telegraph in May 2008 that YouTube consumed as much capacity in 2007 as the whole Internet did in 2000.1 Videos, which account for about 99% of all bytes transferred, are growing in terms of file size and length: 90% of videos last longer than three minutes, up from under one minute in 1997

Increasing complexity and size of content

The mobile industry has shifted its focus from ‘mobile Web’ (WAP) to full-screen display of standard Web pages. Web pages are also becoming more complex, with increasing numbers of multimedia-rich pages. The average Web page size has more than tripled in the past five years2

Services/content

Affordable pricing and bundling

Reduced prices and unlimited tariffs improve affordability (particularly for traffic-intensive services)

1.2 Wireless network traffic from voice and data services will increase dramatically by 2015

The total volume of wireless network traffic generated from voice and data services in 2015 will be nearly ten times that in 2008 in developed regions, as shown in Table 1.2. Of this total, data will account for 94% by 2015 and almost three quarters (74%) will be generated indoors. Wireless network traffic in developing regions will follow similar trends, but will lag behind and will not increase to the same extent in the period to 2015.

1 http://www.telegraph.co.uk/news/uknews/1584230/Web-could-collapse-as-video-demand-soars.html.

2 Website Optimization, LLC (Ann Arbor, MI, 2008), Average Web Page Size Triples Since 2003. Available at http://www.websiteoptimization.com/speed/tweak/average-web-page/.

Operator strategies for network evolution: the road to LTE 3

© Analysys Mason Limited 2009 1: Wireless data growth will drive network evolution

Table 1.2: Key forecasts for mobile traffic in developed and developing regions [Source: Analysys Mason, 2009]

Traffic metric Developed regions Developing regions

2008 2015 2008 2015

Total traffic per month

57PB 557PB 50PB 307PB

Traffic per mobile user per month

56MB 455MB 22MB 83MB

Percentage of data in total traffic

49% 94% 7% 79%

Percentage of total traffic generated indoors

54% 74% 34% 62%

Revenue per megabyte

USD0.86 USD0.12 USD0.57 USD0.15

These forecasts assume a Base case scenario, as shown in Figure 1.1. In the Downside scenario, traffic per customer will be four times higher in 2015 than it was in 2008, while in the Upside scenario, traffic per customer will be 22 times higher. The variability of these forecasts implies that operators must carefully predict the demand for traffic from their customers and build a degree of flexibility into their plans.

Figure 1.1: Global wireless network traffic growth, 2008–2015 [Source: Analysys Mason, 2008]3

0

500

1000

1500

2000

2500

3000

2008 2009 2010 2011 2012 2013 2014 2015

Tota

l wire

less

net

wor

k tra

ffic

(PB

per

mon

th)

Base caseDownsideUpside

3 For more information, see Brydon, A. and Heath, M., Wireless network traffic 2008–2015: forecasts and

analysis, Analysys Mason (Cambridge, 2009).


1: Wireless data growth will drive network evolution © Analysys Mason Limited 2009

1.3 The volume of traffic will erode revenue

The combination of a substantial rise in wireless network traffic per user and a relatively modest increase in ARPU will result in a significant decline in revenue per megabyte between 2008 and 2015: from USD0.86 to USD0.12 in developed regions, and from USD0.57 to USD0.15 in developing regions. The cost of maintaining the existing network will soon begin to erode operators’ profit margins and will eventually outstrip revenue, as shown in Figure 1.2.

Figure 1.2: The cost of data traffic using legacy networks, and revenue [Source: Analysys Mason, 2009]

Time

Volume ofnetwork traffic

Cost ofexisting network

Revenue

Voice era Data era

Network costs outstrip revenue

Time


Cost ofexisting network

Revenue

Voice era Data era

Network costs outstrip revenue

In developed countries, mobile broadband revenue is overtaking that of voice (developing regions will lag behind due to a greater initial reliance on circuit-switched voice). Mobile broadband revenue in Europe is expected to grow from USD9.07 billion in 2008 to USD33.91 billion in 2014, and is a key element of operators’ business models.4 In order to support usage-intensive applications profitably and maintain an optimum cost structure, network operators will need to develop their network architecture with the aim of reducing the network carriage cost per megabyte.

Together, these conditions point to the need for: a data-optimised network architecture, a network carriage cost per megabyte that realises a return on investment and a level of indoor service that is comparable to outdoor coverage. The introduction of network enhancements, such as HSPA, HSPA+ and LTE – possibly in combination – will address these needs, but operators are facing complex technology choices.

4 For more information, see Hatton, M., Mobile broadband in Europe: forecasts and analysis 2009–2014,

Analysys Mason (Cambridge, 2009).

© Analysys Mason Limited 2009 2: Operators need to understand, and make, a complex array of technology choices

2 Operators need to understand, and make, a complex array of technology choices Operators will face a number of strategic decisions about deploying a data-optimised network. When upgrading their networks, operators must consider the following key issues:

• the advantages of LTE • other technology options – CDMA, TD-SCDMA and WiMAX • the implications of existing assets and legacy services for future deployment plans • the provision of indoor coverage.

2.1 LTE offers a compelling proposition

LTE is the latest development in the UMTS family of standards. It is the response from the UMTS standards body, 3GPP, to the rising demand for data among mobile subscribers. LTE is an all-IP network in both the core and access sub-networks and, compared with legacy networks, is optimised to carry data at a significantly reduced cost, as shown in Figure 2.1.

Figure 2.1: Typical mobile network costs per subscriber per cellular sector [Source: Analysys Mason, 2009]

0

10

20

30

40

50

60

70

80

40 80 160 320 640 1280

Customers per sector

Ope

ratin

g co

sts

per s

ubsc

riber

(E

UR

per

mon

th)

HSDPALTE

1GB per customer

2GB per customer

4GB per customer

1GB per customer

2GB per customer

4GB per customer

8GB per customer


2: Operators need to understand, and make, a complex array of technology choices © Analysys Mason Limited 2009

Compared with previous mobile technology generations, LTE:

• offers significant performance improvements • has the support of the mobile industry • can offer significant opex and capex savings • offers improved use of spectrum.

LTE offers significant performance improvements

LTE offers reduced latency, improved quality of service and the ability to operate with advanced antenna systems, which can improve spectral efficiency.5 For example, LTE with 2×2 MIMO is approximately 2.5 times more spectrally efficient than HSPA.6

LTE has strong industry support

Radio access standardisation of LTE was finalised at the end of 2007 and the remaining specifications are expected to be completed by the end of March 2009.

Already 85% of the world’s mobile subscribers are on GSM/UMTS-based networks and the number is expected to increase. Both base-station and handset vendors are taking advantage of this huge installed base of subscribers. Existing 3G infrastructure and handset vendors will offer LTE-compatible technology and devices, creating a multi-vendor market with interoperable products and potentially cost-effective upgrades from existing equipment, which will reduce costs and increase choice through competition. Base-station vendors are designing LTE base stations that use existing base-station sites to provide the same coverage as GSM and UMTS services. Base-station vendors claim that they will attempt to enable MNOs to re-use existing assets wherever possible, minimising the requirement for investment.

