DVB Project's Technical Case for DVB-T Adoption in Mexico: COFDM vs. 8-VSB and Digital Transition Issues

Contents

Executive Summary 2

Glossary of DVB standard names 3

I — Technical Issues 4

II — Possible Application 12

III — Economic Issues 23

Appendix II-1 The UK Digital Terrestrial Platform 33

Appendix II-2 Philips DTX 6371 digital terrestrial receiver 39

Appendix II-3 Canal+ MediaHighway applications 41

Appendix II-4 DVB Multimedia Home Platform 46

Appendix II-5 Sophisticated solutions for optimal bandwidth efficiency in DVB-T Systems 54

Appendix II-6 Nokia Media Screen interactive portable DVB-T receiver and turbo Internet browser 62

Appendix III-7 System features and Added Value Services for Handicapped Viewers 63

Appendix III-8 European Service Model for Digital Terrestrial Television-Features and Dates of DTT services in Europe 64

Appendix III-9 STMicroelectronics Solutions for Digital HDTV 66

Appendix III-10 Comparative Reception Testing Of 8VSB and COFDM in Baltimore 72

Executive Summary

Considering digital terrestrial television, DTTB, DVB realised that the transmission systems should operate to existing receiver antenna installations, targeting the very important market of set-top-antennas particularly. The COFDM modulation scheme chosen provided the ideal, cost effective and rugged alternative to more primitive single carrier systems. Because it is a member of the DVB family, DVB-T operates with all the applications being used across the other DVB transmission media e.g. satellite and cable.

An added feature of the DVB-T transmission system is that the use of simple gap fillers means that coverage can be greatly increase for any single transmission frequency. If you choose your options carefully and synchronise your transmitter network, you can even operate a nation-wide service on a single frequency. The concept of a single frequency network is an extremely powerful and flexible tool to anyone providing digital terrestrial television service especially in built-up high margin areas like Mexico City. The flexibility of the DVB-T systems also means that it suits the 6MHz channel characteristics of Mexico. Significantly, the differences between 6MHz DVB-T equipment and 7MHz or 8MHz DVB-T equipment are extremely small.

The DVB response to the Mexican questionnaire also deals heavily with the issues of applications and the business models already employed on DVB-S, DVB-C and DVB-T networks to drive the transition from Analogue to Digital. Many broadcasters view new services, such a HDTV, as key. In the UK, the addition quality afforded by 16:9 extended definition services is been coupled to interactive services to offer competitive offerings to the satellite and cable systems (which run DVB as well).

This document also deals extensively with the issue of the convergence between computer, television and telecommunications. Multimedia Home Platform is perhaps the most important tool in the DVB arsenal on the road to convergence. Coupled with innovative applications from manufacturers and service providers, MHP promises a level playing field for all elements of the industry driving it towards convergence.

In short, the DVB provides all the solutions that the broadcasting community needs to ensure an economically viable transition to the digital era. With the hundred of manufacturers, producing thousands of products for millions of viewers around the world: DVB can help.

Glossary of DVB Standard names

Introduction

The following text is the DVB response to the questionnaire from the Mexican Secretary of Transport and Communications on the issues associated to the choice of Digital Terrestrial Television systems.

DVB responses follow each question, with the questions in Italics and the associated responses in normal text. In addition, we have decided to include a series of annexes complimenting the text.

A. Methodology

Standard promoters of ATSC and DVB are requested to analyse this document and summit responses to the issues and questions contained in it, with as much documented support as needed as well as with an executive summary. This documentation must also be presented orally by ATSC and DVB counting on the technical and professional support required for the occasion. Based on the conclusions that will arise from the analysis of the submitted information, a set of field tests might be defined in order to obtain the technical data that supports the particularities of the Mexican case.

2. Questionnaire

1.1 What is the standards behavior for the particular conditions of Mexico City?

DVB: DVB-T has been fundamentally designed to increase the coverage over existing analogue services at much less power. In particular, DVB-T has been designed to perform in the type of multipath environment encountered with the high rise buildings in Mexico City. Reception in both high rise and low-level dwellings will be possible using this system regardless of multipath conditions. This has been proven in the UK, which now has a countrywide DVB-T service providing up to 36 television channels to all types of dwellings. Perfect reception is now possible in buildings and environments, which previously could only receive pictures with extreme ghosting. Reception using only set top antennas is now common place.

Field trials in Singapore, Australia and Hong Kong have also shown this to be possible using DVB-T. It should also be noted that recent comparison field trials in the USA by Sinclair Broadcasting Group have proven the superior performance of DVB-T over 8-VSB in these circumstances.

1.2 What results could be obtained for this particular case regarding coverage of the signals?

DVB: Even in conditions of extreme multipath it is expected that the existing analogue coverage can be replicated with DVB-T using average transmitted ERP levels more than 15dB lower than those of existing peak of sync analogue transmissions. It is also likely that this power can be further reduced.

In the UK DVB-T, power levels 20dB lower than those of the analogue service are used almost replicating the existing analogue service coverage.

Recent developments in DVB-T chip sets have further improved performance by exhibiting a further 2dB decrease in receiver C/N threshold.

1.3 Which are the propagation characteristics of the signals based on experimental results, and how do these differed from the predictions obtained from the theoretical (Longley-Rice) methodology? If possible, which weight factors have been obtained for a better application of the Longley-Rice methodology.

DVB: Considering propagation characteristics of COFDM signals, the Longley/Rice method is fully applicable for the coverage prediction for such cities as Mexico City. Simpler knife edge methods or Okumura-Hata for the higher UHF frequencies or even the old ITU-R 370 model, but with the enhancement of the "terrain clearance angle", might lead to useful results.

The question is very specific (adapted weight factors for Longley/Rice method for a given area: Mexico City). Our experts, in order to produce significant figures need the specific data of the City itself. Unfortunately, in the given time, they could not find the measurement data of Mexico. One of our members, L&S Hochfrequenztechnik from Lichtenau, Germany, is presently trying, through its Radio Communications department, to find such data.

If, and as soon as, they find these data, we will further inform you about the output.

2. It has been stated that 8 VSB has some problems regarding multipath.

2.1 What are the specific facts that support the Sinclair statements? Provide comments on this procedures and results.

DVB: It has been stated many times in the past and again recently by ATSC that the 8-VSB system was never designed for anything other than fixed reception. The prime aim of the system was to deliver HDTV signals to the home environment. This may be the reason why the transmission method used was based on existing single carrier techniques. These techniques maximised data rate transmission at the expense of ruggedness of reception. ATSC maintains that it is up to the receiver manufacturers to solve the signal impairment problems encountered in the transmission path.

Recent statements from the ATSC have shown that they do not disagree with the Sinclair findings but cite the fact that the problem with reception was due to the use of early receivers. This has led to many hyped press statements that a new range of miracle chips will be available soon. Even if this is the case and these chips do improve performance the projected cost of the silicon quoted by the manufacturers will make the receivers more expensive and will also completely eradicate the myth that ATSC receivers will be cheaper than those receiving COFDM.

Conversely, the design of the DVB-T system was based around providing a transmission system that would be extremely rugged to transmission impairments whilst maintaining a high bit rate. The DVB-T approach has always been to provide a system that can deliver data faultlessly to the receiver regardless of whether that data is HDTV, SDTV, audio or other data services. The flexibility of the DVB-T system has been shown to allow the faultless reception of data at speeds of 300km/h albeit at reduced bit rate. Critically, broadcasters can choose the broadcast parameters to suit their service and reception scenarios, but the receiver is a single flexible DVB-T receiver automatically tracking the changes in transmission parameters.

