Editor's note: Mr. Dermot Nolan has been an effective and forceful advocate of the COFDM modulation system, and, in particular, for the modulation choice made in Europe--the DVB-family.
Mr. Nolan has long presumed that the 8-VSB system developed the US will go down in defeat and leave the United States in a quandary which will take three-to-five years to remedy. That is how long it is thought to take if the FCC opens up the standard to further evaluation and change. No "fast track" (where a decision can be made in a year or less) is able to occur as long as competitive options seek admission to the FCC standard. Nolan takes aim at those competitive likelihoods below.
Many experts believe that should the standard open at the FCC the US effort in terrestrial DTV (and especially HDTV) will come to its end. HDTV will then be relegated to private usages in business and other specialty applications while the consumer is led to a standard definition set top box less for picture quality than for the ancillary uses of digital transmission. The "wireless" initiatives going on around the world seek spectrum, and the DTV spectrum in the US, and elsewhere, is prime real-estate.
Dale Cripps
______________________
TECHNICAL ANALYSIS
Both DVB-T and ISDB-T are MPEG2 /COFDM digital television systems. DVB-T was finalised in December 1997 and ISDB-T in April 1999. There are many similarities between the two systems:
1 Both use COFDM multiplexing and a variety of modulation standards such as QPSK, 16QAM, and 64QAM. ISDB-T also uses DQPSK, the DAB standard.
2 Both support MPEG2 transport streams in SDTV and HDTV modes
3 Both have robust reliable receivability on fixed outdoor, portable indoor or mobile outdoor antennas
4 Both support hierarchical modulation schemes
5 Both support Multi-Frequency Networks, Single-Frequency Networks, and on-channel domestic gap-fillers.
The main differences are that DVB-T is a general purpose DTV system with hundreds of operational modes, it does not use band segmentation, and the current implementations do not use time interleaving. DVB is either interoperable or easily configured to be interoperable with QAM cable, DSS/DVB-S, DMVDS, DSL-DTV. DVB also supports Dolby surround sound.
ISDB has a more restricted set of operational modes, does use band segmentation, and does employ time interleaving. ISDB is not currently interoperable with internationally deployed digital television platforms such as QAM cable, DSS/DVB-S, DMVDS, DSL-DTV. In essence ISDB is a derivative of DVB-T which can be viewed as a composite system embodying the functionality of DVB-T and DAB. ISDB does not currently support Dolby surround sound.
Time interleaving, as seen in the recent Brazilian tests of global DTV systems, can help in impulse noise performance. However there is a cost penalty for introducing time interleaving due to increased receiver complexity and processing delay, which adds to receiver costs in markets contestable with cable, satellite and other digital delivery systems. There are also likely to be cost implications for the distribution of signals to time-interleaved single frequency networks since the stations have to be synchronised both in frequency and time offsets. This may require services to be fed by fibre optic baseband distribution networks, generally not an issue in OECD countries depending on the pricing of PTO tariffs, but which may be an issue in other countries. Other solutions to impulse noise problems, which generally occur in weak signal strength areas, are now due to be released by a number of DVB-T silicon vendors later this year. These schemes trap the propagation of corrupted symbols, which have been affected by impulse noise before they can travel through the receiver and also use intelligent error concealment in the MPEG decoder. Solutions, which avoid the use of time interleaving in COFDM, may lead to lower long-run deployment costs for broadcasters, consumers, and network operators. Time interleaving is implemented by ISDB-T and is an optional extension to DVB-T for countries that may wish to consider it.
Band segmentation in which the RF channel is split up into discrete segments with different RF characteristics can be implemented. In ISDB-T around ten segments are devoted to video services with three to audio services. Introduction of band segmentation increases receiver hardware and software costs, as the receiver must be able to extract the different bands, which behave differently.