Consequently, a number of major operators have confirmed that they have a long-term strategy to move to LTE, building confidence throughout the industry that the value chain of base-station and device manufacturers and vendors will be able to create economies of scale and support a variety of devices and services. Product development for LTE is in progress and the first LTE systems will go to market at the end of 2009.

5 Spectral efficiency is a measure of the ability of a radio bandwidth to carry information, and is measured in bits

per hertz.

6 MIMO multiple antenna technology improves the range and capacity of a communications link for a given bandwidth.



LTE can offer significant opex and capex savings

Compared with legacy networks, LTE improves spectral efficiency and requires fewer nodes to deliver the same network capacity, or a greater capacity can be achieved for the same number of nodes. Furthermore, LTE employs a flat-RAN architecture, which also reduces the number of network nodes. Jointly, the benefits of improved spectral efficiency and a flat-RAN architecture reduce network carriage costs and create a cost-growth curve that tends to track revenue rather than demand, allowing the operator to maintain a healthy profit margin, as shown in Figure 2.2.

Figure 2.2: LTE can deliver profit by significantly reducing opex and capex [Source: Analysys Mason, 2009]


Cost oflegacy network

Revenue

Voice era Data era

Legacy network costs outstrip revenue

Cost ofLTE network

Profit


Cost oflegacy network

Revenue

Voice era Data era

Legacy network costs outstrip revenue

Cost ofLTE network

Profit

Compared with previous access technologies, like GSM or UMTS, the smaller size of the LTE access node brings other operational gains, including easier and quicker siting of the node, lower power consumption and a much smaller space requirement.

Deploying LTE networks could allow operators to switch off their GSM networks, which would offer further savings. However, this is one of the more contentious benefits. Switching off GSM would simplify network management, with the corresponding operational cost savings and the earlier an operator can decommission its GSM network, the greater such gains are likely to be. At the same time, there would be costs associated with:

• migrating the GSM customer base and other legacy services • the loss of GSM roaming and the attendant revenue stream • the loss of revenue from GSM circuit-switched voice.

The costs will have to be carefully weighed against the savings.



Many operators have tried RAN sharing as a way of reducing costs. When RAN sharing fails, it is often because operators are not able to agree how to apportion value to existing assets. Deploying a new (LTE) network offers operators the chance to plan a shared access network without being concerned about how to value existing assets. Sharing an LTE RAN will help operators to justify the investment. Indeed, network sharing may be the only way that operators can afford to invest in LTE.

LTE offers improved use of spectrum

The choice of spectrum is an important one for operators as it directly affects capex and opex. LTE offers a greater degree of spectrum flexibility compared with earlier generations of mobile technology. It is available in a wider number of carrier frequencies and various bandwidths are available from 1.25MHz to 20MHz. In comparison, UMTS has a fixed, and therefore relatively inflexible, bandwidth requirement of 5MHz.

LTE also supports both paired and unpaired radio spectrum, using its FDD and TDD modes. LTE is deployable in a range of spectrum bandwidths (from 1.25–20MHz), allowing MNOs to use current spectrum allocations, and has also been designed to use the IMT-2000 extension band, which we discuss in more detail in Section 4.1.

In later releases of the standard, LTE will be able to add together blocks of spectrum, or concatenate spectrum, for different parts of the radio spectrum to create one large bandwidth. This will be a major step forward in improving capacity, spectral efficiency and harmonising spectrum usage worldwide.

2.2 There are other technology choices, including CDMA, TD-SCDMA and WiMAX, but most operators will migrate to LTE

LTE is the natural evolutionary successor to HSPA (and HSPA+), and it could also offer a smooth migration from all other the major systems, including CDMA, TD-SCDMA and WiMAX, as shown in Figure 2.3.



Figure 2.3: All technology roadmaps can lead to LTE [Source: Analysys Mason, 2009]

GSM EDGE

W-CDMA

CDMA 1×

WiMAX 16d

TD-SCDMA

WiMAX 16eTDD

EV-DO Rev A/B

HSPA HSPA+

EDGEEvolution

TD-SCDMAEvolutions

UMB

WiMAX 16mTDD/FDD

LTE Rel.10FDD/TDD

LTE Rel. 8/9FDD/TDD

GPRSGSM EDGE

W-CDMA

CDMA 1×

WiMAX 16d

TD-SCDMA

WiMAX 16eTDD

EV-DO Rev A/B

HSPA HSPA+

EDGEEvolution

TD-SCDMAEvolutions

UMB

WiMAX 16mTDD/FDD

LTE Rel.10FDD/TDD

LTE Rel. 8/9FDD/TDD

GPRS

However, LTE is not the only solution for a data-optimised network. It is important to consider the role of other technologies, including CDMA and TD-SCDMA, and particularly WiMAX.

WiMAX 16e was initially presented as competing with LTE, however deployments have shown that this not the case. WiMAX 16e uses 2.3GHz, 2.5GHz and 3.5GHz licensed spectrum and 5.8GHz unlicensed, and the fact that WiMAX 16e is TDD has allowed it to find a number of applications in parts of the world where spectrum is an issue. Moving into 4G, the specification for the next generation of WiMAX – IEEE 802.16m – is very similar to LTE in terms of frequency, spectral efficiency, latency and advanced antenna systems. As a consequence, WiMAX 16m will compete with LTE for the same customer base. LTE will have the advantage of being able to draw on a significant subscriber base, but as there is no single service that demands LTE, it is plausible that there will be strong competition from WiMAX based on price. The outcome of this competition will depend on the willingness of members of the LTE and WiMAX value chains to subsidise price in order to see their technology succeed.

Because LTE will evolve from the vast base of GSM/UMTS networks, LTE will have major commercial advantages over WiMAX. LTE will provide performance similar to that of WiMAX 16m, so there is no justification (from a performance perspective) for UMTS operators to change to WiMAX. Discussion with leading vendors and operators reveals that most MNOs will upgrade their networks to LTE rather than WiMAX.

We predict that UMTS technologies will dominate wireless broadband services, and will have possibly twenty times as many users as WiMAX by the end of 2015. It is likely that most WiMAX deployments will be in developing regions, where it is already finding ready markets. We believe that 92% of the 98 million WiMAX customers at the end of 2015 will be in developing regions. While the majority of MNOs will follow an LTE roadmap,



WiMAX will be one of the technologies available to UMTS operators that might want to take advantage of opportunities in developing countries and/or countries where ADSL coverage at broadband speeds is poor. Telefónica is one such example.

Telefónica is working with WiMAX in selected markets in order to understand its potential in specific scenarios

In Latin America, lack of spectrum is a big issue. With the exception of Brazil, Telefónica does not have dedicated 3G spectrum in any of the Latin American markets in which it operates. Furthermore, in most of these markets, Telefónica has spectrum allocations of 2×12MHz or 5MHz at 850MHz, and 10MHz or 20MHz at 1900MHz, which is half the amount of spectrum it has in European markets. According to Telefónica, “We have to make the most of the spectrum we have to deliver the capacity that our clients demand. We are exploring options that we have with our 2.5GHz and 3.5GHz assets. WiMAX continues to be an option in very specific market scenarios, but we have clearly defined a development path that focuses on 3GPP, not WiMAX.”

