| Wireless Internet Applications |  ... insights for success http://www.strategyanalytics.com |
|
A number of different technical broadcast/ multicast solutions are currently being considered and trialed by cellular operators, broadcasters, and existing TV network owners for delivering mobile TV content:
These solutions all necessitate the deployment of a dedicated network or satellite that will broadcast signals to devices that are fitted with the appropriate radio receiver.
As has been described in previous reports4, the principal rationale for rolling out dedicated broadcast networks is that the performance/ topology characteristics of mobile cellular networks are insufficient to handle mass market demand. Although new cellular networks are both more efficient and offer faster transmission than previous generations, they are not sufficiently optimized to be able to deliver high volumes of broadcast content to large numbers of mobile devices at prices that consumers will find attractive.
On the other hand, vendors such as Nokia, Alcatel, Qualcomm and Samsung have a vested interest in pushing technologies and standards for which they hold IPR: DVB-H, DVB-H+, FLO and DMB respectively. Consequently, each camp has heavily promoted the benefits of their own technology while highlighting the shortcomings of rival implementations, including MBMS.
There are 5 attributes that need to be considered before deciding which technology should be adopted for launching mobile TV, (Figure 7):
A major factor impacting the timing of mobile broadcast deployments is the availability of radio spectrum in the appropriate frequency band. Spectrum availability and capacity for digital TV varies widely by country, and is dependent on the extent to which the spectrum required is being used for other purposes, such as analog TV. We will now evaluate the major solutions on offer.
T-DMB services have been designed for DAB networks that operate in the VHF III and L-Bands. In countries where the required VHF and L-Band spectrum capacity exists, (DAB networks have been built out across 40 countries6), and where DAB networks cover a significant percentage of the population, DAB can be leveraged to provide a fast and cost effective route to deliver mobile TV. The DAB Forum has stated that advanced T-DMB trials are being conducted in 14 countries (China, France, Germany, Hong Kong, Italy, Mexico, the Netherlands, Norway, Peru, South Korea, Spain, Taiwan, Turkey, and the UK). Examples of early DAB based mobile TV deployments include:
However DAB faces the following major challenges to widescale deployment as a mobile TV delivery solution:
The technical characteristics of DVB-H were described in detail in our previous mobile broadcast report7. In summary, DVB-H is based on DVB-T, and has been adapted to support a number of features required for handheld reception, including time slicing for power saving and forward error correction for additional mobility. DVB-H can share spectrum and investment with DVB-T if hierarchical modulation or multiplexing is adopted.
We believe that DVB-H will become the dominant global mobile digital broadcast technology. The DVB-H format has been selected as a mobile TV standard by ETSI and has widespread support from operators, chip vendors, handset manufacturers.
Despite these advantages, DVB-H faces some major challenges that will impact its deployment:
In many countries UHF spectrum is currently occupied by analog TV services, and is unavailable for use for DVB-H until the process of analog switch off is initiated. In Italy, this was originally being timed for as early as end 2006, but has now been pushed back to 2008. In contrast, UHF based services cannot be launched nationwide in the UK before the analog signal is turned off in 2012. In the US, broadcasters have been mandated to achieve analog switch off before 17th February 2009.
Therefore, the use of the UHF band for DVB-H or FLO is largely contingent on when the government or local authority has planned for analog switch off and the regime it adopts to reallocate the spectrum. The Digital Video Broadcast group, who are proponents of the DVB-H format, claim that in Europe most administrations are planning for six to seven DVB-T multiplexes in the UHF band, while reserving one or two for DVB-H. Examples of two contrasting methods used by governments to allocate UHF spectrum are described below:
Figure 4 illustrates the timeline for UHF spectrum availability (nationwide) in a number of European countries. However, the switching off of the analog signal will happen on a regional basis within countries, with the larger more densely populated cities being left until the very end. For example, during April 2005 the local governments of Sardinia and Valle d'Aoste signed an agreement with the Italian Ministry of Communications to switch-off their analogue signals in early 2006 although full national switch off has been timed for 2008.
