PLANNING STUDIES INTO THE ALL-DIGITAL FUTURE AND DIGITAL SWITCH-OVER SCENARIOS C.R. Nokes, N.J. Laflin, D.J. Darlington BBC R&D, UK ABSTRACT With digital terrestrial television now launched in the UK, the Government has made clear its intention to switch fully to digital broadcasting possibly starting as early as 2006. Before this can be done, they require near-universal coverage of digital services, and full access by viewers. This paper describes the results of planning studies conducted to determine the optimum way to provide universal coverage of DTT services, and concludes that an evolution of the current DTT plan is the best way forward. It shows that the use of analogue conversions will be an extremely important part of this process, but notes a number of difficulties to be resolved and outlines a possible migration strategy. It also considers the diverse range of requirements for digital television broadcasting from European countries, and explains the need to revise the Stockholm Plan the European frequency plan for television broadcasting. INTRODUCTION In November 1998, the UK launched a digital terrestrial television (DTT) service the first country in the world to do so. Throughout 1998 and 1999 a programme of DTT transmitter building was carried out, and there are now 80 DTT transmitters on air, covering about 86% of the UK population. These DTT transmitters were mostly planned using UHF channels that were interleaved with the analogue channels, and caused little interference to the existing analogue network. The UK Government envisages eventually switching from analogue to full digital broadcasting, possibly beginning as early as 2006. Before any such switch-over could commence, the Government indicated they wanted to ensure that virtually the entire UK population was covered by digital transmissions, and that most consumers have access to digital equipment. BBC R&D has started to study how these goals might be achieved, adopting two approaches. Firstly, we have undertaken planning studies to determine what would be the ideal means to provide universal DTT coverage. Secondly we have started to consider the process by which we might migrate from the current UK mixed analogue/digital plan to the final fully digital scenario. This work has been carried out separately from the work in the UK Independent Television Commission s (ITC) Genesis project. It is envisaged that the eventual switch-over to alldigital transmission will free a number of UHF channels for new uses, possibly including new broadcasting applications. One of the purposes of the work has been to determine how many channels will be freed, and which channels would be the most appropriate. This paper presents the results of the studies carried out by BBC R&D to date. However it should be noted that these are interim results, and considerable further work will be required before a fully developed plan can be produced. The paper also considers the diverse range of requirements for digital television broadcasting from European countries, and explains the need to revise the Stockholm Plan the European frequency plan for television broadcasting. UK PLANNING STUDIES THE ALL DIGITAL FUTURE Having planned the UK DTT network and left its implementation and roll-out to the transmission providers, the frequency planners at the BBC were in a position to spend more time on consideration of the all digital future. Essentially we were asking ourselves the questions: what will the licence fee payer want from terrestrial DTT in the future how will the BBC (in association with the other multiplex operators) be best able to deliver this what coverage of the UK will be achievable A range of planning studies was undertaken to try to determine what would be the ideal means to provide universal DTT coverage within the UK.
These studies have made use of a network of 128 transmitters, comprising all high power main stations, relays and some extra gap filling relays. At each site, the effective radiated power (ERP) was set to 10dB below the analogue power with a maximum of 50kW and a minimum of 250W. No external interfering sources were included. For each condition to be considered, the coverage was calculated for 95% locations, or good coverage as defined in the Chester 1997 Agreement (1). The coverages of several different types of network configuration were calculated to allow the following planning issues to be studied: Fixed, portable or mobile reception Multiplex capacity Number of channels per multiplex required Multi-frequency networks (MFN) Single Frequency Networks (SFN) national, regional or sub-regional Guard interval FFT size (2K/8K) possibilities for spectrum release Results Of The All-Digital Planning Studies A summary of the results from the studies for fixed reception is shown in Figure 1. Although this work is continuing, the following initial conclusions can be drawn for the options we investigated: 1. The national SFNs do not provide adequate coverage for the 128-station network in the UK. 2. Improved coverage is achieved with the regional and sub-regional SFNs, especially with the longer guard intervals. However it should be noted that these imply a reduction of the data capacity. 3. The 4-channel MFN provides reasonable coverage, but it would probably be very difficult to extend the network beyond the initial set of sites. 4. The 6-channel MFN provides good coverage, and should allow scope for extending the network beyond the initial set of sites. 