Similarly, Orange has launched WiMAX in Africa, taking advantage of the availability of WiMAX licences where 3G licences are not yet available.

Orange is pursuing an opportunistic approach to WiMAX

Orange has launched WiMAX in four African countries – Botswana (July 2008), Cameroon (April 2008), Central African Republic (December 2007) and Mali – as the first step to entering the broadband market. According to Orange, WiMAX is well-suited to the delivery of fixed services and as such offers an advantage in developing countries where broadband network deployments are limited: “We are taking an opportunistic approach, launching services in specific African markets, where WiMAX licences have become available and there is currently no 3G infrastructure.”

Although WiMAX offers good radio performance, Orange identifies a number of limitations, including interoperability roaming, security for authentication (by not offering SIM cards) and limited shipments of handsets and base stations, which mean that it would not consider deploying the technology outside countries with limited fixed infrastructure.

CDMA, a narrowband technology similar to W-CDMA, has been widely deployed globally, notably in the USA and parts of Asia. CDMA was pioneered by Qualcomm, but Qualcomm shifted its interest to LTE in November 2008, essentially closing CDMA’s evolution path.

TD-SCDMA is a proprietary technology based on TDD W-CDMA. It was devised in China to compete with Western technologies and to reduce the cost of intellectual property rights being paid to Western companies. While there is limited global interest in TD-SCDMA, it will have a significant impact in China, where China Mobile has adopted it as the 3G



evolution technology.7 It remains to be seen whether TD-SCDMA will evolve or join the LTE path. It is part of 3GPP specifications from Release 4 onwards.

2.3 Existing assets and legacy services will influence future deployment plans

One of the key questions facing operators is how much they will be able to maximise the use of existing network assets. By the end of 2008, 3.5 billion mobile subscribers were using networks based on the GSM family of standards, of which only 350 million subscribers were on 3G UMTS. In theory, operators could migrate a substantial number of subscribers to UMTS before moving on to LTE. However, this ignores local regulatory constraints, availability of suitable spectrum, local legacy services and the relatively limited roaming capability of UMTS.

3G spectrum has not yet been fully allocated in many countries and UMTS could be deployed more widely, but operators will need to weigh the cost of upgrading older technology against the cost of deploying a new technology with a longer lifespan ahead of it. In countries where UMTS is deployed, many operators feel that their existing assets are not yet fully utilised. For example, in November 2008, Vittorio Colao, CEO of Vodafone, commented that Vodafone’s European networks were running at only 32% of their capacity.8 Operators must carefully weigh up the longer-term improved profit of LTE networks compared with the investment that is lost if operators do not maximise their UMTS networks. For this reason, many operators will consider using LTE as a capacity overlay where required.

What are the reasons for deploying HSPA+?

In developed countries, many operators with UMTS HSPA networks have to decide whether to upgrade to HSPA+ or whether to move directly to LTE. A number of operators, including Swisscom, Telefónica and T-Mobile, will carry out software upgrades to HSPA and HSPA+ without deploying MIMO. Typically, these operators have already deployed HSPA across a large proportion of their networks, offering HSDPA at 7.2Mbit/s in major cities.

7 The Chinese telecoms regulator issued China Mobile with a technology-specific licence for TD-SCDMA in

January 2009. It is expected that China Mobile will begin trials in 2009 and will launch the technology commercially in 2010.

8 The Financial Times: http://www.ft.com/cms/s/0/18d0ebb0-cf85-11dd-abf9-000077b07658.html.



There are a number of arguments in favour of migrating straight to LTE from HSPA, as shown in Table 2.1.

Table 2.1: SWOT analysis of migration from HSPA directly to LTE [Source: Analysys Mason, 2009]

Strengths Opportunities

• Improved spectral efficiency at 20MHz

• Offers speeds of up to 144Mbit/s downlink and 50Mbit/s+ uplink speeds in <20MHz spectrum

• MIMO evolution path will be more likely for LTE

• Competitive differentiation

• Wider choice of carriers and bandwidths. LTE TDD will be available

• Can re-use GSM sites

• Future proofing against capacity demand

Weaknesses Threats

• Significantly more investment required in infrastructure, because LTE is based on OFDMA, a completely new modulation scheme (whereas HSPA+ is based on W-CDMA). LTE requires a new set of radio access and core infrastructure components – not only are the radio access algorithms different from legacy networks, but also the signalling and control protocols from the access to the core are significantly different

• Backward compatibility – handover from LTE to HSPA+ is complex, compared with handover from HSPA+ to HSPA

• LTE chips will be expensive initially, because they will require a new semiconductor ecosystem to support them

• Significant disruption to existing services and redirection of resources

• If LTE operators are not able to secure 20MHz spectrum, they will not be able to realise the advantages of LTE in terms of speed

• May cause operators to overextend themselves financially

• Anticipated data demand may not be realised

Other Tier 1 operators, including Orange, have conducted public trials of HSPA+ with hardware MIMO: clearly there is a feeling that HSPA+ still has growth potential, as shown in Table 2.2. This is particularly true where operators can benefit from using HSPA+ as a high-capacity overlay when demand is great.



Table 2.2: SWOT analysis of migration from HSPA to HSPA+ (Release 7) [Source: Analysys Mason, 2009]

Strengths Opportunities

• HSPA+ uses the existing modulation scheme (W-CDMA), whereas LTE requires a new set of radio access and core infrastructure components

• Spectral efficiency of HSPA+ (when MIMO is included) is close to that of LTE in 5MHz spectrum

• Backward compatibility – handover from HSPA+ to HSPA is relatively simple (the connectivity between RNC and the core (SGSN and GGSN) remains as before), compared with handover from LTE to HSPA+

• Handsets and terminals are widely available

• HSPA+ USB modems are likely to be cheaper than LTE equipment, which will require a new semiconductor ecosystem

• If operators are unable to secure 20MHz spectrum, they will not be able to realise the advantages of LTE in terms of speed

• Reduces capex

• Less disruptive upgrade path

• Allows investment to match demand – smooth upgrade path

Weaknesses Threats

• HSPA faces serious coverage issues, as increased data usage causes cell shrinkage. To counter this, network operators would have to split cells to handle the load

• The backhaul networks (being dimensioned for voice and low data rates) are not ready to support mass adoption of mobile data services. In addition to investment in MIMO equipment, mobile networks would need upgrades to the backhaul and the core network, which will be costly

• Requires another site visit to upgrade to LTE

• Competition from operators that have migrated directly to LTE

• Running out of capacity

Orange plans to deploy HSPA+ as soon as the technology becomes available, which is likely to be in mid-2009.

Orange plans to upgrade to HSPA+ (Release 7) as soon as the technology becomes available

Orange plans to upgrade its network gradually, focusing on high-density areas, as and when capacity is required: “HSPA+ allows you to select the options you want to include in your network and, as such, you can have focused deployment in areas where there are issues with loading or performance. Secondly, it is unlikely that we will see a complete deployment of LTE before 2011 or 2012. If you want to improve the user experience and cater for the growing demand for dongles and new terminals like the iPhone, you need to upgrade.”



