Changing satellite techniques and their potential in D-Cinema
Preface
This is an article I offered for publication to the BKSTS, but after a couple of revisions it kind of fell by the way-side and has laid neglected on my storage since. So here it is, an analysis of satellite technology from the viewpoint of its use in D-cinema distribution. I hope it proves useful to someone and that some insight can be gained into the various technologies available.
(copyright Bob Hannent, do not reproduce without permission, yada, yada, yada...)
Technology Introduction
It may be surprising to hear that satellite communications has had very few revolutions in technology since its introduction in the 1960's. The frequency of operation has increased dramatically, the traditional C-Band transmissions at 4 to 8GHz, which dominated the early commercial satellite distribution, was displaced by Ku-band at around 10-18 GHz with the introduction of domestic satellite use in Europe because of the advantages of increased gain (amplification) with frequency for the same antenna size. Permitting much smaller dishes (65cm to 90cm) more suited to the high density populations of Western Europe than the 2 meter C-band dishes seen in rural locations around the world.
The increased ease of production of solid-state microwave components has not only reduced the component cost and yield but contributed to the ability to the development of technologies at further frequencies. With the current digital age, where before microwave circuit fabrication was expensive, it is now fairly modest to build a simple device to provide conversion and amplification at above 10GHz and with scientific applications exploring the region of the EM spectrum approaching the wavelength of light.
The other revolution in satellites was the introduction of digital transmissions for mass-market distribution. The European DVB standard developed with ETSI and the members of the EBU has standardised the way in which we work with digital media distribution worldwide[1]. Very few analogue transmissions remain in Europe and those channels that are still broadcasting analogue usually have a digital equivalent and are already in the planning stages of ceasing analogue services. The BBC's last Ku-band analogue transmission of BBC World to Europe was ceased earlier this year because of the reduced need and viability of an analogue a transmission in Europe. Digital receivers are now at a very modest price, in continental Europe hardware has been available in supermarkets for some years and worldwide often provided free through subscription subsidisation finance models.
Telecommunications has been taking advantage of satellites since they have been in orbit and digital technologies were applied early on because of the relatively high cost of putting a satellite in orbit. The current generation of communications satellite may cost approximately between $150 million and $400 million each depending on: payload complexity, launch and insurance. The worldwide growth in digital markets has reduced production costs of all hardware and even the specialist market of satellites has taken advantage of production volume discounts. Production volumes are obtained through the strong standardisation of signals which has permitted silicon manufacturers to manufacture high volumes of identical processing chips to meet the demand of not only multiple box manufacturers but also across multiple markets. The same receiver circuits can be used in a professional satellite modem as well as in a high quality domestic satellite set-top box. The telecommunications market has also benefited substantially from standardisation which has allowed the use of mass produced components enabling the manufacture of modems that cost less than one 10th the cost of traditional low volume application specific hardware.
The increase in frequency has not only provided benefits in antenna gain, giving smaller hardware, but has also given us increased bandwidth availability through the increased density at the higher harmonics of the EM spectrum. Satellites transmit in what are referred to as 'beams' towards the earth. These beams provide a the coverage or 'footprint' of the satellite. One satellite may have multiple beams within the surface area of the planet onto which it is facing thus increasing the total coverage. The trade-off that increasing frequency has introduced, is that footprints have become narrower. Whereas with C-Band it was relatively common to cover an entire continent or hemisphere with one beam. Now, with the use of Ku-Band the operators prefer to have better coverage in geographically regional or zonal area. Increasing microwave frequencies are more subject to absorption by atmospheric water such as heavy rain clouds. The potential disruption from heavy storms is offset by error correction algorithms which can afford more redundant data with increasing bandwidth availability and the increased gain per antenna diameter at higher frequency.
Another factor which has changed over the past 20 years of digital transmissions, has been the sophistication of signal processing. Originally the extraction of the binary information from the Phase Shift Keyed (PSK) modulated signal may have been done using discrete components. However, today the number of components has been reduced to the point where the signal is simply digitised and all the decoding is done 'in software' by generic Digital Signal Processing (DSP).
Nearly all transmissions made to consumers in the domestic market employ Quadrature PSK modulation which can signal two binary bits per change in phase by having four different phase states to use for the signal. A receiver detects the phase of the signal and determines which binary values match that 'symbol'. Four phase states is only the most common system, the possibilities of soft-decoding have allowed flexible receivers capable of receiving higher densities of transmission encoding. Each of these increases gives more 'bits per symbol' but the increasing density makes the transmissions more sensitive to noise and distortion. Many vendors actually ship receivers which can decode multiple forms of modulation but restrict their capabilities with code keys so that they can sell increased capability after market with no hardware change.
A satellite signal will leave the transmission station ('teleport') through a sizeable satellite dish on earth, travel through the atmosphere, into space for some 35,781km and then the satellite will receive the signal. Satellites are rather dumb devices consisting of simply frequency converters and amplification. The signals it receives are simply converted from the higher end of its range to the lower end of the band (at Ku-band around 14GHz is shifted to around 11GHz) so that the transmissions do to interfere with reception. The new frequency is then amplified and transmitted back to earth to the pre-set footprint. Only a very few broadcast satellites offer processing on-board as this not only adds to the cost, but restricts the potential uses. Dumb satellites have shown they can not only work with analogue and digital but can also handle increasing densities of data with no changes to their systems.
The journey the signal takes is not an untroubled one, the signal is essentially microwave radiation. As it is absorbed well by water, any substantial rain clouds may have significant affect on the signal level. Also, space weather can be a factor, the background radiation and noise of space is not insignificant to a signal which is travelling over such a long distance and often being received by quite small antennae. In satellites signal degradation is expected and the solution to this is to provide an extra data overhead to act as redundancy information. Sophisticated algorithms provide some protection to the signal, often enough to compensate for over the 200dB of signal losses en route. Different and newer error correction patterns are being used more and more to provide even greater protection at higher efficiency for the overhead. One may say, that the signals are more reliable and more efficient than ever.