Opening remarks by EU Commissioner Reding at CeBit 2006 highlighted that achieving a more rapid path towards common spectrum allocation for mobile TV services across Europe, before 2012, will require a significantly greater effort on the part of European governments.
In summary, using the UHF band for mobile TV is complicated by the following uncertainties:
The timescale and spectrum allocation uncertainties have not deterred many European operators from participating in terrestrial DVB-H pilots (Figure 9). Furthermore, a number of carriers in Europe have already stated that they will adopt DVB-H for mobile TV:
O2 UK is in discussions with Ofcom (the UK regulator) to find spectrum for use with a DVB-H network before the end of the year, with use of the L-Band a possibility. However, this would mean compromising between coverage and quality to enable an early launch. Use of the L-Band for DVB-H would also require additional network investment compared to the more optimal UHF frequencies. According to figures from the Digital Video Broadcast group, DVB-H over L-Band would require an addition CAPEX of 160% on top of DVB-H over UHF.
Figure 9 Publicly Announced Mobile TV Service Trials and Pilots
In the US, Crown Castle/ Modeo acquired 5 MHz of unencumbered nationwide spectrum through the Federal Communications Commission auction 46 in 2003. We provide the following guidance on DVB-H deployment in the US.
Like Qualcomm, Crown Castle/ Modeo will offer wholesale access to the broadcast network to service providers. Like Qualcomm's MediaFLO (discussed below), it will also perform basic content aggregation functions to provide an end to end offering (Figure 10).
Crown Castle has stated that its deployment of a nationwide network covering 30 markets and a population of approximately 75 million will require funding of $500 million, equating to $6.67 per head of population. However, the network CAPEX for a DVB-H network varies significantly based on the modulation and spectrum that is used. Transmitting on higher frequency bands requires a greater number of transmitters, (because signals travel shorter distances at higher frequency), which ultimately leads to higher CAPEX. The modulation used also impacts total cost. For a given coverage area using 16QAM (16-state Quadrature Amplitude Modulation) instead of QPSK (Quadrature Phase Shift Keying) for a DVB-H network operating in the UHF band can add as much as 160% to the overall network cost. Using 16QAM modulation allows for greater capacity than using QPSK, (which equates to more channels or faster streaming speeds). However, 16QAM adds more cost because modulating the signal using 16 QAM instead of QPSK will require either more sites or higher transmitters, since QPSK offers superior coverage to 16QAM.
Figure 11 provides a cost of coverage comparison between DVB-H and T-DMB for an area consisting of 26,174 square kilometers, (190 sq km urban area, 3,908 sq km of suburban area and 22,076 sq km of area), conducted by the DVB organization. The comparison clearly indicates that DVB-H in the UHF band offers a more cost effective solution compared to DAB based networks, assuming that new networks need to be built for each.
System |
DVB-H |
DVB-H |
DVB-H |
DAB |
DAB |
Spectrum |
VHF 200 MHz |
UHF 600 MHz |
L-Band 1460 MHz |
VHF 200 MHz |
L-Band 1460 MHz |
Transmitters |
241 |
93 |
246 |
209 |
239 |
Repeaters |
492 |
198 |
588 |
420 |
572 |
Capex (EURM) |
49.3 |
22.4 |
58.2 |
42.7 |
56.7 |
Cost/Service (EURM)* |
3.8 |
1.5 |
5.3 |
10.7 |
14.2 |
Cost (EUR/bit/s) ** |
15.03 |
5.97 |
20.71 |
40.23 |
53.49 |
* Assuming 250 kbps/service
** Using QPSK Modulation for DVB-H.
S-DMB uses the Digital Media Broadcast format to deliver video streams, but uses a satellite with a network of repeaters and gap fillers to distribute the TV signal. SK Telecom's DMB service uses the 2.630 - 2.665 GHz range, providing a total bandwidth of 25 MHz. A basic representation of this network is summarized in the figure below.