It is, however, clear that Orange is focusing on LTE in the long term. It is currently in discussion with LTE equipment vendors and aims to begin trials in 2009, with its first deployments envisaged for 2010.

Orange’s initial LTE deployments will focus on dense, urban areas where the demand for network capacity is high. From there it will extend coverage opportunistically. Orange’s strategy will also vary between countries, depending on regulatory constraints.

As we will see in Section 2.4, the question of whether to upgrade the macrocell network to HSPA+ or LTE is closely linked to indoor coverage, which is a major consideration in operator access network deployment strategies.

2.4 LTE will not be deployed as a standalone layer, but as a complement to a dedicated indoor coverage mechanism

Data traffic has grown exponentially in the past 12 months, and an increasing proportion of data usage is occurring indoors. We predict that more than 86% of data traffic will originate indoors in developed regions and more than 60% in developing regions by 2015. To be successful, operators will need to fulfil subscribers’ expectations in terms of coverage. Operators will need to improve the 3G user experience in indoor environments to cater for this growing usage trend. Direct indoor coverage using femtocells or Wi-Fi is one way of achieving this.

The widespread use of a direct indoor coverage solution will affect operators’ plans to deploy LTE by offloading the demand from the macrocell network. Figure 2.4 shows the growth in traffic from 2008 to 2012 for a mobile operator with 10 million subscribers. Demand will outstrip HSPA capacity in 2010–2011 and HSPA+ capacity by 2012. However, if indoor traffic is offloaded to a direct indoor solution network, HSPA and HSPA+ will offer sufficient capacity for 2012 and beyond.



Figure 2.4: Network traffic generated by an example mobile service mix, split between indoor and outdoor usage, 2008–2012 [Source: Analysys Mason, 2009]

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

2008 2009 2010 2011 2012

Tota

l dow

nlin

k tra

ffic

per m

onth

(ter

raby

te)

Outdoor Indoor Indoor capacity overhead

LTE (20MHz)

LTE (10MHz)

HSPA+ (10MHz)

HSPA (10MHz)

Networkcapacities

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

2008 2009 2010 2011 2012

Tota

l dow

nlin

k tra

ffic

per m

onth

(ter

raby

te)

Outdoor Indoor Indoor capacity overhead

LTE (20MHz)

LTE (10MHz)

HSPA+ (10MHz)

HSPA (10MHz)

Networkcapacities

LTE (20MHz)

LTE (10MHz)

HSPA+ (10MHz)

HSPA (10MHz)

Networkcapacities

There are two direct indoor solutions: femtocells and Wi-Fi.

Femtocells

Femtocells are small UMTS base stations, which users can install in their homes and connect to mobile operators’ switching networks through the home ADSL line. In contrast, femtocells for businesses are usually installed and integrated into the existing IT network of the business by a trained installation team provided by the operator. Backhaul is via the existing external connectivity. While the femtocell solution is technically elegant, it is not straightforward to implement and has at least three major hurdles to overcome.

• The business case is challenging. • The success of the solution depends on the use of an ADSL backhaul, which will

disadvantage the mobile operator, unless its parent company also owns the backhaul. • Operators will have to learn to manage millions of femtocells, as opposed to tens of

thousands of base stations, which was their previous experience.

Much has been said about the technical challenges of a femtocell deployment associated with managing interference and handover, but these have largely been overcome.

Mobile operators must carefully consider their position before relying on a femtocell solution to meet the increasing demand for capacity. A number of operators, including T-Mobile, TeliaSonera and Vodafone, are conducting femtocell trials. The hype



surrounding femtocells seems to have diminished and operators are doing more work behind the scenes to determine the validity of a femtocell solution.

Telefónica will deploy femtocells alongside LTE

Telefónica started considering femtocells more than three years ago. It has been testing several vendors’ solutions, both for access and coverage.

The operator has a wide scope of operations and the choice of technical solutions for improving in-building coverage differ in each market. According to Telefónica: “In Spain, for example, Telefónica is a main player in the fixed broadband market, with significant market share. In other markets, where Telefónica does not operate a fixed network or has less-comprehensive coverage, it may decide to adopt different approaches.”

There are some technical and commercial challenges to be overcome, but Telefónica believes that the main technical problems have largely been addressed.

If successful, the number of femtocells that an operator may deploy could reach into the millions. For a mobile operator, the deployment of such large numbers of nodes and their management is more akin to a domestic Wi-Fi deployment rather than a network of base stations, which is typically tens of thousands. This change of scale brings operational challenges. Telefónica feels that its ability to manage large numbers of femtocells will benefit from its experience of managing its fixed network.

Telefónica expects that HSPA femtocells will first appear as access in public areas. The operator is exploring the motivations for deploying femtocells in the home, in order to anticipate its customer needs and expectations, but it is also considering alternative solutions for improving in-building coverage. Potentially, LTE deployments may count on femtocells as dedicated capacity solutions.

According to Telefónica, when deploying femtocells in an enterprise, the operator business case will need to account for on-site customer support, as there may be issues in terms of self-installed self-optimised solutions that may make use of already existing IT infrastructure. This is not the case for femtocell deployments in the home, where easy self-installation and ‘plug-and-play’ functionality will be mandatory.

Wi-Fi with ADSL

Femtocells are not the only means of providing indoor coverage. Wi-Fi, combined with WiMAX or ADSL, is another serious contender. Wi-Fi routers are commonplace in both homes and businesses. These routers are self-installed devices, and have a number of advantages over the femtocell direct indoor solution. The business case for Wi-Fi is well proved and the solution is technically efficient. Furthermore, because Wi-Fi routers use unlicensed frequency bands, operators are not required to manage interference with an outdoor network. However, this ease of use also has its disadvantages. Wi-Fi routers do not offer a carrier class of service, which is particularly important for business deployments. Wi-Fi installations are not bound by the type of service level agreements that an enterprise



would expect. Finally, because of the nature of the Wi-Fi – plug and play – installation, enterprises tend to be casual about security and often do not implement a full link encryption.

Orange is deploying a UMA/Wi-Fi solution to improve indoor coverage

Even with the deployment of advanced data-optimised network technologies, Orange recognises that indoor coverage can be limited because of site density. Site density is expensive to increase and, as a result, it is considering UMA/Wi-Fi solutions to offload traffic to its fixed network.

Orange operates a fixed–mobile handset service, Uniq, based on UMA/Wi-Fi, in countries where it has fixed DSL coverage, including France (October 2006), Poland (April 2007), the UK (November 2006) and Spain (December 2006).