5. Mini-SFNs could still be useful to provide gapfillers, and may be required to complete a network for portable coverage. There are currently 6 DTT multiplexes in the UK. If an MFN with 6 channels per multiplex is selected for the all-digital frequency plan, a total of 36 UHF channels will be required, out of 46 channels available to broadcasting in the UK. This suggests that either one additional multiplex could be accommodated, or up to 10 channels made available for new uses, possibly including new broadcasting applications (e.g. networks for mobile broadcasting). A study into which channels could most easily be made available for such applications, has shown that releasing the channels above 60 would cause the minimum disruption. UK SWITCH OVER SCENARIOS The results presented above suggest that, at least for the time-being, for the UK transmitter spacing and terrain, and given the data rates required, there is no advantage to be gained from the use of SFNs for the initial transition to an all-digital plan although SFNs may still be required eventually to Percentage population 100% 90% 80% 70% 60% 50% 40% 30% 7 µs (2k only) 28 µs (2k or 8k) 56 µs (2k or 8k) 112 µs (8k only) 20% 10% 0% National SFN 4 channel regional SFN 4 channel subregional SFN 4 channel MFN 6 channel MFN Figure 1 Population coverage for 128-station networks, for fixed reception at 95% locations
fill coverage gaps. This finding is helpful because it implies that the all-digital plan can be a development of the current MFN (interleaved frequency plan), and so the switch-over may be evolutionary rather than revolutionary. The preceding general discussion hides the problems that will become apparent in some areas when the details are examined. Crystal Palace Consider for example the transmitter that serves London and the South East of England Crystal Palace. The analogue transmitter serves approximately 11.6 million viewers, but the current best DTT transmitter serves only about 8.8 million viewers. Therefore 2.8 million viewers of the Crystal Palace analogue service are currently unable to receive DTT signals from Crystal Palace. This discrepancy is mainly caused by the need to restrict the DTT power radiated in certain directions to prevent interference to transmissions in other areas, mainly in neighbouring countries such as France, Belgium and The Netherlands. It is therefore extremely unlikely that it will ever be possible to extend the coverage of these DTT transmissions from Crystal Palace to serve the remaining 2.8 million viewers. Some, but by no means all, of these viewers may be able to receive DTT transmissions from other transmitters. This problem is illustrated in Figure 2. The outer region is served by analogue transmissions, but is not yet served by DTT transmissions. Analogue Conversions The only practical way to extend the coverage of the DTT transmissions from Crystal Palace (without impacting on the analogue and digital assignments in neighbouring countries) is by making use of the existing assignments for analogue broadcasting. This process was foreseen in the Chester 97 coordination agreement (1), and is known as an analogue to digital conversion: an analogue assignment may be converted to a digital assignment with a mean power 7dB below the power of the analogue assignment. Continuing with the example of Crystal Palace, we have carried out an initial study of the coverage achieved with digital conversions. For this, our prediction program converted all the relevant stations (wanted and unwanted) to their digital equivalents as specified in Chester 97: thus, it reduced the power by 7dB (but fixed the maximum allowable converted power at 100kW). It also maintained any HRP or VRP applied to the transmitters. Under these conditions, the coverage of the predicted digital transmission was 1% greater that that of the original analogue transmission in other words the 2.8 million viewers who are currently not served by DTT would be served following an analogue conversion. Therefore, in spectrum planning terms, the use of analogue conversions looks very attractive at Crystal Palace, and at some other stations where it will be virtually impossible to replicate the analogue coverage otherwise. Ipswich Oxford Watford Reading LONDON Croydon Dover Southampton Brighton Analogue DTT Figure 2 Coverage of Crystal Palace analogue and current DTT transmissions
Logistics of conversions Whilst from a spectrum planning point of view the use of conversions appears very attractive, in terms of both the network provider and the viewer it raises some serious questions. Network provider issues for conversions The use of conversions will pose some difficult challenges for network providers, but it would appear that these can be solved (although they may involve additional costs). The issues fall into two main categories the transmitter equipment and the distribution network. Some of the equipment at a transmitting station will not be capable of direct re-use for digital signals, and will require modification or replacement. Therefore, the only way a conversion could be carried out without a long break in service is by the use of a temporary transmitter alongside the existing transmitter. The conversion would then be carried out by switching the feeder. The distribution network poses a more difficult problem for the network providers. Currently, many of the stations in the analogue relay network are fed by off-air transmissions from parent main stations. When the main station is converted to digital, it may be that some relay stations need to continue to transmit analogue signals for some time. These will either need to be given a separate line feed of analogue signals (which would be very expensive), or make use of the digital satellite or terrestrial transmissions. One problem with using DTT transmissions to feed analogue relay stations is that currently there is no provision for the carriage of teletext within the DTT signal (since digital text is provided instead). Therefore a new mechanism would need to be found for delivering teletext data to a relay station, once its