Simulcast is an option of particular interest to broadcasters who have to continue to satisfy existing analogue listeners for several years to come, but wish to introduce DRM services as soon as possible. In many cases these broadcasters are restricted in the ways in which the digital service can be introduced. For example they may have a single MF assignments and no prospect of receiving an additional frequency assignment to start a digital only version of their service. They may also be keen to avoid having to make a short-term investment in an additional transmitter and/or antenna and site to start a digital service on a new frequency.
These broadcasters would like to be able to transmit simultaneously both the existing analogue service and a new DRM service, with the same content, whilst using the existing transmitter and antenna. This option is probably most applicable to broadcasters with LF or MF assignments, where there is generally less freedom to use new frequencies, although there may be similar SW applications where NVIS is used for domestic radio coverage. In an ideal world these broadcasters would like to be able to transmit a service using single channel simulcast (SCS), so that both the analogue and digital signals are contained wholly within the assigned 9 or 10 kHz channel.
Strictly the term simulcast can be taken to describe the simultaneous transmission of more than one signal carrying the same programme content. In this context it often describes the simultaneous transmission of analogue and digital versions of the same programme from the same transmitter and therefore from a common location. However, it could also mean that only the antenna is common, as well as that both transmitter and antenna are common to the two services. In some cases it could be more economic to add a new lower powered transmitter for the DRM service, feeding the same antenna, rather than making extensive modifications to an older less suitable transmitter, currently carrying the analogue service.
DRM supports a number of different simulcast options. Currently the supported simulcast modes require the use of additional spectrum outside an assigned 9 or 10 kHz channel (Multi-Channel or Multi-frequency Simulcast, MCS). The DRM signal can be located in the next adjacent upper or lower channel and can occupy a half or whole channel depending on the bandwidth option chosen. Significant testing, both in the laboratory and in the field, has been carried out to determine the optimum level of DRM signal needed to provide a good quality DRM service, whilst avoiding significant impact on the continuing analogue service. The conclusion is that a satisfactory compromise can be obtained when the DRM power level is around
14-16 dBs below the adjacent analogue signal. In an ideal world it would also be possible to transmit both an analogue and a digital signal within the same channel (9 or 10 kHz) so that the analogue service could be received, without interference from the digital signal, on any analogue receiver. At the same time the digital service could be received in high quality audio on a digital receiver. However, although promising proposals for a SCS option are currently being evaluated, certain compromises will almost certainly need to be made. Amongst these are likely to be a reduced digital service data rate, which will adversely impact on audio quality, and a reduced service area compared to the analogue service if interference to the analogue service is to be avoided. In the case of the analogue service there is likely to be some impact on the background noise level due to the presence of the digital signal, and the impact is likely to be dependent on the design of the analogue receiver. Nevertheless, there is optimism that most of these problems will be overcome, or significantly reduced, as a result of the ongoing development work.
Even if single channel simulcast may prove a difficult goal to achieve, the other options mentioned above, which require wider bandwidths, can already be implemented. These options will still allow some reduction in transmission equipment investment by allowing the use of the existing antenna and/or transmitter that already carries the current analogue service.
3.4 Service planning
3.4.1 DRM overview
Planning procedures within the AM broadcasting bands below 30 MHz need to be considered in two parts. Within the AM bands contained in the LF and MF part of this spectrum, there are pre-existing regional plans which lay down the fixed assignments or allotments to be used for transmissions by each member country of the ITU. In the HF bands, planning is done on a much more flexible basis, which takes into account the diurnal, seasonal and solar variations in propagation when the allocation of spectrum is determined. In the case of MF and LF spectrum two agreements are in force, the Geneva 1975 Agreement, which covers ITU Regions 1 and 3 and employs a 9 kHz frequency grid, and the Rio Agreements of 1981 and 1988, which cover Region 2 and employ a 10 kHz frequency grid. In the case of HF planning, all three regions use the same frequency grid of 10 kHz and planning, for most countries, is carried out through the auspices of the informal HFCC/ASBU/ABU-HFCC coordination process, with the resultant twice-yearly plan being registered at the ITU by administrations.