With DVB-T, it is not necessary to provide special receiver equalizers to improve multipath performance. In fact the DVB-T transmission system provides enough transmitted data for the original signal to be reconstructed very quickly providing extremely fast dynamic multipath response and large Doppler shift performance.

2.2 What are the specific results on this matter obtained for the New York City case (Manhattan)?

DVB: While DVB does not have access to specific results from trials carried out in New York, the city presents a number of key issues for a television broadcast system. The high rise buildings present significant reflections to a receiver — a situation in which DVB-T thrives. Additionally, New York is interesting because it is in these high rise buildings that the high margin consumer lives and works — the consumer who will be prepared to pay for the exciting new services digital television technology will bring.

2.3 There have been statements from Motorola and NxtWave, establishing that VLSIs have been developed for the elimination of the ATSC multipath issue, however it has also been stated that production will take some time. How do these chips solve the problem? What is the feasibility of counting with prototype receivers that could incorporate these adjustments? Are these chips oriented towards dynamic or static multipath conditions?

DVB: At the time of writing no test results for either chip have been published. The initial hyped performance data was based on software simulation and used neither real silicon nor exhaustive modelling.

In the simulation data recorded during the Sinclair trial was used, however to our knowledge this did not provide total real world data and certainly did not provide data to prove dynamic correction.

Both chips have significant increases in complexity and silicon area to provide the large equalizers needed to correct for static multipath. This results in significantly higher costs and it remains to be seen whether the final chips will meet the simulated performance let alone provide the dynamic correction required in real world situations and how much the C/N performance will be degraded under these conditions.

Such dynamic multipath situations are those experienced when receivers are operated with indoor antennas or situated near roads or airports where fast dynamic changes occur.

Because of DVB-Ts proven capability in the mobile environment, which is more severe, the problems have never been a limitation.

3. DVB and ATSC use different modulation schemes

3.1 What is the useful bitrate capacity that could be used for the different foreseen applications at 6 MHz, and how is this handled or assigned in terms of transmission components? What is the bit error rate obtained on each case and how does this affects reliability?

DVB: The following table is drawn from the DVB-T standard: EN 300 744 Version 1.2.1:

Thus, the ATSC data rate of 19.3Mbit/s would be approximately similar to 18.7Mbit/s in DVB-T. Sinclair in Baltimore tested this same scenario and DVB-T reception was found to be vastly superior to ATSC.

Again, it is worth pointing out that the choice of transmission parameter is up to the broadcast for his/her specific scenario and the DVB-T receivers, being capable of all modes, can track changes.

4. It has been stated that ATSC has a better response than DVB regarding impulse noise and that it is also more efficient in terms of power to obtain the same areas of coverage.

DVB: What are the technical elements that support this statement?

It is true that this is often stated but as with most ATSC information this is largely based on laboratory experiments and not real world situations.

A main requirement in the USA is that the low VHF spectrum should also be used for digital transmission. In general, the problems with man made interference are much worse at these frequencies than they are at UHF.

It is also very important to define the type of impulse noise and the length of impulses before definitive measurements can be made. Real world experience is already available for DVB-T but we have not seen any for ATSC. In fact, there are those that suggest that ATSC receiver design is such that the problem is more acute for ATSC than it is for DVB-T. The recent demonstration of TV reception on seven trams at the IBC exhibition clearly showed that in this situation impulse noise was not a problem for DVB-T. We would consider that reception of DVB-T in trams to be most severe impulse noise environment possible.

Where some impulse noise problems have been experienced in the UK these have normally been cured by repairing faulty antenna/feeder installations or by changing the Set Top Boxes to later versions. Some early set top boxes were designed without the automatic gain control circuits, included in most receivers, with the consequence that small noise spikes would often saturate the receiver causing the failure of the receiver to operate correctly. Recent boxes no longer have this problem.

5. For the implementation of DTV in Mexico, the second channel assignment might be done using the channels 20 to 69.

5.1 What are the reception requirements in terms of antennas, cables, etc.?

DVB: For the reception of DVB-T, there are no special requirements for antenna installations. The existing system used for analogue can also be used for DVB-T.

This major advantage of DVB-T over ATSC allows for the purchase and installation of Set Top Boxes without the need of specialist advice. This allows the sale of units in existing TV/Audio shops.

5.2 What are the threshold levels for the signal to be detected?

DVB: The threshold levels for reception are detailed in the following table taken from EN 300 744 Version 1.2.1:

DVB: Yes the excellent multipath performance of DVB-T provides perfect reception in areas where only poor quality analogue pictures were available before

This has been proven during the installation of over 400,000 STB, s in the UK.

Let us not forget that one of the key criteria behind the development of digital terrestrial television was to increase (not just match) the coverage over analogue television.

6. In Mexico, the licensing scheme for open television does not include the possibility of having a license to cover the whole country, actually licenses are limited for a coverage area. However, due to the orographic characteristics of the country, an important element to achieve good coverage is the use of repeaters (high and low power transmitters) with the constraint for assignment of having an efficient use of the spectrum.

6.1 What is the digital signal behavior for repeaters and how feasible is it to use cochannels for this purpose? What is the maximum level of the primary carrier without having reception problems on the retransmitted service (repeater)?

DVB: DVB-T has the property of working well with simple on channel repeaters. This is the principle of Single Frequency Networks. Such SFNs can operate at a very local level with a simple amplification and re-transmission, or can operate at regional of even national levels when the transmission sights are co-ordinated.

This situation is an extremely practical arrangement. Take for example the case where you receive a DVB-T signal on a roof-top antenna in an appartment block. Rather than having to re-distribute this signal through the cable network, you can simple insert a bandpass filter and amplifier in the antenna and re-transmit the signal to set-top antennas in the building.

Concerning the set-up for this arrangements, the most important feature of the installation is not necessarily the relative power of the two "transmitters" (main transmitter and repeater), rather the guard interval used the DVB-T mode being transmitted. Essentially you set up the guard interval such that this provides protection in your population of receivers from echos from either of the transmission sites.

6.2 Is there feasibility for the use synchronic low power repeaters that could be used for gap filling purposes (reinforcing the signal at specific parts of the coverage area)

DVB: The design of the DVB-T system allows for on frequency repeaters by design. This feature allows for the use of single frequency networks SFN allowing maximum spectral efficiency.

Where the transmitters region to be covered is over 50km from the main transmitter additional transmitters can be added to extend the coverage range of the transmission. In a single frequency network the same transmit frequency can be used providing that the transmitters are synchronised together. This is simply done by using GPS derived frequency sources at each transmitter.

Within the 50km radius from the transmitter, simple repeaters can also be used to provide additional coverage. These repeaters simply comprise of an antenna pointed at the transmitter, a bandpass filter and amplifiers to provide more signal. It would also be advisable to provide a bandpass filter before the signal is fed to the transmit antenna. The only requirement for this type of repeater is that the isolation between the receive and transmit antennas is greater than the gain of the system.

6.3 How easy the local programming insertion could be done under conditions of satellite distribution of signals as well as for repeaters?