There are currently two commercial mobile digital TV services in operation, both of which are based on the S-DMB standard:
The satellite band can also be used for DVB-H. Alcatel's DVB-H+ solution, which is a hybrid satellite and terrestrial DVB-H solution, aims to circumvent the low and fragmented availability of spectrum in the UHF band by proposing a solution that uses DVB-H over the unlicensed satellite UMTS band (2170-2200 MHz). Alcatel has filed for satellite orbital positions for European coverage. However, although spectrum is available, the process that will be used to allocate frequency is complex and will involve the ITU and national telecoms regulators.
Given that DVB-H+ is a Europe wide solution, Alcatel has partnered with other leading players in order to lobby the EC for harmonized spectrum allocation across the 25 EU member states. Alcatel claims that the big five network operators, (who we interpret to mean Vodafone, Telefonica, Orange, T-Mobile and TIM), have expressed an interest in DVB-H+ in addition to several broadcasters. Alcatel also claims that a number of handset vendors have expressed an interest in supporting the standard. In February 2006 Sagem announced its intention to offer DVB-H+ enabled handsets with the first prototypes available for field trial in the second half of 2006, in time for commercial product availability in 2007. Alcatel is expecting to initiate trials of DVB-H+ in 2006, with full scale service launch feasible by end 2007.
Despite this, we remain concerned by the potentially high upfront costs associated with satellite solutions. The DMB satellite launched by the Mobile Broadcasting Corporation (MBCo) cost $250 million and additional investment for gap fillers and repeaters are also required in order to overcome the inherent line of sight deficiencies associated with satellite platforms. In South Korea 4,800 terrestrial repeaters had been deployed by June 2005 at a cost of WON 120 billion ($120 million) to TU-Media.
Based on initial calculations SK Telecom expected to invest a total of WON 316 billion (~$327 million) in the project between 2003 and 2010, with profitability being achieved after 3 years on annual sales of WON 280 billion (~$290 million). By 2010 it was expecting annual sales to grow to WON 4,000 billion (~$4 billion) and to achieve a 30% internal rate of return.
In November 2004, Qualcomm announced plans to build its own nationwide, wireless multimedia network. Qualcomm stated then, that the network would start operating in 2006, broadcasting over the 6 MHz of 700 MHz TV spectrum purchased at auction by Qualcomm during 2003.
Qualcomm subsidiary, MediaFLO USA, will act as the commercial vehicle for services. MediaFLO will use OFDM (Orthagonal Frequency Division Multiplexing) technology based transmissions from high broadcast towers to cover a city with an average of only 2 - 3 transmitters. Qualcomm plans to invest $800m in the project during the next four to five years. MediaFLO is also aggregating some basic TV channels such as news and weather to offer on a wholesale basis to carriers in time for 2007 launch.
FLO offers higher performance specifications than DVB-H at a lower cost. While we are not able to independently verify the extent of its advantage, we understand the following statements to be true:
We are providing the following guidance with respect to the US deployment of FLO:
While Qualcomm retains ambitions to market FLO overseas, we are not confident that it will gain significant traction outside of the US. Despite superior performance attributes, we currently anticipate that it will achieve only a few deployments in countries in Asia Pacific and Central and Latin America (CALA) due to the following reasons:
Multimedia Broadcast Multicast Service (MBMS), which is specified in release 6 of 3GPP's roadmap, provides carriers with a solution for delivering point-to-multipoint services, such as mobile TV, over their cellular networks. 3GPP2's CDMA equivalent is termed BroadCast and MultiCast Service (BCMCS).