188.8.131.52 Regions 1 and 3 – LF and MF planning
Within these two Regions only Region 1 currently has assignments for and uses the LF band. Therefore the majority of assignments for both regions are in the MF band. Under the existing GE75 Plan, existing assignments are listed with their power, antenna details and transmitter location. Any change to this situation, for a particular assignment, requires a recalculation of the transmission parameters to ensure that the protection ratios for other assignments in the Plan, which might be adversely affected by the change, do not deteriorate by more than 0.5 dB. This is also the means by which new assignments have been and can be introduced into the Plan. In September 2002 the ITU Radiocommunication Bureau published Circular Letter CCR/20 under which the RRB with Rules of Procedure to provide the possibility to introduce DRM transmissions into the MF band in Regions 1 and 3 and the LF band in Region 1. Until this issue is agreed by a competent conference the following course of action may be taken by administrations on a provisional basis.
In the case of existing assignments already within the GE75 Plan the ITU-R Letter allows these to be converted to DRM assignments on the basis that they operate with an average DRM power at least 7 dB below that of the currently assigned analogue DSB service carrier power.
In the case of new assignments, which it is proposed should be introduced under the existing GE75 Plan, planning is carried out as if it were to be a new analogue DSB Assignment. If such a new analogue assignment is allowable within the plan, then it may be introduced as a DRM service, provided it is operated at an average power level at least 7 dB below the allowable new analogue assignment.
In both the above cases it is important to note that only DRM Modes A and B using 9 kHz bandwidth are approved for use under this change in the Rules of Procedure.
184.108.40.206 Region 2 – MF planning
The introduction of DRM services in the MF band in Region 2, within the confines of the Rio 1981 (R81) Agreement, is much more problematical. This is due to a stipulation to the effect that § 4.2 of Annex 2 to this Agreement imposes on the classes of emission, other than A3E (that is DSB with full carrier), the condition of being receivable by receivers employing envelope detectors. The later Rio 1988 (R88) Plan, which extends the allowable extent of the MF band in this Region, does not impose such a similar condition. However the ITU RRB did not currently feel able to make a determination for a draft change in the Rules of Procedure for either agreement and so DRM services are not currently envisaged as feasible within the MF band in Region 2. This does not entirely preclude the use of DRM transmissions in this band should an Administration wish to authorise its use within its territory on a non-interference and non-protected basis.
The RRB discussed in its determination the question of whether simulcast systems might be allowable under the R81 plan, as they were receivable on a receiver employing an envelope detector. However the Board expressed concern about the bandwidth requirements of such systems, as they generally required between 20 and 30 kHz of spectrum to accommodate both the analogue DSB signal and the digital counterpart.
Except for a single channel simulcast version of the DRM system (see § 3.3), which was not specified at the time of the RRB’s determination, all other DRM simulcast proposals involve the use of between 20 and
30 kHz of spectrum. In some Region 2 territories such a system option would be potentially allowable within the terms of locally applied spectrum masks with which broadcast services in the MF band must comply. These spectrum masks are generally more relaxed than the ITU–R transmission spectrum mask and envisage lowered but significant levels of energy being radiated up to 10 or 15 kHz away from the assigned channel centre frequency. In such cases the DRM hierarchical transmission modes could be operated in conjunction with an analogue DSB signal to occupy a total of 20 or 30 kHz of spectrum. The analogue signal, at full assigned power, could occupy 10 kHz of spectrum with the base and enhancement DRM transmissions occupying 5 or 10 kHz of spectrum immediately above and below the analogue signal.
220.127.116.11 Regions 1, 2, and 3 – HF bands
Due to the diurnal (day/night-time), seasonal and sun spot related variations in propagation which take place in the SW bands, planning requires that frequency schedules are generally valid for only a six month period. For the majority of international SW broadcasters and operators this requires that intended transmissions are coordinated informally through the HFCC/ASBU/ABU-HFC in order to reduce the potential for interference to a minimum. This procedure is equally being observed for the introduction of DRM transmissions into these bands. Under current coordination procedures DRM transmissions may be introduced under similar principles to that in the MF bands. That is the service is first coordinated as if it were an analogue DSB service and then a DRM transmission substituted with a power level at least 7dB lower than the allowable analogue transmission. The provisional protection ratios adopted during WRC03, for the protection of analogue DSB transmissions from DRM transmissions, show small variations according to DRM mode and modulation. However, in all cases, these variations are smaller than the precision of the propagation prediction tools and can be discounted for the purposes of coordination.