DVB: The DVB set of specifications is designed as a coherent group of transmission standards with associated Service Information (EN 300 468). While the physical layer of the transmission standards differs, the software layer is common amongst all DVB broadcasting media. The net result is that local insertion of programs into a DVB-T broadcast, either from DVB-S satellite or local generation is extremely easy: that is the way the system was designed.

Furthermore, it is important to note that this same Service Information layer is that being used in countless DVB-S services around the world, and indeed in Mexico. Therefore, manufacturers and implementers have had a number of years of experience in how insertion is done reliably. DVB is the only system to offer such features.

The diagram below indicates possible broadcast and reception scenarios for locally inserted DVB programming. The DVB multiplexor generates the Service Information to facilitate the locking of the receivers to a particular service. In addition, the SI can be used to construct a programme guide.

7.1 What are the differences (if any) between the MPEG-2 used by DVB and the one used by ATSC?

7.2 What is the level of compatibility between the two systems regarding audio and video compression schemes?

DVB Both ATSC and DVB are MPEG-2 based digital television technologies. The main differences between the two lies in their choice of transmission mechanism, the Service Information layer (to use the DVB term) and in their choice of recommended audio technology. It is important to note therefore, that the differences in MPEG terms are far less than the similarities between the two systems.

Transport Layer: Both ATSC and DVB use MPEG transport stream as the main transport mechanism. This is defined in ISO/IEC 13818-1 and provides an extremely flexible and robust mechanism for delivering MPEG video, audio and data services. In addition to the transport stream, both ATSC and DVB have specified a Service Information layer, which in DVB’s case, is common across all the transmission media, e.g. satellite, cable, terrestrial and microwave.

Video: Both ATSC and DVB use MPEG-2 video as defined in ISO/IEC 13818-2. DVB is often much criticised for not placing more emphasis on HDTV, the so-called Main Profile @ High Level MPEG-2 video option. DVB is very much a market based set of standards and can, and will, be used in whichever way broadcasters feel can best reap the benefits and revenues associated with digital television services. Thus in Europe, DVB is being used for Standard Definition Television, as specified in the Main Profile @ Main Level of MPEG-2 video, while in Australia, DVB-T will be used for HDTV applications similar to the US environment. Critically, the chipsets involved are the same and thus we are seeing a single truly world-wide standard for digital video, namely MPEG-2.

Audio: There are three digital audio compression systems offered for use in conjunction with MPEG-2 video for digital television applications. They are all surround sound capable and are: Dolby AC-3, MPEG Layer II and MPEG Advanced Audio Coding (AAC). The ATSC exclusively recommends the use of Dolby AC-3 audio, while the majority of DVB applications currently use MPEG Layer II stereo. Indeed the DVB recommends that MPEG Layer II audio offers the best technical solution particularly in the light of legacy stereo applications similar to the millions of DVB-S decoders in the market. In order to cope with demands of countries wishing to incorporate Dolby AC-3 audio in their digital TV transmissions by satellite, cable and terrestrial, DVB has devised mechanisms to incorporate Dolby AC-3 into its streams thus offering the broadcasting community the ultimate choice.

8. ATSC and DVB will be used instead of analog standards

8.1 What is the feasible compatibility level between the digital terrestrial television standards used for open television and the ones that are, or could be, used for the restrictive television technologies (DTH, MMDS, Cable)? What are the cost implications for each case?

DVB: As you are probably aware, DVB is the most popular standard for digital satellite, cable and MMDS applications around the world. DVB’s success is based on a set of coherent technically advanced standards for each of the media concerned. Thus DVB adopters benefit from the reduced costs associated with the enormous economies of scale associate with millions of receivers being deployed in the market. In addition, experience that has been gained from the deployment of DVB-S systems since the end of 1994 can be used for the deployments of cable and terrestrial systems.

Transitions between the different media with DVB are also facilitated since they all use the same DVB Service Information layer.

9. Australia and Singapore and actually Brazil, are countries that have established testing schemes for the evaluation of the standards.

9.1 What have been the test-rigs used for each one the cases (instrumentation, methodology)?

DVB: The enclosed CD-ROM contains all the details concerning the Australian test set-ups. Australia was the first country to fully and independently test both ATSC and DVB-T systems and all agree, from the engineering community, right up to the chief executives that they have made the right choice in going with DVB-T.

9.2 Base on the achieved experience, are there some comments regarding these test rigs?

DVB: A significant number of both field and laboratory trials have been carried out in DVB-T. If you add to this, the number of independent trials carried out in Australia, Singapore and Hong Kong, there is now a lot of information available.

DVB has been involved directly in a number of these trials and thoroughly recommends extensive study of the testing mechanisms proposed. This is especially true of the situation in Australia. However, as time goes on, our experience in testing digital terrestrial television systems increases, and Mexico stands to gain a lot by adding its valuable know-how to the work of the other countries who have carried out extensive independent comparative tests into ATSC and DVB-T systems.

9.3 Is there data obtained from these tests that could be extrapolated for the Mexican case?

DVB: While Mexico has a specific set of requirements and will doubtless develop an implementation scenario and set of services to suit its own case, physics is the same everywhere and therefore, the extensive preparatory work that went into the testing in Australia and Brazil can be useful. In particular, issues of interference, overall coverage and noise immunity are relevant the world over and therefore, DVB feels that Mexico can learn a lot from the experiences of other who have carried out tests in digital terrestrial systems.

9.4 What is the feasibility of providing equipment for measurement, tests and demonstrations? If so, when could these be done?

DVB: The DVB Project Office is the central co-ordinating body for the DVB members. Indeed, the Project Office has been involved in most of the official comparisons between ATSC and DVB-T. The Office acts in a co-ordinating role, bringing together industry partners, suppliers of DVB equipment, and co-ordinating their involvement in tests. Once a decision is made to carry out official testing into the comparisons between the various systems, the DVB Project Office can provide help in obtaining equipment from the DVB members and help in testing parameters and methodologies wherever appropriate and applicable.

The DVB’s philosophy in the technical evaluation of its systems has been to let the technology do the talking. In every instance of technical evaluation of the different digital terrestrial systems on offer, the DVB-T system has been proven to be the most advanced and coherent system available.

Testing of television systems is a time consuming and complex process demanding significant commitment from all sides, DVB will ensure that it will do whatever it can to help.

10.1 What are the current and future possibilities of this?

DVB: DVB made the important choice of offering a multi-layer structure, where all layers are independent of the others, and as many of these are common across different delivery media as possible:

Service operation and programme management is also now dealt with using sophisticated computer systems. They deal with:

• traditional broadcasting (see ETR 154 (DVB-MPEG) ) like sequencing, insertion, etc.

• subscriber management,

• 7 free to air (FTA) channels and an education channel by BBC, based on a yearly licence fee of US$160.

Although generally including simple navigators (compared to future generations), first generation receivers display, for the viewer, basic programme details of the present and following programmes on all the terrestrial multiplexes. This uses information transmitted in SI and is exchanged between multiplexes and cross-carried such that viewers have information on programmes on all terrestrial multiplexes whichever multiplex they are tuned to. The system has the capability to be extended to carry a longer schedule but this has not yet been implemented.

The multiplex operators believe that programmes sell subscriptions, and the implementation of data services is regarded as a somewhat lower priority. Both the DVB-S satellite and DVB-T terrestrial platforms launched in 1998 have grown rapidly without a common API.