Band segmentation loses the concept of the DTV channel as a broadband pipe in regulatory regimes where service allocation is made at the band level. At the technical level band segmentation leads to loss of frequency diversity which means that if part of the band is lost, the entire signal can be lost. This is one of the reasons why the designers of DVB-T rejected band segmentation. There are other ways to achieving the same service split: eg hierarchical transmission, which is implemented by DVB-T and ISDB-T. Loss of frequency diversity is generally considered to be a bad idea and therefore band segmentation may prove to be unattractive for many broadcasters and consumers.
MARKET MATURITY OF DTV SYSTEMS
DVB-T is fully operational in the UK with around 750,000 consumer receivers installed and there are now around 500 DTV transmitters on air in the UK, eighteen months after launch. The DVB-T application is multichannel FTA/pay EDTV. Elsewhere DVB-T services have started in Sweden and Spain which, like the UK, are targeted at the fixed rooftop/portable indoor markets. Other European countries will be introducing DVB-T later this year with similar service portfolios
In Germany and Singapore DVB-T will be piloted as a portable/mobile DTV service portfolio later this year, with fully operational services in 2001.
Next year DVB-HD services will be launched in Australia and Singapore. In Australia DVB-HD may use hierarchical modulation which was demonstrated by DVB/Sinclair at NAB2000.
DVB hardware is sourced from multiple international CE and silicon vendors in an increasingly crowded field, including low cost Far Eastern manufacturers.
ISDB-T is currently a demonstrator system. There is one transmitter on-air in Tokyo and several hundred hand -built receivers exist. Present plans indicate that the deployment timescale for ISDB-T in Japan are between 2003 and 2005 although there are considerable uncertainties in this because of:
1 Costs to broadcasters of DTV conversion are estimated at $14 billion and the commercial broadcasters would like the Japanese Government to pay the conversion costs. 2 Commercial deployment of ISDB-T in Japan is likely to be a lot later than currently predicted. This is because frequency planning in Japan is exceptionally difficult, there are many thousands of analog transmitters in use, each main station has on average 80 repeaters, and MPT has indicated that analog stations and viewers will have to be converted in order to clear spectrum for digital use.
3 Commercial broadcasters and, to some extent NHK, prefer digital DBS as this offers a faster route to market, is being rolled out next year, does not have the frequency clearance problems, has time-to-market advantages, and conveniently preserves the existing analog broadcasting market structure for several years.
4 Japan must make a multibillion dollar investment in a fibre optic distribution network to distribute the services to the transmitters which will be configured as Single Frequency Networks because of the spectrum congestion problems in Japan. Estimates indicate that this fibre network will not be in place before 2004/2005.
5 There are no off-the-shelf silicon solutions for ISDB-T and no consumer STB's, as Japanese industry will have to agree the specifications and then build them.
6 In several respects the situation mirrors the early years of DVB-T. In 1996 it was predicted that DVB-T would be in widespread deployment throughout Europe by 2000. The UK launched in 1998, due to silicon availability, licensing and frequency coordination. Countries where frequency planning is more difficult (through either congestion or coordination constraints) are launching later eg France, Germany, Netherlands and Italy in 2001/2.
It is no accident that the three countries that launched DVB-T first were UK, Spain and Sweden, all on the western seaboard with fewer coordination and congestion problems. Therefore those promoting an early launch of ISDB-T seem to be neglecting the very severe congestion, conversion and coordination problems which the Japanese DTV industry will face.
In the next three-to five-years the only practical instantiation of a COFDM DTV system is DVB-T.
This has profound implications for countries that want to launch a DTV system in that timeframe or which may need to change their DTV system, for whatever reasons, in timeframes that remain competitive with cable and DBS. Furthermore current commercial conditions and frequency congestion indicate that the launch of ISDB-T in its core home territory is likely to be later rather than sooner. This has implications for availability of hardware, software and services in the ISDB-T DTV domain.
It is also highly likely that the DVB-T system will innovate, develop, and extend its reach during this timeframe. Current development efforts in interactive services, silicon solutions, and product availability tend to strongly support this hypothesis. Very large growing markets and high levels of vendor competition, innovation and rivalry buoy this effort.