3: Spectrum issues will affect MNOs’ LTE deployment strategies © Analysys Mason Limited 2009

3 Spectrum issues will affect MNOs’ LTE deployment strategies

Operators’ strategies for LTE will be heavily influenced by their choice of carrier frequency. There was a time when the decision was not so complicated and operators merely decided whether or not to buy what was on offer from the regulator. Now, spectrum is becoming available in many new bands, and offering operators a much wider choice than ever before. The four most important developments in current spectrum matters are:

• the ITU allocation of a 3G expansion band in 2600MHz • the refarming of GSM 900MHz (and possibly GSM 1800MHz) • the availability of spectrum in the 700–800MHz range, as a result of the switchover from

analogue to digital broadcast TV • the refarming of wireless cable 1700–2100MHz spectrum in the USA for Advanced

Wireless Services.

3.1 Operators will deploy LTE in the IMT-2000 expansion band

In 2000, the World Radio Congress identified new spectrum for 3G technologies in the 2500–2690MHz frequency band. This is known as the IMT-2000 or 3G expansion band. Many countries plan to auction 2500–2690MHz spectrum in the next two years, if they have not done so already, as shown in Figure 3.1.

Figure 3.1: Timeline for 2500–2690MHz spectrum auctions [Source: Analysys Mason, 2009]

3Q05

New Zealand

(December)

Sweden (May)

Norway(November)

July2008

January2009

2Q2009

2Q2010

January2008

Hong Kong

Spain,Denmark,Germany

Austria, Belgium,

Switzerland,Finland

4Q2009

France

Auction status:CompletedExpected

3Q05

New Zealand

(December)

Sweden (May)

Norway(November)

July2008

January2009

2Q2009

2Q2010

January2008

Hong Kong

Spain,Denmark,Germany

Austria, Belgium,

Switzerland,Finland

4Q2009

France

Auction status:CompletedExpected


© Analysys Mason Limited 2009 3: Spectrum issues will affect MNOs’ LTE deployment strategies

Unlike UMTS, LTE operates over a range of bandwidths, from as little as 1.25MHz, but with greater spectral efficiency at bandwidths above 10MHz, ideally 20MHz. The 2600MHz band offers good bandwidth, and most network operators have publicly declared an interest in deploying LTE in this band. However, compared with lower frequencies, the range of a 2600MHz cell is limited. This can be overcome, but requires a high density of cells, at greater expense. As a consequence, most operators are considering using the IMT-2000 expansion band as a capacity overlay where there is a greater demand for capacity and range is less of a concern, for example in urban areas.

3.2 The timing of GSM 900MHz (and GSM 1800MHz) spectrum refarming is critical

Operators are unable to speculate about whether they will use LTE in any refarmed spectrum: in most cases, it is not known when the spectrum will become available and, therefore, operators do not know how mature LTE may be at that time. Furthermore, they do not know what other spectrum may become available in the meantime.

We believe that, where refarming is already permitted or regulatory approval is imminent, there may be a good business case for a GSM operator to deploy UMTS900, rather than LTE, in order to provide 3G rural coverage, increase 3G capacity, or both.

Operators that are launching UMTS900 networks are expecting the greatest usage to come initially from mobile broadband services based on USB modems or datacards, and envisage that voice services will follow as subscribers replace their 2G handsets with phones supporting UMTS900/2100 – this may take several years, especially in the current economic climate. If early switch-off of the GSM network is shown to offer significant cost benefits, operators will need to subsidise 3G handsets in order to encourage their subscribers to upgrade more rapidly, but the costs and benefits of doing so would require careful consideration.

Elisa and DNA have already deployed UMTS900

Operators, such as Elisa and DNA in Finland, have already deployed UMTS900 in rural areas and plan to create nationwide UMTS900 networks as quickly as possible as an improved coverage layer (and UMTS2100 as a capacity layer). Both operators intend to decommission their GSM networks as soon as possible, with Elisa aspiring to do so by 2015.

Where refarming measures are delayed or take time to implement, it becomes increasingly likely that an operator will migrate to LTE900, rather than UMTS900. There are a number of benefits of deploying LTE900, rather than UMTS900.


3: Spectrum issues will affect MNOs’ LTE deployment strategies © Analysys Mason Limited 2009

• UMTS deployments require a full 5MHz of spectrum, whereas LTE, which can be deployed in spectrum bands as small as 1.25MHz and up to 20MHz, allows for the allocation of spectrum to be more flexible and maximises the use of available spectrum.

• The coverage of GSM and LTE is similarly predictable, because like GSM, and unlike UMTS, the range of an LTE cell is unaffected by traffic on the cell. Therefore, GSM sites can readily be used to deploy LTE, yielding considerable capex savings.

If operators migrate to LTE900, then GSM900 is likely to be maintained for voice services, global roaming and certain M2M applications, possibly as late as 2018–2020.

The future use of any 1800MHz spectrum that an operator holds is linked to the future of GSM900. Unlike UMTS900, for which the supply of network equipment and devices is rapidly increasing, there is relatively little vendor backing for UMTS1800 equipment or devices at present. As a result, we regard refarming to LTE at 1800MHz as a more probable outcome than deploying UMTS1800. Operators are therefore likely to retain GSM services at 1800MHz until refarming at that frequency is required for capacity reasons, and will then deploy the most spectrally efficient technology at their disposal at that time. How quickly migration to LTE1800 becomes feasible will depend on the availability of the LTE standards, network equipment and devices. Vendors are claiming that operators want to deploy LTE in the 1800MHz band. If this is true, then we expect network equipment and devices to be available within the next three years. In the short term (3–5 years), it is likely that LTE will not replace GSM1800, but will be deployed in addition to GSM1800.

3.3 Digital dividend spectrum may be used for mobile communications

Many countries are converting from analogue to digital broadcast TV – this is known as the digital switchover. Some of the spectrum that was committed to analogue TV broadcasting may be freed for other uses, including mobile communications – this spectrum is known as the digital dividend. Some countries in Europe, such as the Netherlands and Sweden, have already completed their digital switchover. Most countries will complete it by 2012 – for example, France (in 2011) and the UK (in 2012) – while some, such as Russia, will complete it as late as 2015. North America is expected to complete the process in 2009.

The sub-bands of freed spectrum that are of greatest interest to mobile operators are: 790–862MHz in Europe; 698–862MHz in the USA and Americas; and 790–960MHz in Asia–Pacific. The benefits of 700–900MHz spectrum are good coverage and possible advantages when providing indoor coverage.


© Analysys Mason Limited 2009 3: Spectrum issues will affect MNOs’ LTE deployment strategies

3.4 The US regulator has made spectrum available for Advanced Wireless Services

In 2006–2007, the FCC auctioned spectrum in two segments (1710–1755MHz and 2110–2155MHz) for Advanced Wireless Services – mobile broadband and other 3G services. The spectrum was previously allocated to MMDS or wireless cable. Most of the spectrum was awarded to T-Mobile USA in 2007.

Advanced Wireless Services spectrum offers reasonable coverage, similar to GSM1800, and is offered in 10MHz and 20MHz bandwidth pairs, providing good capacity. Therefore, the spectrum could be used for coverage only, or for coverage and capacity.

In summary, the availability of new spectrum allocations and the opportunity to refarm existing GSM spectrum are two key areas that will enable LTE deployments. In the future, operators should be aware of the various forthcoming spectrum auctions within their particularly market(s) and align them with their RAN evolution strategies.