18.104.22.168 The 26 MHz SW/HF band
The 26 MHz broadcasting band allocation is seldom used for traditional short-wave broadcasting. This is due to the frequency being too high for reliable sky-wave propagation during most of the 11-year sunspot cycle in most parts of the world. To a lesser degree, the same is true for the 21 and 19 MHz bands. These bands, particularly the 26 MHz one, could easily be used for DRM broadcasting to a more local audience. Tests in Europe have produced very encouraging results. In the UK tests were part of a local single frequency network of 3 transmitting stations for which the power used was only 10 watts per transmitter. Another test using a single 100-200 W transmitter at a high altitude site close to Geneva showed excellent coverage and quality around the city.
For the line-of-sight services, which are proposed within these bands, Modes A, or B are likely to offer the optimum results. It may sometimes be possible, in some countries and with regulatory approval, to employ the wideband 20 kHz option to improve the audio quality still further. To obtain the best performance from this type of service, it is likely that it will need to be planned in a similar way to an FM service. That is with the antenna at a high level, with respect to the coverage area, and with average powers in the range of 100 to 200 W. It must be recognised, however, that for a period of the sunspot cycle around its maximum, significant interference may be experienced to the local service area. This interference is most likely to be caused by high power international 26 MHz transmissions, as conditions will then make these possible. There may also be interference from other, more local, low powered transmissions, if efforts are not made to minimise sky-wave radiation from them.
22.214.171.124 Near vertical incidence sky-wave (NVIS)
This type of propagation is typically used for in-country SW coverage in tropical zones. The "near vertical" geometry causes multiple reflections between ground and the reflecting ionospheric layers. The result is illustrated in Fig. 27, where several significant reflections are seen to arrive at the receiver antenna. It has been observed during transmissions that at certain times of day, such as dawn and dusk, these reflections can have similar energy and be spread over a period of several milliseconds. In order to prevent destructive interference it is important to ensure that these reflections arrive inside the guard interval otherwise the system will fail.
At the same time as these multiple impulses are observed they can also be subject to high values of Doppler spread. This is due to the constant movement of the reflecting layers and is more significant compared to long path reflections, due to the fact that for NVIS the movement represents a greater proportion of the ground to ionospheric distance. The result of the conjunction of these two phenomena is simultaneously high values of delay and Doppler spread. This can only be overcome by the use of a long guard interval in conjunction with wider frequency spacing for the OFDM carriers. However, because the signal strength can be quite high due to the short paths, signal to noise ratio is often not the limiting factor in NVIS and so 64QAM may be useable for the MSC. Even so, due to the frequent need to use Mode D because of its higher resistance to Doppler and delay spread, the usable data rate of this mode, in a 10 kHz channel, will be quite low. This low data rate may force the use of CELP+SBR audio coding, rather than AAC, unless it is possible to use the 20 kHz wideband option. In this case AAC+SBR becomes possible providing near mono FM, or even stereo quality in good conditions.
126.96.36.199 Single Frequency Networks (SFNs)
Although analogue synchronous networks are often used to provide extended coverage, there will always be problems with mutual interference in at least some parts of the overlap areas. This usually requires the use of additional frequencies to supplement coverage in these areas. With careful design, this problem can be all but eliminated in the case of a DRM SFN. Figure 28 shows a much-simplified arrangement for a DRM SFN, using 6 transmitters. In the area of overlap between areas 1 to 4 it can be seen that signals may be received from all four transmitters at the same time. Provided these signals all arrive within the guard interval they will reinforce each other and reception should be improved in this area over that obtainable from any one transmitter. It is important to note that the transmitted signals must be identical for reinforcement, rather than interference, to occurrence.
Care will need to be taken however to ensure that the network continues to work effectively after dark. Then sky-wave propagation may allow more distant transmitters in the network to contribute signal into the local service area of parts of the SFN. If the propagation path is of sufficient length, and the signal strength is high enough, it may cause interference due to the sky-wave signal being delayed by more than the guard interval. Preventative measures to be taken could include ensuring that sky-wave radiation is minimised by suitable antenna design and changing to a more robust transmission mode, with a longer guard interval, during times of sky-wave propagation.
SFN operation is, in principle, possible using two or more MF or SW transmitters providing service entirely using sky-wave propagation. However the technical requirements are quite onerous, since each of the signals must be timed to arrive simultaneously over the whole of the coverage area. Otherwise they will cause mutual interference rather than reinforcement. This may require real-time monitoring of signals received at several points in the intended coverage area.
Without this, predicting the propagation transition time from transmitter to receivers in the coverage area may prove difficult to achieve sufficiently accurately in advance.