As an interim solution to the future Multimedia Home Platform common API of DVB and due to the urgent market request, data services on DTTB have recently commenced in the UK using ISO standard MHEG5. All the operators are currently in process of rolling out "enhanced" information services. (Enhanced implies viewer interactivity by selection only. The use of return path is not yet implemented but will be added by download later, knowing that return path hardware and middleware are already implemented in first generation IRD's as shown in Appendix II-1).

Information services include a full programme guide. Teletext Ltd., which has a licence for VBI (vertical blanking interval) text services on ITV and Channel 4 analogue, has rights to 700kb/sec on the equivalent digital multiplex. Teletext plans to operate an information service which includes a full on-line programme guide, news and commercial features. One should note that there is extremely high penetration of teletext services in the UK analogue TV market.

True interactivity has recently launched on the Sky digital DVB-S satellite platform and there will be great interest to discover how important this feature is to growing the market. Terrestrial channels will consider this in planning the implementation of interactivity in their present and future programmes. For instance, BBC has already tested and announced interactivity on its News, Weather Information, Sport and BBC Knowledge programmes.

The current "free" set-top box market means that the box is built down to a price and that the opportunities for really advanced interactive services are limited by the lack of processing power, memory and hard disc. The facilities that are provided on a PC fundamentally require PC resources and cost PC prices. The market for interactivity on a TV or set-top box is a different market — the software has to be very simple and very robust to appeal to users who do not already have a PC.

This is precisely the aim of the Multimedia Home Platform being presently defined in the DVB-MHP standard to offer a really consumer terminal at reasonable price and very high user-friendliness. The Java programming aims at allowing download upgrades. See later paragraphs on the subject.

Transport and distribution networks are at the very heart of convergence. From content creation site (be it fix or mobile), to the programme provider and from there to the broadcasting centres, DVB offers all types of "convergent" transport/contribution media: on telecom networks (DVB-PDH, DVB-SDH, DVB-ATM, through fibre optics, coaxial cables, microwave links) and via satellite links (DVB-DSNG, DVB-S). Distribution networks are also highly relevant in the convergence process, being all computer-controlled as networks and as transport-stream distributing systems. Whichever the transmission medium (DVB-S, DVB-T, DVB-T, DVB-MS, DVB-MC, DVB-CS) is used, they include common concepts allowing interoperability and a common service management through the use of the Service Information System (DVB-SI) down to the user. Distribution of interactive services means also return channels. These channel are also fully defined in DVB through DVB-RCC, DVB-RCL, DVB-RCDECT, DVB-RCGSM, DVB-RCCS and fully interfaced with the service tools and the user's receivers.

Receivers are another extreme point of convergence. DVB receivers are rapidly moving from simple tuner/receiver/decoder interfaced with, or integrating a display, towards a complex piece of equipment that performs many more functions than a PC in a completely transparent manner for the user.

Presently available receivers, although far from including the high degree of flexibility of the future Multimedia Home platform, offer already a large number of functions related to interactivity and allow numerous "convergence related" services. See, for instance, the technical specifications of the Philips set-top-boxes for the UK market (free to air by BBC, Digital 3 & 4,SDN and pay-TV by ONDigital) .

10.2 What are the short and medium term possibilities of interactive television?

DVB: The interactive services already offered in Europe and other countries of the world on satellite and cable are planned to be rapidly transferred to DTTB, as soon as DTTB starts in the given countries. In addition to the already very attractive interactive TV services offered in the UK today (see § 10.1 above), the services delivered by DVB-S today by Canal+ are a good illustration of the short and medium trends of DTTB.

They include:

• Electronic Program Guides

Similar information can be found on the web site of French TPS (Satellite Pay TV subsidiary of the free to air TF1 broadcaster): http://www.tps.fr/Demain/Interactivite/Intro.htm, or on the BSkyB interactive services site called Open: http://www.bib.co.uk/

Note that these service and platform implementations are based on DVB standards with, in addition some proprietary solutions for some specific issues, like their API, the specifications of which are not yet finalised within DVB.

This is precisely the task DVB is presently finalising for the next generation of receivers. This new generation is called the Multimedia Home Platform. Georg Lütteke (Philips, EACEM) describes its basic function requirements . Although these receivers are not yet on the market, they have already been presented at IFA, Berlin August 1999 with operational Internet services by the local TV station. They were also on show at IBC, Amsterdam, and September 1999 and at Telecom'99, Geneva in October 1999.

The industry is planning to offer them on the market soon after the final DVB-MHP specification is released (expected end 1999).

The attached summary of a presentation of the MHP by some industrial DVB members aims at showing its capabilities and services possibilities:

• slide 5 shows how 4 MHP prototypes from different manufacturers could interact with 6 different services from various broadcasters;

Most implementations of commercial DVB-MHP platforms will probably include a hard disk and interface with equipment for data processing (PC's) and storage (CD-R, DVB-R) as well as with future home-networks. Significantly, DVB has already specified in its set of standards, the protocols and interfaces required for this convergent connectivity.

The high bit-rates of DVB-T data broadcasting open the way, with such receivers, to plan data services to private, to professionals and to business users as well.

10.3 What are the technological trends of the systems in terms of:

10.3.1 Number and kind of services at 6 MHz, including video, audio, data, interactivity, information, etc.

DVB: Firstly, we should noted, as explained in § 10.1, that DVB has dissociated the service and transmission layers completely. This means that all services from wherever in the world can be distributed around the world using the DVB transmission standards without any post-processing. They can be received on any DVB IRD, whether it operates with 6, 7 or 8 MHz bandwidth, provided the appropriate the service conditions chosen by the broadcaster fit the available bitrate.

As explained in § 3.1, the available bit-rate in DVB ranges from 3.7 to 23.8 Mbit/s for 6MHz (4.9 to 31.7 Mbit/s for 8 MHz) while ATSC offers only a fixed bit-rate of 19.3 Mbit/s. In practical terms, the UK uses today parameters that allow 24.13 Mbit/s which would give 18.1 Mbit/s in a 6 MHz channel. Interestingly, the Sinclair tests in Baltimore (see Appendix III-10) used a DVB-T mode giving 18.66 Mbit/s in a 6MHz channel.

To this, we should highlight the fact that, within a given multiplex, DVB offers the possibility for adaptive statistical multiplexing. This system performs a dynamic share of the bit-rate capacity between the different services and allows an extra practical capacity of up to 25%, as implemented today in satellite transmission (see Appendix II-5 "Sophisticated solutions for optimal bandwidth efficiency in DVB-T Systems" by Roel Janssen . For a 5-program multiplex, as that described here in Fig II-1.1 traditional model, the increase would be about 17%, allowing the addition of 3.74Mbit/s for an extra SDTV channel or a Data Service for instance.

If the bit-rate needed for a give service depends on the information density of the content (simple or complex scenes, static or dynamic images…) it also depends on the degree of compression of the signal. This element remains a decision of the content provider. For instance, now in the US, digital satellite operators who have no image quality competitors use excessive compression schemes that make DTV from satellite in the US much less appealing than it could be. On the opposite European broadcasters, like the BBC, chose optimal compression for high image quality. The result is that a BBC EDTV images (16*9, 720 pixel/line*576 lines) look as nice as ATSC HDTV programmes broadcast in the US and much better than US digital satellite.