parent main station was converted. Of course, since the eventual aim is to switch all analogue transmitters off, it could be argued that the temporary loss of teletext from the relay station was acceptable. Viewers issues for conversions By far the most complex issue in dealing with conversions is dealing with the viewers. If an overnight conversion were used as described above, all viewers would need to have digital receiving equipment ready for use prior to the conversion. However some would have no means to check this equipment prior to the conversion, so this could pose serious difficulties and is likely to prove extremely unpopular. In previous conversion exercises, (such as the conversion of every UK house to burn North Sea gas in the sixties; 405 to 625-line television conversion, and the Channel 5 VCR retuning exercise), a mechanism was available to allow the conversion to be carried out gradually. In the case of North Sea gas conversion, small areas could be dealt with at any one time. In the case of 405 to 625-line conversion, dual band transmissions were used, and dual standard receivers were available for many years. In the case of the Channel 5 VCR retuning exercise, retuners were able to visit viewers with a test generator prior to switching on the new transmitter, to prevent interference from occurring. One method by which viewers could be given more chance to test their digital equipment prior to analogue to digital conversion is if short digital test transmissions could be used. These could start during non-peak viewing, and gradually be extended into peak hours, to encourage all viewers to make the checks. This also suggests that a digital television receiver that was capable of automatically switching from analogue to digital and back would be extremely useful. An alternative to the temporary test transmissions mentioned above would be the temporary use of channel 36, if it could be allocated to broadcasting. Its use would however imply the need for two transitions the first from analogue to digital in channel 36, and the second from digital in channel 36 to digital on the converted analogue frequencies. The benefits of this need to be investigated further. Other approaches that are worthy of further study include the conversion of only one channel at a time, and a pro-active approach, similar to that used by Channel 5. Overall migration strategy The above discussion has concentrated on the use of conversions as a means to replicate the analogue coverage. However, in areas where conversions are not being used, many of the issues remain relevant. The overall migration process could start by extending the existing DTT network where it is cost effective to do so. Analogue conversions could be used in areas where there are serious coverage deficiencies, and digital satellite or cable services could be relied on in areas where it is either impractical or too costly to provide DTT transmissions. The migration strategy will need to be developed and agreed with both the Government and the whole of the UK television industry in the coming months and years. It will need to ensure that the best route to digital can be achieved, which
maximises availability of services, minimises costs to broadcasters, and minimises inconvenience to viewers. A key part of this migration will be to ensure that as many viewers as possible have digital-ready television receivers in the run-up to the switch-over ideally, all new television sets and VCRs sold today should be digital-ready! REVISION OF THE STOCKHOLM PLAN THE INTERNATIONAL PERSPECTIVE The foundation for all television planning in Europe is the Stockholm Plan of 1961 (2) known as ST61. In general terms the Stockholm Plan provided the possibility of three programme coverages in each country. Some countries including the UK had a target of four programmes. Of course, the UK's beneficial geographic shape and position being an island with only one international boundary helped in this respect. Since 1961 two major developments have taken place. The first is the building of many thousands of low power relay stations to fill-in coverage holes and the second is the introduction of new networks, usually with less than national coverage, but with a commercial target. For example in the UK we have analogue Channel 5 with less than national coverage, and more recently we have introduced the six digital networks. Overall, the total of number of television transmitters in Europe now exceeds 80,000. Given the Stockholm plan how is it possible to identify sources of spectrum for digital networks in Europe? Three options can be can be considered: 1. Some countries have not implemented all of their analogue networks or assignments and these may be converted to digital. 2. To interleave between existing analogue channels (as done in the UK) although in some countries these interleaved channels have been used for non-national analogue networks. 