4: Device availability will be key to realising data growth © Analysys Mason Limited 2009

4 Device availability will be key to realising data growth

The successful deployment of any new access technology relies on reasonably priced devices being readily available. Deployment of LTE and the take-up of advanced data services by subscribers are no exception to this rule.

4.1 LTE-compatible devices will be available from 2H 2009

Handset vendors are likely to release LTE-compatible handsets at the beginning of 2010, as shown in Figure 4.1, working under the assumption that the first commercial LTE networks will be available as early as the end of 2009. The USA awarded spectrum in the 700MHz band in March 2008 (mostly to operators AT&T and Verizon) and it is likely to be one of the first countries to launch LTE (along with Japan). Qualcomm and Motorola are likely to be among the first vendors to deploy LTE-enabled devices.

Figure 4.1: Expected LTE device availability [Source: Analysys Mason, 2009]

3Q02 3Q03 3Q04 1Q05 3Q05 1Q06 3Q062009 2010 2011 2012

Handsets

Laptops with

embedded modules

USB modems

and datacards

3Q02 3Q03 3Q04 1Q05 3Q05 1Q06 3Q062009 2010 2011 2012

Handsets

Laptops with

embedded modules

USB modems

and datacards

LTE devices will be introduced in phases.

• USB modems or datacards will be the first devices to be introduced and are likely to be available in 2H 2009. These device forms are already common with UMTS. LTE will initially support data-intensive services for laptop users.

• Handsets will be available approximately one year after the first USB modems are available. Handsets will initially be for data only. Voice handsets will probably not be available until LTE VoIP is commercialised. This will first require the standardisation of


© Analysys Mason Limited 2009 4: Device availability will be key to realising data growth

VCC, part of IMS service continuity, which is due in Release 10 of 3GPP standards, and not before 2010.

• Laptops with embedded modules are likely to be made available 1–2 years after the introduction of USB modems. The design cycles are longer for embedded solutions. The embedded modules, particularly for LTE, require additional support for MIMO.

Clearly, the growth in demand for data from subscribers, which is the catalyst for LTE development, will also have an impact on the design of handsets and devices with embedded modules. The likely innovations are larger, higher-resolution handset screens on which to view video content, larger capacity for storing music tracks and dedicated keys to use solely for gaming. Long battery lives will be another key factor among the specifications.

Within 2–3 years, the dominant device for mobile broadband access will be the laptop with an embedded module. In the longer term, the delivery of LTE services will not be limited to traditional devices, but will instead extend to a range of devices including cameras, cars, energy monitors, environmental sensors, health-monitoring devices and many other machines.

As with any new technology, the retail cost of LTE devices is expected to come in at the top end of the market, targeting enterprise users and early adopters. We expect unit cost to fall as a result of economies of scale and competition. To this end, members of the value chain are working hard to ensure that take-up of LTE devices is strong. The Embedded Mobile Intiative, an industry group led by the GSMA, is aiming to foster collaboration between consumer electronics and industrial product manufacturers to encourage the adoption of embedded 3G modules, while developing common technical standards and platforms. Members of the initiative include operators and vendors, such as Huawei, KTF, Rogers Wireless, Smart, Softbank, Telecom Italia, Telefónica, Telstra and Vodafone. Qualcomm estimates that it will take “a couple of years” for LTE devices to become widely available.

As unit costs fall, laptop manufacturers will increasingly fit broadband access modules as standard. This threatens to separate the device and access markets, and MNOs will lose control of an important element of the value proposition.

4.2 The first LTE-compatible devices are likely to support four frequency bands

The first devices (USB modems and datacards) in Europe and Asia Pacific are likely to support four frequency bands: 900/1800/2100 and 2600MHz, GSM and W-CDMA. Devices to support refarming of 900/1800MHz and possibly 2100MHz, will be available


4: Device availability will be key to realising data growth © Analysys Mason Limited 2009

from 2010. Commercial launch, depending on demand, will follow at a later date, following a period of interoperability testing.

The 700MHz frequency band, auctioned recently in the USA, will pose a number of challenges for device manufacturers, including:

• the need for a larger antenna due to the longer wavelength • channel interference from adjacent frequency bands that are used for TV broadcasting • image channel interference due to GPS in band harmonics.

The characteristics of the 700MHz band will constrain LTE devices, resulting in a more onerous design process, the cost of which has to be absorbed in the value chain.

4.3 Vendors have agreed a standardised royalty framework for LTE handsets

In April 2008, seven major vendors – Alcatel-Lucent, Ericsson, NEC, NextWave Wireless, Nokia, Nokia Siemens Networks and Sony Ericsson – announced that they had reached an agreement for a framework for licensing intellectual property rights relating to 3GPP LTE. The framework sets a single-digit percentage royalty or, in the case of notebook computers, a fee of under USD10 on equipment using LTE technology. The framework for royalties should allow operators to avoid the delays and costs incurred when disputes over intellectual property rights disrupted the development of 3G handsets.

In summary, a number of key steps have already been taken to ensure that devices will be available in time for the launch of the first LTE networks. On-going co-operation between operators, application developers and handset manufacturers is essential if a reasonable choice of devices is to be available in time for future commercial launch dates. While vendors and operators are confident about device availability – expecting them to be ready in sufficient numbers for LTE launches – the success of LTE is particularly sensitive to the price and availability of devices. One would do well to remember the lessons of early deployments of UMTS, which were beset by the limited number of reasonably priced handsets. In the case of LTE, the range of frequencies and variety of device types may exacerbate the problem. If devices are not available or are available in insufficient quantities, this would represent a major barrier to the adoption of mobile broadband services over LTE.

© Analysys Mason Limited 2009 5: There are a number of pathways to LTE

5 There are a number of pathways to LTE

For the purposes of analysis and illustration, we consider possible strategies for the four main types of operator: the combined GSM and UMTS operator, the UMTS-only operator, the GSM-only operator and the CDMA operator.

5.1 The combined GSM and UMTS operator

Operators with both 2G and 3G spectrum, for example 900/1800MHz and 2100MHz, respectively, are likely to retain GSM voice at 1800MHz, upgrading through GPRS and even GPRS enhanced to support VoIP as it matures. Spectrum in the IMT-2000 expansion band, 2500–2690MHz, will be used as a capacity overlay in urban areas. The challenge is to provide reliable mobile broadband coverage indoors. In this, operators have three realistic choices:

• direct indoor coverage with femtocells and/or Wi-Fi • macrocell coverage using 900MHz • macrocell coverage using 2100MHz and/or 2600MHz.

With direct indoor coverage using femtocells/Wi-Fi, the quality of coverage is likely to be far better than macrocell coverage. Wi-Fi has the advantage of being a reliable technology that can be deployed today with a viable business model. Femtocells are a natural progression for the mobile operator from a GSM/UMTS family background, but the business case and technical challenges may yet prevent the widespread deployment of femtocells. Both Wi-Fi and femtocell solutions depend on the ADSL backhaul. This may make the mobile operator unacceptably vulnerable to the fixed operator and may be a serious weakness that fixed operators could exploit, especially as our predictions are that mobile operators may have taken more than 20% of fixed operators’ broadband business in developed European markets by 2013. Of course, a converged operator, with both fixed and mobile networks, would not feel threatened by fixed–mobile substitution.