We can also consider that, in the coming years, compression techniques will further improve (see work in the MPEG world) and allow more programmes for the same bit-rate.

ii) Multiple service broadcasting

DVB: Another important issue is the interoperability between DVB satellite, cable and terrestrial delivery systems. The independence of services from the transmission platform is the major reason why, DVB-S service have been operation all around the world to millions of receivers since 1994. Service providers are now introducing value-added services like pay-TV and iTV which will be identical across the different DVB transmission platforms. The broadcasting scene is moving to multiple platform broadcasting (satellite, cable and terrestrial) with similar and sometimes identical services. A practical example of this can be found in the UK: digital terrestrial receivers are all fitted with the DVB Common Interface (DVB-CI) and can easily be fitted with DVB-CI based "sidecars" which convert a standard DVB-T receiver into a DVB-S/DVB-T receiver.

The DVB technology, with its full interoperability between the different platforms, allows and, in a way, favours the move of the broadcast business from a purely vertical business (from content creation to distribution to the home) into a more horizontal business (when requested by the broadcasters). In this case, the broadcast industry is focussing on offering content-based services, leaving the high investment business of transport/contribution and distribution platforms to specialised companies. An example of this is the UK situation where BBC sold its DTTV infrastructure to NTL. That runs it now for all DTTV broadcasters.

This makes easier the share of a given multiplex by different services provided eventually by different broadcasters.

Section 15 shows the economic interest of such an approach to the DTTB business.

The number of services per multiplex depends significantly on the types of services and the quality level decided by the broadcasters themselves. Figure 2 shows three different types of approaches corresponding to different broadcasting business models.

The second model is the HDTV oriented model allowing viewers non-equipped with expensive HD receivers to have also access to the programmes.

The third model is the one that seems to develop rapidly in the world. It is based on the present trend in the satellite and cable worlds with good commercial success: TV programs (including interactive ones), data broadcast (for professional use for instance) and radio programs. DVB-T, with its COFDM modulation technique offers, in addition, a type of service that the two other platforms are not at all able to offer: programs to mobile viewers. Remark that such service to mobiles can be TV oriented (football matches, advertising to commuters…) or interactive data oriented (traffic information, stock exchange, e-mail, etc.)

These schemes are only examples. It is clear that the world of broadcasting will change significantly in the future. Such open standards as DVB will allow broadcasters themselves to re-engineer their own business with such tools.

A simple example will illustrate this: DVB is now preparing the requirements for a return-path through the DTTB antenna: this may simply allow full autonomy of the broadcasters for interactive services and avoid depending on telephone operators -potential TV competitors to broadcasters with such technologies as xDSL- for interactivity services.

iii) Added Value services

DVB: With the ability of DVB-T to transmit to portable receivers (with omnidirectional and set-top antenna) numerous new types of services are open for creation: "TV anywhere anytime" becomes a reality. Specific services to the youth (video-clip, sport results and comments), educational programs (with interactive facilities including audio and video-telephony with educators), business to business for mobile professionals (company LANs access) and many more services in addition to those already launched will develop.

10.3.2 Special devices for services different to television.

DVB: The ability of DVB-T to transmit data as well as TV on 6, 7 or 8 MHz channels with bit rates up to 24 Mb/s in a 6MHz channel in practical conditions has opened the way to new services as mentioned in previous paragraphs.

Receiver equipment manufacturers have already started the conception of fully new types of receivers for services related to this type of new services.

The most impressive one for its innovation and convergence oriented design is the Media Screen prototype demonstrated at IFA consumer show in Berlin, August 1999 with operational Internet services by the local broadcaster. It was also on show at IBC, Amsterdam in September 1999 and at Telecom'99, Geneva in October 1999. A photograph of this device is show in Appendix II-5 . It consists in a portable, battery powered, DVB-T receiver of the size of a lab-top computer. It includes a Linux based Internet browser and e-mail facilities. A small 10-cm antenna allows both DVB-T reception and GSM return-path for any interactive services on the move. The LCD colour screen displays TV programmes as well as any Internet content. Significantly the upstream element of the Media Screen uses digital mobile GSM telephony, while the downstream elements, including the digital TV and turbo Internet pages, come via the DVB-T receiver.

This device alone, which is one DVB member’s view of how the world of broadcasting, telecommunications and computing can converge, is one of the most significant product announcements in the field for some time.

Industrials in the world are now considering other new types of interactive services based on the perfect reception of high speed internet capability of DVB-T in its local, regional and national area of distribution. Multimedia information kiosks, very large animated advertising screens, etc. are in project and many more types of equipment might be expected in the coming future in relation with these unique reception features of DVB-T.

10.3.2 Handicapped features and content awareness

DVB receivers already available offer facilities for elderly and disabled persons.

Visually impaired: For teletext and data services, possibility to adapt the displayed fonts to the quality of viewing of the user can also be implemented in the receivers.

Physically handicapped: the implementation of multi-vendor remote control units allow specialised niche market manufactures to create special units adapted to various handicaps, as already done today for analogue TV.

Ethnic & linguistic minorities: the possibility for numerous sound channels is one of the most exciting improvements brought by digital television. DVB, based on the MPEG2 audio standard, supports the option to embed up to 7 language channels in a multi-channel audio stream. Each of these additional language channels requests 64 kbit/s in addition to the 5.1 surround sound channel (see " DVB Audio" by Leon van de Kerkhof .

For a detailed description of the facilities already offered in the UK, see " System features and Added Value Services for Handicapped Viewers", by Peter Marshall (DTG) Appendix II-6

10.3.4 Links to emergency services.

Taking into account the fact that DVB-T includes all facilities for interactive services, including soon a direct-through-the-antenna terrestrial return path, the creation of such service should be one of these types of new services that we can expect the broadcasting industry to create and implement at short or medium notice.

The first needed element is the portable/wearable alarm knob. As such devices already exist with connection to the telephone set, they should be easily adaptable to interface the MHP receiver for instance.

The type of service that a broadcaster might offer on this base can be very attractive. It can make use of Internet telephony for direct contact (including videophony) for direct human communication with emergency experts. It can furthermore use the high bandwidth video and datacasting functions to provide the user with multimedia information and demonstrations of what to do, or not to do.

10.3.5 Reception on personal computers?

As already described in § I0.3.1 the Multimedia Home Platform is going to be at short notice the interface to the home PC and its peripherals.

Nothing prevents PC users either to install DVB-T PC cards similar to those already on the market for DVB-S and use this for professional activities or just to watch TV on a PC. Knowing the increasing proportion of one-person homes in developed countries, watching TV on a PC, although less appealing, might develop significantly in some countries.

Furthermore, in a world where more and more young people use PC's for studying and working (often lap-tops), the time will certainly come when "converged" terminals, like the Media Screen prototype of Nokia, will be a very common way to get access to any infotainment services through the broadcaster, not only at home but in any leaving place.

11. Mobile reception could be attractive in cities where people spend time on public transport for commuting. In fact, mobile reception could be used for other applications.

11.1 What are the possibilities of each standard to accomplish these types of services?

DVB: This is a very strong point of DVB.

COFDM modulation technology has been chosen for terrestrial broadcasting by the DVB Technical Module, because it really makes use of all digital electronics to satisfy the most severe requirements of the broadcasting members of DVB (working in the DVB Commercial Module).