3. To make use of channels above 60 (that is above 790 MHz) which in most countries in Europe have been reserved for military operations though this is not the case for the UK. In some countries this is the only spectrum that can be used for the introduction of digital television. Obviously, no provisions were made in the Stockholm Plan for adopting any of these three options for the introduction of digital television. Within "CEPT Europe" (43 countries), this has been facilitated by means of the Chester 97 Agreement (1) on the criteria and procedures to be used in the initial introduction of digital television in Europe. The Stockholm 61 plan and the Chester 97 Agreement are by no means ideal to meet the objectives of a fully digital broadcasting scenario since they are optimised for an "analogue world". In order to make the most efficient use of the available spectrum and to release the full potential of digital television services, the majority of CEPT administrations have indicated that a new frequency plan for Europe should be developed. Further, they consider that this should be done under the auspices of the ITU and should include all 56 countries in the European Broadcasting Area. CEPT ideas for an ITU Conference With regard to the preparation of the conference, it is proposed that the ITU-R should prepare the necessary bases required for planning. A twosession conference is envisaged with the sessions separated by about two years. The first session would determine the principles for developing a plan, which would include technical parameters as well as migration strategies, and the means to achieve equitable access to the spectrum. The second session, which would produce the required Plan, is anticipated sometime around 2005, but clearly this will be a decision for the ITU. Range of requirements Before any plan can be prepared it is necessary to determine each country s position with regard to coverage and capacity targets; their requirements for service type (fixed, mobile or portable coverage); network structures (existing networks, denser or even cellular networks); planning model (MFN or SFN); and DVB-T variant the standard offers a wide range of options to suit different requirements of capacity and robustness. Various scenarios could be developed from combinations of the above factors. However, what is clear is that any single scenario is unlikely to be appropriate for all countries, or even for all areas within a single country, and consequently a mixture of service requirement models will be required. Planning Options Within the EBU and CEPT, members have been considering the pros and cons of various planning
options. To date, three main options have emerged as described below. Conversions As described before, conversions re-use high power analogue assignments for digital transmissions. Conversions are very attractive because they provide the simplest means of matching the analogue coverage of each transmitter. Since existing assignments are used, transition can take place by a series of bilateral agreements between countries. However, as well as the difficulties mentioned previously, conversions do not necessarily produce the optimum plan for efficient use of the spectrum, and they restrict the scope for SFN planning (and hence for portable and mobile coverage). Also, they may not lead to equitable access to the spectrum, and make spectrum release more difficult. Existing DTT network This option would make use of existing digital networks, in countries where they had already been developed. This is clearly attractive as transmission equipment and services would already be in place, and co-ordination agreements should already be in place. There is scope for further extension and improvement when the analogue service is switched off. As for conversions, this approach may not lead to equitable access and is not optimised for spectrum efficiency. Initial coverage will be limited and spectrum release may be difficult. Re-plan In this scenario planning could start with a "clean sheet" and the existing ST61 assignments would not necessarily be taken into account. This is more likely to lead to equitable access to the spectrum and should be better optimised for spectrum efficiency. It should make it easier to release spectrum. However, it would mean a more difficult transition, probably with significant breaks in service. New transmitting equipment and receiving antennas would be required, and synchronisation of the transition between countries would be difficult. New co-ordination agreements would also be required. Need for flexibility countries indicates that the long term aspirations are likely to be very different. From the UK perspective, it is clear that the proposed conference would have an impact on the DTT planning process in the UK. As is nearly always the case with international conferences, it is anticipated that the final solution will involve some compromises by all parties. At the current time it is too early to specify what the real issues will be. CONCLUSIONS This paper has reported the results of a range of planning studies to determine the optimum network, given the UK s constraints of transmitter spacings, terrain and required data rates. It shows that, given these constraints, there is currently no advantage to be gained from the use of SFNs, and so an evolution of the existing MFN DTT network should ultimately provide the optimum network for the UK. It has also been pointed out that, within this evolution, in some areas there is considerable advantage in planning terms from the use of analogue conversions. Conversions could form an important part of a migration process that includes making use of digital satellite and cable services in areas where it is either impractical or too costly to provide DTT transmissions. We have also discussed the diverse range of requirements that are anticipated for digital television broadcasting from European countries, and explained the need to revise the Stockholm Plan. It is anticipated that the ITU will arrange a conference to produce a revised Plan in about 2005. REFERENCES 1. CEPT, 1997. The Chester 1997 Multilateral Coordination Agreement relating to Technical Criteria, Co-ordination Principles and Procedures for the introduction of Terrestrial Digital Video Broadcasting (DVB-T). CEPT Multilateral Co-ordination Agreement, 1997. 2. ITU, 1961. Final Acts of the European VHF/UHF Broadcasting Conference, Stockholm, 1961. International Telecommunication Union, 1961. Whilst the requirements of the various countries have yet to be established in a formal sense, early information provided by a limited number of