If it transpires that a profitable femtocell business case is not possible, and/or the operator does not wish to rely on ADSL backhaul to provide its mobile broadband offering, then the operator will need to provide indoor coverage with the macrocell network. Whether the operator decides to buy 700MHz spectrum when it becomes available or to replace 900MHz GSM with data-optimised access technology, such as LTE or HSPA+, depends on


5: There are a number of pathways to LTE © Analysys Mason Limited 2009

when the spectrum becomes available, and how strong the demand for data is predicted to be. We believe that the operational costs of revisiting sites to make hardware upgrades mean that most operators will deploy LTE in 900MHz when it becomes available. It is likely that the 2G/3G operator will take advantage of incidental indoor coverage from 2100MHz, however the site density required to achieve good coverage indoors at 2600MHz will need to be weighed against the cost of acquiring new, low-frequency spectrum (900MHz).

5.2 The UMTS-only operator

Operators with 3G-only spectrum face problems different to those of their 2G/3G counterparts. Firstly, they would almost certainly like to acquire spectrum at lower frequencies, which can then be used to provide data coverage in rural areas and indoors. In rural areas, due to the limited population, the demand for capacity tends to be less, and coverage is the most important consideration. So a mobile operator may want to acquire some low-frequency spectrum in order to provide a thin (few sites and big area) layer of coverage. The operator would then need to decide between deploying LTE or HSPA+ in this spectrum. Probably, an operator would opt for LTE, given the improved capacity, and opex and capex cost savings. However, as with all other operators, the 3G-only operator must align this strategy with the provision of indoor coverage. In this case, the same arguments apply as for the combined 2G/3G operator apply. It is also highly likely that 3G-only operators would want to acquire 2600MHz spectrum to be used as a capacity overlay in heavily populated areas.

5.3 GSM-only operators

There are many operators with only GSM networks, particularly in emerging markets, where there may be limited fixed infrastructure, limited competition within the fixed telecoms market and low fixed broadband penetration. In developing countries, 900MHz GSM will probably be retained because of its good rural coverage and the high proportion of voice usage. It is likely that wireless broadband with desktop CPE will be used as a fixed broadband substitute.

Before LTE, the upgrade path for GSM-only operators would typically be through UMTS R99 and on to HSPA, as the demand grew for data services. Many operators are now considering moving directly to LTE, assuming that local spectrum regulations allow this strategy.

In developing countries, it is likely that some GSM-only operators will consider using WiMAX, due to its ability to operate in licence-exempt spectrum and its TDD capability. WiMAX lends itself well to mobile broadband use in countries where the frequency allocation is non-standard.



5.4 CDMA operators

The CDMA2000 standards (CDMA2000 1× RTT, CDMA2000 1× EV-DO, and CDMA2000 1× EV-DV) are the direct successors to IS-95 or cdmaOne. The technology is deployable in the 450, 700, 800, 1700, 1900, advanced wireless services and 2100MHz bands. It is deployed in regions that include Africa, Asia, the Caribbean and Latin America, Europe, the Middle East and North America.

Qualcomm was the lead sponsor of UMB, the data-optimised generation of CDMA2000. UMB was designed to compete with LTE for the wireless broadband market. Despite there being many operational CDMA2000 networks worldwide, the CDMA family of technologies lost favour with operators during 2007–2008. In November 2008, Qualcomm announced that it was ending development of the technology, favouring LTE instead; this effectively closes down the CDMA2000 evolutionary path for operators using CDMA2000 technology.

Most CDMA operators have now publicly declared that they will upgrade their networks using LTE. However, according to the Release 8 standard, LTE is not offered in all CDMA2000 bands, most notably 450MHz. We assume that LTE will be offered in this band in later standards in order to support the evolution to a data-optimised access network architecture of a substantial number of 450MHz CDMA networks worldwide.

CDMA operators will almost certainly upgrade their networks to LTE. Verizon Wireless in the USA is likely to deploy one of the earliest commercial LTE networks. The operator was reportedly planning to launch LTE services in 2009, but the US Senate confirmed in early February 2009 that the switchover to digital TV in the USA would be delayed by four months. This delay could jeopardise Verizon Wireless’s LTE launch schedule, because the operator had been expecting to use newly vacated 700MHz spectrum.

For CDMA operators, there is considerable debate about how to provide reliable indoor coverage, either femtocells/Wi-Fi on the one hand, or LTE at low frequencies such as 700MHz on the other.

In summary, there are as many pathways to data-optimised RANs as there are operators. The reasons for these different strategies vary from operator to operator, and possibly from site to site. Some sites will be the latest Node B allowing software upgrade, the remainder will require hardware upgrades. In the case of sites that require hardware upgrades, it may be more cost-effective to upgrade to LTE rather than HSPA+. In developed markets, we expect widespread deployments of LTE by 2015. Most operators will evolve through the various technologies: GSM, GPRS, EDGE, W-CDMA, HSDPA, HSUPA and HSPA+ and then move onto LTE. There will also be anomalous situations where early adopters in emerging markets leapfrog from GSM directly to LTE. We expect this to occur in rapidly evolving telecoms markets in South-East Asia. The same will also be true for a number of


5: There are a number of pathways to LTE © Analysys Mason Limited 2009

CDMA operators, including Verizon Wireless, which is expected to be one of the first operators to deploy an LTE network.

A number of operators have announced plans for LTE deployment, as shown in Figure 5.1.

Figure 5.1: LTE by geography/operator [Source: Analysys Mason, 2009]

68

AT&TTrials: nowLaunch: 2012

China MobileTrials: nowLaunch: 2010

= 2G & 3G operator

= CDMA operator

SwisscomTrials: TBALaunch: 2012

TeliaSoneraTrials: nowLaunch: 2010Vodafone

Trials: nowLaunch: 2010

Verizon WirelessTrials: nowLaunch: 2010

TelefónicaTrials: nowLaunch: 2010

T-MobileTrials: nowLaunch: 2010/2011

France TelecomTrials: nowLaunch: 2011/2012

NTT DoCoMoTrials: nowLaunch: 2010

TBA = To be announced

68

AT&TTrials: nowLaunch: 2012

China MobileTrials: nowLaunch: 2010

= 2G & 3G operator

= CDMA operator

SwisscomTrials: TBALaunch: 2012

TeliaSoneraTrials: nowLaunch: 2010Vodafone

Trials: nowLaunch: 2010

Verizon WirelessTrials: nowLaunch: 2010

TelefónicaTrials: nowLaunch: 2010

T-MobileTrials: nowLaunch: 2010/2011

France TelecomTrials: nowLaunch: 2011/2012

NTT DoCoMoTrials: nowLaunch: 2010

TBA = To be announced

Technology, devices, spectrum and demand, along with the operator’s business strategy, are the main factors that will determine the timing and strategy for LTE deployment for any particular operator. More than ever these will be different for each operator. Operators will need to consider their local choices carefully, in order to achieve an optimal position to deal with future data demand and demands on their revenue. We expect operators to differentiate themselves based on their understanding of the interplay of these factors and their subsequent technology choices and deployment strategies. Ultimately, there are as many pathways to a data-optimised RAN as there are operators.