The use of the "guard interval" parameter, eventually associated with 16QAM or QPSK modulation instead of 64QAM, allow to cope with the Doppler effect related with dynamic echoes and the mobility of receivers.

Numerous experiments, field trials and even pilot projects have already shown the strength of DVB-T to offer stable video, audio and data to mobile receivers.

Too articles in the DVB CD-ROM give detailed information on the technical aspects of some of these trials:

"DVB-T and DMB in Mobile Environments" by R. Burow et al. and "DVB-T Mobile Reception trials in Cologne" by DTTV-SA .

11.2 If applicable, what is the trade off resulting from it?

DVB: The trade-off for strengthening signal reception is bit-rate reduction. Depending of the choices made, as a function of local topology and quality of service decided, mobile reception would need, for example, to reduce the 22 Mbit/s static service bit-rates to 11Mbit/s (16QAM, CR2/3, TG1/8) or eventually to 5.5 Mbit/s (QPSK, CR2/3, TG1/8) for longer distance reach in 6 MHz channels.

11.3 Are there other mobile applications being envisaged, apart from digital television?

DVB: Digital television to the commuters, in great cities public transportation in great cities like Mexico is an obvious application. Singapore is presently preparing the implementation of such a concept. We highly recommend the Mexican authority to get in contact with the Singapore Broadcast Authority, Chairman Mr. Hock Chuang LIM ([email protected]).

Germany considers such services as extremely important, especially for long distance highway travel and in large cities.

Despite this however, many DVB members feel that the major market demand will come, as already mentioned in § 10.3.2 about corresponding equipment, are infotainment services to mobiles. The access to real time information on traffic conditions, on weather, on transportation facilities (buses, trains and planes) are usual needs while you are already on the move. Gambling, horse racing, sports at any time are also expected to represent a significant market. With the development of Internet and mobile telephony, the access to broadband Internet services (information and entertainment like games) will be a favourable area for broadcaster to offer.

12. High definition is one of the features that could be consumer attractive for the new digital television products.

12.1 What are the possibilities of each standard to accomplish this requirement?

DVB: High Definition is standardised in DVB as well as EDTV and SDTV. In fact, DVB covers the whole MPEG2 range of TV formats as seen in DVB-MPEG standards .

There is absolutely nothing preventing broadcasters to use DVB-T for broadcasting HDTV. The fact that DVB-T offers implementation parameters allowing up to 23.8 Mbit/s for a 6 MHz channel as compared to the fixed 19.3 Mbit/s offered by ATSC. This bit-rate advantage may be used for the simulcast of the same programme with SD or ED image formats for the viewer majority without HDTV displays.

In this view, Australia precisely unanimously chose DVB-T to implement its HDTV broadcasting strategy after its extensive comparison with ATSC. The reader will usefully analyse the report of the DTTB Selection Panel that has considered all elements of the two standards and concluded that "while Europe had chosen not to go HDTV, the DVB system was capable of HDTV and had fully defined it in its standards during 1997. […]. The Panel considered that both systems permitted any current HDTV standard to be used." .

Australia is preparing the launch of its HDTV programmes and will start commercial broadcasting 1 January 2001. Article "DVB and HDTV" by Ken McCann and Charlie Sandbank explain in detail how HD might be implemented.

Singapore, after comparisons between ATSC, DVB and ISDB, has concluded on this point that "High Definition Television for viewers means good quality pictures and sound. HDTV application is possible in all three systems, […] Equipment for the DVB standard is expected to be available by the year 2000." .

12.2 If applicable, what is the trade off resulting from it (see question 3)?

DVB: The full use of the maximum bit-rate is not necessarily optimal in all environments and some trade-off are usually made between bit-rate and C/N threshold (or coverage area) and/or ruggedness to fixed and variable echoes.

For example, the London Crystal Palace transmitter uses parameters offering 24.13 Mbit/s (about 9.3 M inhabitants for 10 kW ERP). This would give 18.1 Mbit/s in a 6 MHz channel, far enough for one excellent quality HDTV programme.

12.3 DVB is using EDTV at Europe for the introduction of digital television, are there devices compatible or expandable to HDTV?

DVB: European broadcasters have chosen to launch DTTV with EDTV image quality for economical reasons. The experience of the HDMAC project has shown that European customers do not consider the improvement of resolution beyond PAL or SECAM as worth more than 20% price increase.

The price of HD displays being significantly higher due to the need for larger image dimension -HDTV is not significantly visible unless the display is larger than 40" and looked at from about 5 meters- and extra-costs for of resolution.

Added to this, the increase of resolution of HDTV up to 720 to 1080 lines over the EDTV 554 lines is made negligible in terms of picture sharpness by the lack of horizontal resolution of present so-called HDTV displays based on picture tubes.

Considering display prices, as typical examples, the present market price (OEM large quantities) of the largest EDTV tube on the market (36" diagonal) is about US$ 640, while similar dimension real HDTV (1080*1920) plasma displays are more than US$ 4 000. In fact, only projection displays can offer intermediate prices with HDTV resolution. This display manufacturing industry estimates that basic price evolution of the high resolution displays in the large display business (30-60+ diagonal) will reduce slowly with time (a few % per year), due to the expected relatively slow penetration (from ~ 15% in 1999 to ~24% in 2005, of which professional is the largest part).

In any case, the bottleneck of HDTV, the display is independent of the standard.

It is usually considered that the period of analog shut-off (2007 to 2009 in the UK as recently announced by the government) may be an appropriate period to introduce HDTV in the frequencies made then available.

Considering the receivers, DVB equipment manufacturers have committed to Australia (and to Singapore), that intends to launch HDTV with DVB-t in January 2001, to put receivers on the market as soon as the market opens.

Three different IC manufacturers are already producing decoding IC's for DVB HD-ED-SD: STMicroElectronics, Philips Semiconductors and Fijitsu.

A Hong Kong company E.F.A. already offers HD DVB-T set-top boxes based on the STMicroElectronics chip-set.

Considering the 500 000 IRD's installed in the UK in the last 10 months, they have no facility for decoding HDTV neither the possibility to be upgraded by simple software download. The estimated increase in cost of HDTV over EDTV IRD's is estimated by the IC industry around 20%, due to the increase of memory size and of signal processing capacity.

In general, the IRD industry generally aims at reaching production costs around US$ 200, for standard IRD's estimating that this price will stay stable with time as IC cost/performance decreases and as features (like HD decoding) are progressively introduced.

13. Another distinctive element of digital television are the audio signals.

13.1 ATSC is currently proposing AC-3 and DVB is using MPEG-2, what is the standard flexibility to migrate to other audio compression schemes?

DVB: DVB is an open standard. This means that none of its elements are proprietary (owned and controlled by a company). This comes from an initial decision of DVB to offer to the users, manufacturers and broadcasters full freedom to develop their market in a free competition environment. This fits with the controlled availability of IPR's at non-discriminatory and fair conditions.

This is also the reason why DVB makes use of the maximum of already existing open standards, like MPEG, DAVIC, etc.

For this reason, have chosen MPEG-2 as video standard, DVB chose also MPEG-2 as surround sound standard in complement to the MPEG-1 layer-2 mono and stereo sounds. The only competitive mature solution was Dolby AC-3. All panels of independent experts considered that both performed equally well. As AC-3 is a technology owned and controlled by a private company DVB and no possibility to guaranty perennality, upward/backward compatibility of future generations nor right of access to the technology to all. It was therefore rejected as the recommended choice for DVB systems.