At the present stage of market development it is now clear that DVB-T/COFDM is a proven, practical and pragmatic DTV solution. Development work began in 1993 with standardisation complete in 1997 and full commercialisation is only now being attained. In contrast ISDB-T was only standardised last year, receiver technologies remain at the demonstrator level, and the 'home' market may not roll out DTV service until 2005. This means that any country which adopts ISDB-T before it reaches full commercial maturity in Japan may be acting as a beta test site for a relatively immature DTV system and will not benefit from scale economies until later on. This is a strategic risk for that country, particularly if ISDB-T fails to achieve critical mass, in which case that country would be technologically and economically isolated from mainstream DTV developments.
DVB-T, which began development in Europe, has subsequently been extended by developers elsewhere for HDTV. It is now being extended to provide global mobile DTV services in collaboration with prestige marques such as BMW, Mercedes, VW, and its hierarchical HDTV/mobile DTV capabilities have created strong interest in the United States. DVB-T is likely to be considerably enhanced in the next three to five years and will leverage global economies of scale and scope.
ECONOMIC ANALYSIS
DVB-T benefits from the fact that it is part of a suite of interoperable and extensible digital television standards. The initial focus on SDTV developments, led by the commercially active cable and satellite sector, led to the formulation of the DVB-T standard, initially for E/SDTV. This was subsequently extended to HDTV and mobile DTV, which are all supported by the same COFDM multiplexing standard. Latest additions include DVB MHP, DVB-interactive and DSL-DTV. The scale economies from a fully finalised set of coherent and interoperable standards are significant. Indeed it has been reported that the ISDB-T developers wish to adopt the DVB MHP for their API environment.
DVB-T is a global standard: it is slated for use in 500 million TVHH worldwide prior to decisions being made in South America, China and Russia. This confers 50% global DTV market share at writing time, compared to 46 million TVHH, which are confirmed for ISDB-T. Its range of options means that it can be configured for individual national, regional and local conditions. Since its launch some two years ago DVB-T broadcasts have begun in thirty countries.
DVB-T silicon is commercially available for 6, 7, and 8Mhz markets, with the number of silicon vendors having jumped from two to fifteen in two years. Full specification DVB-T chips are available for under $10 and it is expected that this will fall to under $5 within the next eighteen months driven by market demand. ISDB-T is available for 6Mhz markets as a laboratory demonstrator and silicon solutions are some way off. In the UK alone there are now twenty five separate IDTVs, see www.ondigital.co.uk, from companies such as Philips, Sony, Toshiba, LG, Grundig, Matsui, and all global leaders in STB's provide solutions such as Philips, Pace, Nokia. PC DTV cards are about to be released by Hauppauge et al. By the end of 2000 it is estimated that 1,000,000 homes in the UK will be receiving DVB-T off-air from 500+ transmitters. At writing time around 750,000 STB's and 30,000 IDTVs are in use by UK consumers.
The vendor community for DVB-T is global. The vendor community for ISDB-T is national
The Nokia Mediascreen, a laptop portable/mobile DTV receiver, demonstrated to acclaim at NAB 2000 and Broadcast Asia 2000, is to be commercialised. Fuba and others are developing global mobile DTV receivers, based on DVB-T, in conjunction with luxury car manufacturers.
Very high commonality exists between DVB-S, DVB-C and DVB-T STBs at the SDTV level. HDTV STB's are governed by global market demand for HDTV services. SDTV markets are being driven by economies of scale flowing through from twenty-five million+ SDTV DVB STB's in use throughout the world for cable and satellite. This market is already twice the size of the standards fragmented US digital DBS market. At writing time ISDB is not compatible with other digital delivery systems.
Popular API's such as Mediahighway, Liberator, OpenTV, etc are ported onto DVB-T systems. DVB MHP will also be ported onto DVB-T systems.