According to Ericsson, a proponent of the MBMS solution, between 7% - 30% of a cell's capacity can be reserved to multicast video content to an entire cell of users at speeds between 64 and 256 kbps. In contrast, a maximum of approximately 12 users at any one time can be supported by a unicast system with data rates of 64 kbps using a W-CDMA cell of 800 kbps capacity. In theory, providing MBMS on a W-CDMA network that has been enhanced by HSDPA will enable the carrier to provide 10 - 20 channels at 128 kbps or 20 - 40 channels at 64 kbps.
MBMS/ BCMCS over 3G+ networks (HSDPA/ EV-DO/ 3G LTE) theoretically provides the ability to support mass market demand for mobile TV services, without the requirement of deploying a new broadcast network or satellite. This would reduce risk to carriers, through the minimization of investment in new infrastructure (assuming that they choose this option rather than purchase of broadcast capacity from a third party wholesale supplier such as MediaFLO/ Modeo), and would remove various time to market concerns, including spectrum availability. On the other hand, the extent to which MBMS over 3G+ would provide a sufficiently high quality and robust platform for mass market TV provision remains in doubt and is an issue we will further investigate this year.
Nevertheless, given the potential advantages, we believe that this option should be thoroughly investigated by carriers. Some movements are taking place in this direction:
Whether MBMS and BCMCS are adequate technical solutions remains to be seen. Vendors such as Nokia and Alcatel have indicated that they believe that MBMS is not an optimal solution.
Although our view may be revised as further work into the performance of MBMS relative to dedicated broadcast networks is established, the lower costs and uncertainties associated with MBMS, combined with still uncertain demand, lead us to conclude that carriers should strongly focus on cellular broadcast solutions as a medium term option.
However, equipment vendors are understandably keen for carriers to move straight to dedicated broadcast solutions and bypass MBMS. Therefore, the downside to MBMS-over-cellular is a potentially limited supply of suitable devices from tier-1 vendors who control two-thirds of the global handset market. Nokia, Motorola and Samsung are focused on DVB-H and T-DMB respectively, but remain mute on MBMS. Although MBMS is likely to be mandated in Release 6 3GPP from 2007 / 2008 onward, lack of tier-1 vendor commitment will result in it penetrating less than 1% of total global handset sales in 2007 (hard-launch expected in H2 2007), rising steadily to 4% of total in 2010.
ISDB-T is the digital TV format adopted by Japan. With ISDB-T, a single TV channel occupies a bandwidth of 6MHz and comprises of 13 segments. One of these segments has been reserved specifically for mobile reception and hence mobile TV over ISDB-T is referred to as "One Seg."
Commitment from national and commercial broadcasters to begin broadcasting to mobile phones using ISDB-T in April 2006, combined with the availability of ISDB-T enabled handsets from Vodafone KK leads us to expect ISDB-T services to be available to mobile phones before the end of the first half of 2006.
Strategy Analytics also expects NTT DoCoMo and KDDI to sell devices that support ISDB-T, making it the dominant mobile TV format in Japan.
4 Mobile Broadcasting: Hype not Justified by Demand. Wireless Internet Applications, Feb 2005
5 In this report we use the term MBMS to encompass BCMCS as well
6 World DAB Organization - The Future of Digital Broadcasting (3GSM 2006 Presentation)
7 Mobile Content Broadcasting: Hype Not Justified by Demand!, Wireless Internet Applications, Feb 2005
8 UHF Channel 55, acquired at FCC auction in June 2003
9 On 23rd February 2006 the Finnish Government announced its decision to award the license to Digita, the owner of Finland's DVB-T multiplex.
10 Further details of which handset vendors are supporting which technologies can be found in the January 2006 report published by our Wireless Devices Service (WDS) `TV Phones: Integration and Power Improvements Needed to Reach 100 Million Sales.'
| Industry Report
Mobile Broadcast TV: Caution Needed as Bubble Grows! March 2006 |
© 2006 Strategy Analytics
All rights reserved. Neither this publication nor any part of it 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 permission of Strategy Analytics. |