DVB is however "open" in a different way: many opportunities exist in DVB standards to establish implementation parameter and configuration choices adapted to the local environment. In this respect, Australia, for instance chose to offer its broadcasters the choice of MPEG-2 and AC-3 for surround sound. A possibility for specific services was available in the DVB standards and DVB tested and documented AC-3 as a additional sound to MPEG-2. This solution, although not recommended by DVB (see comments above) can be used in any country.

14. One the issues to consider for the adoption of a standard is flexibility on the systems and growth capacity, as well technological compatibility between receivers and transmitters of different generations.

14.1 What is the status of both standards for this issue?

DVB: The three major rules for DVB standards are:

• flexibility (they should allow broadcasting new types of services for the future 30 to 50 years)

It is in fact the market that mostly controls compatibility: if non-compatibilities appear, the non conform products are rapidly identify and have to be corrected. Some actual laboratory tests have been implemented. For DVB-T, an important test has been performed by the VALIDATE project, supported by the European Commission. Its results are presented in the article "Report on Multidimensional DVB-T Interoperability Tests" by ACTS Validate Project Rennes, 8-12 June 1998 . The full report (68 pages) is available from http://www.bbc.co.uk/validate.

The results are impressive, since the comparison was involving numerous manufacturers of transmitters and receivers and several equipment generations.

Considering software/middleware upgrade, very difficult subject for consumer products that cannot accept non-bug-free solutions, the very powerful facility for downloading software automatically is a great asset of DVB. This facility is essentially made to allow the upgrade of EPG's and API's that are still being optimised according to service needs. Already four upgrades have already been operated in the first 10 months operation in the UK. This is now a routine procedure. The process takes place at night when the IRD's are in stand-by mode. They are not detected by the user, apart from their effect on the service (new presentation of the EPG for instance).

15.1 Specifically, what are the possibilities foreseen by each standard? Is it possible to achieve gradual introduction of the receivers or set top boxes, something that could be similar to what one can do with a computer? What is the modularity and scalability of the receivers?

DVB: DVB standards have been specifically designed, to the demand of the broadcasters and operators working together in the "Commercial Module", in such a way that they can offer the broadcaster any business model he wishes, for the launching of his services as well as for their long term operation.

In general terms, the upgrading scenario which is considered by a large part of the broadcasting and equipment manufacturing industries is the following:

• the improvement of image quality and of interactivity will be parallel. The basic reason is related to market issues: there is a demand for both. Even in the USA where HDTV is the result of an industrial/political push, there is a significant demand for interactivity on the TV (on cable and satellite). In Europe, as well, the viewers are expected to ask for HDTV when HD displays become accessible.

The possibility to receive DTTV signals without replacing immediately your analog TV set significantly reduce a the high financial hurdle to the TV of the future. This allows the fastest start of the business.

The SCART plug, offering RGB connection, or the AVLink, are by far the best solution since they offer a very high image quality, without any noise nor analog interference of any kind related to the analog re-modulation. The presence of SCART plugs on all TV sets in England is one of the major reasons of the great success of DTTV: it offers both plug-and-play and high image quality).

From the first generation on, all DVB-T receivers include the capacity for upgrades by Automatic Downloading. This process is operated by the broadcaster, during the night, in a very short time, while the receiver is in its stand-by mode. It allows either upgrading the IRD software itself as well as elements of the API related to the service provider (change of the visual aspects of the Electronic Programming Guide for instance). Such operations are now done routinely in the UK.

The MHP (see § 10.2) will play an important role. As an IRD it will constitute the major piece of hardware for the DTV system in the home, able to receive from any medium (terrestrial, satellite and cable) all video and sound formats and all types of data. Its upgrades -based on the use of a Java machine and Java programming- will be made through automatic downloading at very high speed. The presence of high memory capacity (IC and hard disk) will allow autonomous operation, but interoperability with PC's and other data storage and processing machines.

As the transmission performance, the analog bottleneck, is drastically improved, the display will become the next speed-of-upgrade limiting factor through its performance/cost ratio evolution.

In fact, the real bottleneck for the overall upgrade of the home equipment will be the content itself. This holds for the traditional video content as well as for the interactive content. We mean here not only the adequacy to a "format" like HD or HTML, but for the really high quality image and emotional content or with the originality and fun provided by interactivity.

16. An element that could facilitate cost reduction is market growth.

16.1 One of the major elements we see that can facilitate the deployment of DTTB is the broadcasting business approach itself .

DVB: The most important business approach now followed by most European countries and being planned by most world countries for DTTB is very similar to the that implemented by satellite operators. It consists, generally through commercial agreements to achieve significant economies of scale through multiple programming.

A given multiplex can offer 4 to 5 programmes or services. These services, whether they belong to the same operator or to several different ones allow to share some of the costs given as typical example in table 1 in Appendix II-1. This appendix is inspired by the UK market, the only successful DTTV service launched today.

Furthermore, DVB allows a series of common features, like the Service Management System, like Service Information, Conditional Access, etc. that allow a number of multiplexes to run on the same platform and make a further significant increase of economy of scale.

Fig 3 shows how sharing costs of 5 programmes in a multiplex and, further, aggregate 6 multiplex into a common business reduces the initial capital per service or programme in a dramatic way.

Figure 4 shows the typical yearly costs in similar conditions for a 1Million-subscriber business.

This means business agreements within a correctly established legal framework.

Such agreements are not always easy to establish, but are worth the effort as shown from Fig 5 that shows the financial balance (revenues-costs) per programme in the same three situations.

This business calculations have been made on the bases of the costs shown in Appendix II-1 and of the multiplex and bouquet aggregations discussed here before. We like to highlight that such investment and operational costs as those of technical buildings have been included while not technical elements like financial costs, human resource costs (education and training) or marketing costs (advertising campaigns) have not been include because they depend too much on cultural and local economic conditions.

The revenues are inspired, as typical values, from the ONDigital model, but could be made as well according to a public broadcaster scheme like the one of the BBC. The advantage of the private sector model is their higher transparency.

In these calculations, we have included the following revenues:

• US$ 120 / year for basic subscription (including ~ US$ 80 for IRD leasing)

With specific measures, it might be presumably possible to reach break-even sooner for the bouquet approach and perhaps to make the one multiplex (5 services) positive too at long term. However there is no doubts that a 1 service / multiplex approach is a basically non-profitable business.

Subsidising the IRD:

Among the broadcasting business elements able to speed up the start and increase the final penetration of DTTB, we have already, in the previous descriptions mentioned the "subsidising" of IRD's by the broadcasters.

This unusual process for Free-To-Air broadcasters can be implemented as soon as they start offering Premium channels. It should be clear that this in no way represents a gift to the user. It can be implemented in different ways:

• Leasing: the IRD is the ownership of the customer after and minimum contract period ( ONDigital solution)

Informing the public is essential. It has play a significant role in the DTTB launching in the UK with a simple slogan "Digital from your own antenna".

The idea of TCS in Singapore to support with practical demonstration such a launching campaign is the installation in all commuter train, metros, buses and taxi of DTV sets based on the mobile proprieties of DVB-T. No doubt all Singaporean citizens will discover, and appreciate, Digital TV very rapidly.