For all operators, finding their way through this maze of complex decisions can be frustrating, but there are a few simple guidelines they can follow, as shown in Figure 5.2.



Figure 5.2: Deployment strategy guidelines [Source: Analysys Mason, 2009]

Get yourtraffic predictions

right

Plan yourindoor coverage

requirements(75% of demand)

Coverage SpectrumDemand Technologymix

1 2 3 4 5 6

Technologyoptions

Phaseddeployment

Understand thecosts andbenefits of

the technologyoptions

Understandspectrum availability,

costs and uses

Plan yourtechnology mixand spectrumrequirements

Optimise aphased deploymentwith the following

constraints:

• minimise networkcarriage costs

• maintain networkservice quality

• meet risingdemand

Get yourtraffic predictions

right

Plan yourindoor coverage

requirements(75% of demand)

Coverage SpectrumDemand Technologymix

1 2 3 4 5 6

Technologyoptions

PhaseddeploymentCoverage SpectrumDemand Technology

mix

1 2 3 4 5 6

Technologyoptions

Phaseddeployment

Understand thecosts andbenefits of

the technologyoptions

Understandspectrum availability,

costs and uses

Plan yourtechnology mixand spectrumrequirements

Optimise aphased deploymentwith the following

constraints:

• minimise networkcarriage costs

• maintain networkservice quality

• meet risingdemand

© Analysys Mason Limited 2009 Key to acronyms

Key to acronyms2G Second generation

3G Third generation

4G Fourth generation

3GPP The 3rd Generation Partnership

Project

ADSL Asymmetrical digital subscriber

line

ARPU Average revenue per user

CDMA Code-Division Multiple Access

CEO Chief executive officer

CPE Customer premises equipment

DSL Digital subscriber line

EDGE Enhanced Data for GSM Evolution

EV-DO CDMA2000 1× Evolution Data

Only

EV-DV CDMA2000 1× Evolution Data and

Voice

FCC Federal Communications

Commission

FDD Frequency division duplex

GB Gigabyte

GGSN Gateway GPRS support node

GHz Gigahertz

GPRS General packet radio service

GPS Global positioning system

GSM Global System for Mobile

Communications

GSMA GSM Association

HD High definition

HSDPA High-Speed Downlink Packet

Access

HSPA High-Speed Packet Access

HSPA+ Evolved High-Speed Packet Access

HSUPA High-Speed Uplink Packet Access

IMS IP Multimedia Subsystem

IMT-2000 International Mobile

Telecommunicationss-2000

IP Internet protocol

kHz Kilohertz

LTE Long-term evolution

MB Megabyte

Mbit/s Megabits per second

MHz Megahertz

MIMO Multiple-input multiple-output (a

smart antenna system)

MMDS Multichannel multipoint

distribution system

MNO Mobile network operator

OFDMA Orthogonal Frequency Division

Multiple Access

Opex Operational expenditure

PB Petabyte

RAN Radio access network

RNC Radio network controller

RTT Radio transmission technology

SGSN Serving GPRS support node

SIM Subscriber Identity Module

SWOT Strengths, weaknesses,

opportunities and threats

TDD Time-division duplex

TD-SCDMA Time Division Synchronous Code

Division Multiple Access

TV Television

UMA Unlicensed Mobile Access

UMB Ultra Mobile Broadband

UMTS Universal Mobile

Telecommunications System

USB Universal serial bus

VCC Voice call continuity

VoIP Voice over IP

W-CDMA Wideband Code Division Multiple

Access

WAP Wireless Application Protocol

Wi-Fi Wireless Fidelity

WiMAX Worldwide Interoperability for

Microwave Access (broadband

wireless based on the IEEE 802.16

standard)

© Analysys Mason Limited 2009 Research from Analysys Mason

Research from Analysys Mason

Analysys Mason provides a portfolio of research services that help organisations to understand major strategic shifts, as well as country- and region-specific trends, in the global telecoms industry. Through a combination of granular market data and forecasts, and independent qualitative analysis and insight, we enable clients to make informed strategic and tactical decisions, reduce risk and benchmark their business performance.

Split across five research streams, which cover core market data, infrastructure and technology, telecoms software, telecoms markets and wider industry dynamics, the research programmes provide a range of deliverables during a typical 12-month subscription period.

In addition to its subscription-based content, Analysys Mason also offers a custom research service, which helps clients to answer specific questions.

Analysys Mason’s research portfolio

Research programmes

Industry Dynamics

Telecoms Markets

Telecoms Software

Technology and Infrastructure

Core Data andForecasts

Industry Strategy Global Growth Markets

Fixed Broadband Mobile Broadband Voice

Consumer Content and Applications

ServiceAssurance

Fixed Networks Wireless Networks

ServiceFulfilment

Billing and Customer

Care

Telecoms Software Strategies

Service Provider Strategies

Network Equipment Manufacturer

Strategies

Service Delivery Platform Strategies

Telecoms Software Forecasts

Telecoms Software Market Shares

Telecoms MarketMatrix

Core Forecasts

Country Reports

Industry Strategy Global Growth Markets

Fixed Broadband Mobile Broadband Voice

Consumer Content and Applications

ServiceAssurance

Fixed Networks Wireless Networks

ServiceFulfilment

Billing and Customer

Care

Telecoms Software Strategies

Service Provider Strategies

Network Equipment Manufacturer

Strategies

Service Delivery Platform Strategies

Telecoms Software Forecasts

Telecoms Software Market Shares

Telecoms MarketMatrix

Core Forecasts

Country Reports

Research streams

For more information about our research programmes and custom research services, please visit www.analysysmason.com/research.

© Analysys Mason Limited 2009 Consulting from Analysys Mason

Consulting from Analysys Mason

Analysys Mason offers consulting services that span the entire business development cycle from strategy development, through planning and implementation, to review. We help clients in the public and private sectors to make the best possible business decisions on many key issues, including investment, strategy, policy, procurement, network rollout and market entry.

Analysys Mason advises clients in many industry sectors

Mobileoperators

Delivering value through strategy planning and

implementation

Mediacompanies

Helping tomaximise revenue in converging markets

Public sector organisationsProviding ICT and

procurement advice for emergency services

and public bodies

RegulatorsEstablishing and

implementing policy frameworks in telecoms

and media markets

Enterpriseusers

Helping to maximise returns from investment

in ICT

Financial institutions

Supporting vendors, sellers and financiers of

industry transactions

Fixedoperators

Defining new strategies, optimising service

portfolios and supporting policy

development

For more than 20 years, Analysys Mason has been working with a range of global clients to resolve business-critical issues by applying its intellectual rigour, passion and expertise.

For more information about our consulting services, please visit www.analysysmason.com/consulting.