DVB-SI works with multiple global languages, multiple platforms, and is now in global use. This facilitates navigation, control, cross-platform interoperability and content design. ISDB-T does not have these capabilities and may seek to adopt DVB MHP.
DVB-T now has fully deployed databroadcasting, teletext, interactive and email services with full DSL-enabled Internet access via second generation DVB-T STB's later this year. No facilities of this nature are supported by ISDB-T.
The extra technical features of ISDB-T, namely band segmentation and time interleaving, are complex and costly to implement. Alternatives to band segmentation exist and time interleaving, a DVB-T optional addition, may be bypassed by alternative technologies.
ISDB-T may be more costly than DVB-T to implement. This is because of the limited national vendor support, the additional technology costs, the higher IPR costs associated with ISDB compared to DVB (or ATSC), the smaller market share, the additional R+D costs required to fully commercialise ISDB-T, and the ten to one market share advantage which DVB-T already enjoys.
SUMMARY
Both DVB and ISDB are MPEG2/COFDM DTV systems.
They are at very different stages of market development and it is clear that DVB-T, the global DTV brand leader, enjoys significant commercial, consumer, economic and strategic advantages over its younger rival.
DVB-T has been fully commercialised, is proven, practical and pragmatic. DVB-T is also extensible and is now fully interactive. It is clear that ISDB is a technically led design enhancement, which combines the functionality of DAB and DVB-T in one platform mixing radio and television services together. This may or may not be desirable depending on individual national regulatory perspectives. These enhancements are either matched by DVB-T or may soon be bypassed by other technology developments, which the DVB camp is now perfecting.
Globally DVB-T now holds a ten to one market share advantage over ISDB-T, a three to five year window before which ISDB-T can be commercialised in its home market, benefits from scale factors that are not open to ISDB. DVB-T also benefits from the chaos in the US arising from the commercial, regulatory and technical failure of ATSC, which now rules ATSC out in other world markets.
In addition DVB-T can now leverage synergies between mobile DTV and UMTS, which positions it very strongly for the wireless digital video communications services in the next decades. Critically DVB-T also supports HDTV and Dolby surround sound, which makes it attractive to those broadcasters who can make a viable business case for HDTV. Even if HDTV fails as a viable digital terrestrial television business those broadcasters will be left with a DTV infrastructure which can easily be repurposed for other service portfolios, provided this is designed in from the outset.
For those countries adopting a DTV system and the trend is moving inexorably towards COFDM, there is only one commercially viable choice, endorsed by the global vendor community, DVB-T. It is also likely that the market 'window of opportunity' for ISDB-T will close outside Japan as countries make choices based on commercial, economic, risk, and strategic considerations.
It is interesting to speculate how different the global commercial prospects for ATSC would now be had COFDM been adopted in 1993 by ATSC.
Today the verdict on making the incorrect technology choice has been handed down by the ABER/SET DTV selectors in Brazil: 'The ATSC system does not fulfill the technical requirements for the continuity of the television broadcasting service'. There can be no more damning indictment of ATSC than this statement. The failure of ATSC to undertake the necessary broadcaster requirements analysis at the 1993 design stage, or alter course by changing technologies has left the door wide open for DVB-T to sweep the global DTV board because of the confluence of events and timescales in the development of the three DTV systems. Had ATSC adopted COFDM in 1993 or participated in subsequent COFDM developments it is likely that DTV market shares would now be more evenly matched. This was not to be and the price for ATSC is now commercial, market and technical failure at the global level.
DVB-T is now well positioned to springboard into the de facto role of global DTV standard, with ISDB-T used either solely in Japan or, at best, by Japan and one or two other countries.
In the US, the most likely scenario is that continued infighting over the choice of DTV standard will close the market window of opportunity for a commercially effective DTV service portfolio in the United States as consumers migrate to cable, DMVDS, DBS, DSL and Internet. Outside the US, with consumer enticing business models, DTV has very good long-term prospects of becoming a wireless 'must-have' service portfolio rather than a bit reduced cable 'must-carry' service.
Dermot Nolan