The consequence of this, as well as of the cost of IRD's discussed hereafter is the penetration curve estimated for UK. This curve fits the present reality (~0.5 Million viewers after 10 months) and the plan of the UK government for analog switch-off within 7 to 10 years with a penetration higher than 95 %. This curve is shown on Fig 6

For the other European countries, all presently preparing their implementation through pilot projects and business preparation, see the "European Service Model" .

Singapore is preparing its launch. It will presumably at the beginning of 2000.

Australia plans to launch all its FTA and commercial services 1 January 2001.

16.2 What are the economic projections envisaged in terms of cost reduction related to number of receivers sold?

16.3 What are the growth expectations in countries that have or could adopt DVB or ATSC?

17. If Mexico adopts DVB, a different standard would be used on the common border between Mexico and the United States.

17.1 What are the economical and technological costs associated with the need of having receivers that could be able to decode ATSC and DVB, as well as for the transmitter needs for handling both standards?

For receivers:

DVB: In such a situation DVB-T receivers should be compatible with ATSC broadcast programmes. They therefore should necessarily:

• Have two inputs and their corresponding front ends.

Technically, this should be a double front-end receivers (or triple is satellite is also to be received), with standards front-end modules. Such IRD's are already being designed by DVB industrials for multiple source IRD's (terrestrial, satellite/cable)

The decoding part is even easier to realise, since a DVB member, STMicroElectronics, already produces a HD decoding chip-set able to decode both DVB and ATSC . The difference is in the software, it should then be programmed to do both.

Such receivers would only be in the high-end due to the limited numbers to be produced (niche market).

Generally speaking, the extra-costs of such specific dual-standard receivers should be limited, compared to the cost of the HD display.

For transmitters:

DVB: The extra investment for broadcasters is also limited. Certainly very small compared to the investment needed for implementing HD.

Content can be the same. Practising HD (for broadcasting towards the US needs all HD equipment, including format (up- & down-)converters.

For the rest, taking into account that DVB-T programmes should be broadast in different multiplexes than ATSC, there should be no extra-costs for the broadcasters.

• A minimum of one HD ATSC programme occupies a multiplex and requires the ATSC real time encoder, and associated equipment, the 8VSB modulator and the corresponding transmitter with the proper directional antenna towards the US territory.

18. It has been stated that ATSC receivers and transmitters are, or could be cheaper than those required for DVB.

18.1 What are or could be the costs associate with this? What are the volume cost trends?

DVB: The statement that DVB-T transmitters and receivers might be more expensive than ATSC for similar functions is incorrect.

Modulators and transmitters

Although modulation technologies are different, the basic requirement placed on transmitter is similar and therefore, cost implications are similar.

While transmitters are basically standard independent, differences could be imagined in the difference ERP needed for the COFDM and 8VSB in favour of ATSC. This advantage of 8VSB on this point is, in theory for a Gaussian channel, by 0.3 dB. Now ancient measurements, like the Australian field trials showed about 4 dB difference. We have already highlighted in article "Facts about DVB-T"- by Martin Jacklin (DVB), Peter MacAvock (DVB) and Andrew Oliphant (BBC R&D) , these measurements were made with a 7MHz DVB-T prototype versus a standard 6MHz ATSC receiver. Recent measurements on present commercial IC's and commercial IRD's bring the difference down to ~2 dB, which is really minor difference.

In reality, the strength of COFDM in real environment shows that, in far field, coverage of both systems is similar for the same ERP level, as shown by Sinclair in their comparative field trial "Comparative Reception Testing of 8VSB and COFDM in Baltimore" by Nat Ostroff and Mark Aitken, Sinclair Broadcast Group :

Furthermore, when considering the powers needed for digital TV as compared to analog, the point appears as a minor issue. The ERP needed for DTTV is about 100 times smaller than for analog. (London: 9.5 Million inhabitants with 10 kW ERP instead of 1 MW for analog for the same coverage).

Receivers

One of the major criteria for DVB to choose its transmission standards was that receivers should be as low cost as possible, even to the eventual detriment of transmitters costs.

The chosen transmission for DTTV, COFDM, finally allows the receiver to be low cost -thanks to the absence of any equiliser- without any increase of the transmitter. Considering the whole IRD, it is however clear that the largest part of the cost is in the MPEG-2 decoding, and in the video-processing. These two functions are independent of the standard. A basic DVB-T decoder is, therefore, in principle cheaper than a ATSC IRD that includes an equaliser, more specifically if future versions will re-enforce this front end part to reduce echo effects on urban reception.

There are other major reasons for DVB-T IRD's to be lower cost, now and in the future, than their equivalent ATSC equipment. These reasons are related to the fact that DVB creates its standards on a concept of platform interoperability and maximal commonality, precisely in order to achieve maximal economies of scale for its receivers, to the benefit of all parties: viewers, broadcasters and equipment manufacturers. This has been explained already in § 16.1

18.2 How do these costs affect in terms of repeaters and/or low power gap fillers?

DVB: In the broadcasting infrastructure, the cost of repeaters and gap-fillers -as well as flexibility for implementation- have been strong arguments for the choice of COFDM by DVB.

This element is, therefore, one of the strong points is of DVB-T as compared to ATSC.

Two types of repeaters are commonly used:

b) Transposers with base-band re-processing to ensure that the re-emitted signal is according to defined quality level.

For digital TV, due to the cliff-effect of DTTV reception, the conditions for using one or the other might change. The use of COFDM as a modulation technology has a significant impact on this issue.

The number of repeaters needed with COFDM will be significantly lower. For single-carrier VSB transmission (being analog or digital) the areas with strong echoes have to be considered as "dark areas" -ghost images for analog, blue-dead images for digital-, while for COFDM, the signal / noise ratio is even increased by the presence of the echoes (constructive interference).

2) The second and major argument in favour of DVB-T on this topic is the possibility for on-channel repeaters without the drawbacks and extreme precautions about in/out signal protection needed by VSB. The constructive interference effect achieved with COFDM allows the use of on-channel repeaters to be implemented for any weak reception area without neither major technical problems nor important equipment and infrastructure costs.

In this framework, the use of even very low on-channel repeaters has been shown as a solution to perfect reception of DVB-T signals even in large re-enforced concrete buildings.

3) The third and decisive argument on this issue in favour of DVB is the possibility to implement, at minimal cost, micro- or local Single Frequency Networks. This technique offers the very big economical advantage. It allows the use of significantly lower power main transmitters without extremely high towers -as usual in flat and building free area like in the US or Canada- and to optimise locally with repeaters and micro-SFN's.

In terms of equipment cost comparison between ATSC and DVB repeaters, let us only consider the electronic part:

b) For a frequency transposing repeater the present market price is around US$ 12k. It can be nor lower for an on-channel repeater.

Considering the number of such repeaters needed in such environments as Mexico City and in mountainous countries as Mexico, the economic issue is very important.

In terms of price evolution, the advantage of DVB will further increase with time. While the countries that have adopted ATSC are based on the concept of very high power main transmitters, with a limited number of repeaters, European countries -and most countries in the world- have already broadcasting infrastructures with a high number of repeaters. As an example, such a rather small country as France has about 12 000 repeaters installed. This situation shows that the market of DVB-T for on-channel repeaters, that has already significantly developed with the launch of UK and Sweden, will boom in the coming years and bring significant price reductions.