188.8.131.52 Coverage planning
At the time of writing there are no planning tools available which have been specifically designed to calculate coverage and availability for DRM transmissions. However a number of DRM Members plan to rectify this situation by setting up a new project to design software planning tools which takes into account the additional propagation parameter needs of the DRM system. For the moment though, it remains necessary to make a calculation of field strength in the target coverage area based on an analogue DSB transmission. This can then be related to the required signal strength for a DRM transmission using a particular combination of robustness Mode, MSC constellation and code rate to provide the necessary SNR for service. For ground-wave services, this method can be expected to provide results close to observed measurements, as the path is simple, and little, if any, multi-path is introduced to cause signal distortions.
For sky-wave services the prediction is much more complex, as the resultant service will depend not only upon the delivered signal strength but on the level of Doppler and Delay spread to which the signal will be subject. Most software based prediction tools either do not estimate these parameters or, if they do, do not produce reliable results. Nevertheless, for the time being, the existing analogue prediction tools will continue to be used, as they are all that is available. However, it is anticipated that new tools will be developed in the near future, which will aim to provide an estimate of these additional propagation parameters. These tools will be designed to recommend the combination of transmission parameters that best meet the needs of a broadcaster for a specific transmission path and target zone.
In general the average power requirements of a DRM transmission will be less than that of the equivalent analogue transmission. In part this is due to the fact that a DRM transmission will have a higher peak to mean ratio than an analogue DSB signal.
A simple analogue DSB signal will consist of a single carrier at zero modulation whilst at 100% modulation there will be the addition of two sidebands which together will increase the power output of the transmitter to 1.5 times the carrier power. The use of power saving, where the carrier level depends on the modulation level, will modify this relationship, so that the average power output and consumption of the transmitter will be lowered compared to the absence of such a system. Because the DRM signal has a peak to mean power ratio of approximately 10 dB the transmitter must be operated in a backed off condition in order to avoid the digital signal being clipped within the various stages of the transmitter. Should excessive signal clipping occur within the transmitter, it would cause the generation of in channel intermodulation products. These products would cause inter-symbol interference and this can impact adversely on the receiver performance.
An important part of assuring the quality of any radio transmission comes from monitoring the transmitted signals within the target coverage area. In the case of analogue services, this has generally been accomplished by using a high quality receiver for signal reception. The signal strength is then read from a calibrated meter, whilst making a subjective assessment of the audio quality. Such an assessment has historically been made by someone in the target area tuning a receiver to the required service and then listening to it in real time. More recently, this manual method has been supplemented by using unmanned remotely controlled or scheduled receivers to receive the signals and record the signal strength, together with a sample of the audio. The move to using a digital transmission system enables the monitoring of reception to be completely automated. To this end DRM has developed a specification and protocol for the control interface (RSCI). If manufacturers of professional receivers use this specification it will ensure that an operator can use monitoring receivers of more than one manufacturer to build a monitoring network, but use the same software to control and download data from all these receivers. Furthermore this opens the possibility for several operators or broadcasters to share the same receivers, if they so wish.
Because a DRM transmission uses digital coding it facilitates the recording of data that can characterise the reception quality. This information can include not only the signal strength and audio quality, which can be assessed from the audio bit error rate, but also continuous parameters describing the quality and nature of the transmission channel. Over time the accumulation of this information should lead to an improved understanding of the propagation behaviour of the ionosphere.
Data acquired by the monitoring receiver can be stored locally and downloaded from the reception site on a regular basis, to provide evidence of the performance of a particular transmission, or accessed in near real time. In either case the most likely method of transmitting this information back to the broadcaster will be by means of the Internet, or if that is not available, by directly dialling the receiver using a telephone line and modem connection.
In some cases it may be possible to permanently connect to the monitoring receiver(s) either via the Internet, using broadband, or a local network connection or, perhaps, via VSAT terminals. In any of these cases it becomes possible to acquire information about the quality of the service in the target coverage area on a near real-time basis. In this case, by providing a real time method for collating and analysing the reception data, it becomes possible to optimise the transmission parameters of the service(s) in real time. This optimisation process requires the employment of a computer system, which amalgamates the reception data from a number of monitoring receivers in the coverage area. Based on this data, the analysis and prediction algorithm within the computer makes near real-time adjustments to transmission parameters, such as the transmission Mode, MSC modulation and code-rate, to achieve a pre-defined quality of service.
A validation of that concept has been done in the framework of the QoSAM project in 2003 and 2004.