Based on all the above, the commercial satellite market is approaching its next phase in transmission technologies. There is an increasing interest in providing services at the 'next' logical frequency band of Ka-band, 27GHz – 40GHz. There is also an increasing use of the new European standard 'DVB-S2' standardising the use of higher data densities in a given bandwidth and more efficient error correction systems.
Distribution Challenge
The traditional model for distributing the content for cinema has been packing-cases containing film and with the large volume of footage it has not been uncommon for the distribution reels to be separated or damaged during transit, especially over wide distributions or with specialist market films. With the number of cinemas in the UK, Europe or world we must expect a certain quantity of failure to deliver within the system.
The interest in D-Cinema is growing and the distribution method at the moment is either magnetic or optical disks. Dispatching of hard-disks has long been a risk as they are vulnerable to heat shock and humidity. Optical disks are better at survival but are not immune to these problems. At present densities optical delivery can require more separate disks which requires more handling of the data through production and reading of the multiple disks. Delivery time is also still measured in days rather than hours or better.
The need is to deliver over a wide geographical area, to relatively large structures in diverse locations. Some buildings may be located in areas of low population density some in highly dense areas. New films are delivered on a periodical basis, with typically at most only a couple of new films per week.
Some may argue that it could be possible to take advantage of new terrestrial telecommunications developments such as DSL which are delivering increasing speeds. But DSL service is highly variable and high speeds may not be possible to the out-of-town multiplex or to the small rural cinema, they are also often contended services which do not deliver maximum speed all of the time, only in burst. Also, the cost of a non-contended sustainable reliable line at high data-rates is based on the customer having access to the line all the hours of the day, all the days of the year. Cinema has a need for more a large 'burst' volume of traffic in a very reliable fashion. The traffic for cinemas is essentially broadcast (multicast) in nature, the films are distributed to the cinemas but little or no traffic is required on the return path. Terrestrial communications are essentially point-to-point (unicast) technologies which may employ multicast techniques.
The very nature of satellite is broadcast but earlier attempts at delivery of films over satellite have perhaps been reliant on traditional techniques which were either expensive or too mass market to provide a viable economical delivery. Perhaps with new technologies it is possible to reduce costs, increase delivery speed and maybe improve total delivery reliability. The standard way to deliver 'data' over satellite in a trial scenario follows one of two methodologies, dedicated SCPC or use an existing 'vsat' satellite based network service.
The first solution is to establish a dedicated “Single Carrier Per Channel” (SCPC) transmission which has the advantage of being a clear link of relatively high bandwidth but in the long term it may be expensive to lease sufficient satellite space dedicated to a particular use. Additionally, until very recently the modems which provide high density of data for the bandwidth given were not common and the use of high spectral density transmissions has not been very popular because of the lack of standardisation and cost. Traditional vsat network delivery has been expensive as it has been based on specialised hardware which has not been able to take advantage of the production volume of consumer hardware. Vsat systems offer some quality of service (QoS) but the target market for vsat has traditionally been geographically diverse transactional based traffic and not so well oriented at broadcasting data.
Application of Technologies
What is needed is a solution which takes advantage of the newer techniques available for transmissions, which is able to leverage the volume cost savings of mass consumer production and can broadcast the large data required by d-Cinema securely to a wide geographical area in a timely manner. A satellite broadcast of data which takes advantage of the standardisation afforded by the European DVB-S2 offering gives scope for all of these. Although at present satellite users and operators are not fully exploiting the offerings of DVB-S2 many have experience of increasing data densities and experimenting with new technologies. Increasing the transmission data density gives the fast transfer rates which would be desired by D-Cinema distribution, but traditionally this would have to be offset by increasing error protection to deal with the worst case availability of weather.
This need not be the case with DVB-S2's new standard of 'Adaptive Code Modulation' or ACM. This no longer transmits at a constant level of error protection and data density but it allows the system to adapt using feedback from receive sites through any number of routes: land-line, mobile, the Internet or return channel over satellite (DVB-RCS). The feedback can be used to help the transmission to be maintained through difficult environmental conditions as well as the potential to substantially increase data efficiency during clear-sky conditions. Conventionally the system would be designed to fail because of environmental conditions for some 0.2% of the year (99.8% availability) this could result in more than 17 hours of unavailability per year and even during clear sky conditions the error protection is utilising more than 30% of the transmission to achieve this availability. ACM offers us just marginally better availability on average given that we cannot escape the power of nature. Its significance is that it allows us to better exploit near clear sky conditions, which prevail most of the time, with less waste for unused redundancy information.
DVB-S2 also offers the ability to split the provision of services so that receiving 'terminals', during degradation of link conditions, can select a more resilient transmission from a selection provided. While this provides a good protection in traditional network connectivity the broadcast nature of the D-Cinema system requires taking advantage of delivering the same information to all locations at once. The DVB-S2 documentation on ACM in the majority refers to individual data links with ACM acting for individual stations. There could be little reason not to apply the principles contained within the document to multicast transmissions and ACM feedback be aggregated to provide a managed service.
Many satellite service providers who offer IP data services and particularly those provisioning DVB-RCS (return channel over satellite), offer additional services to assist their users make the most of their satellite link. Most importantly for this application is that many of them offer data casting services. Data casting is where a bulk volume of files or data is provided from the customer and is transmitted by the operator to all its remote locations as a data broadcast.
Most Ka-band capable satellites provide numerous small footprints, or 'spots', instead of larger continental/regional coverage. This affords another advantage to d-Cinema as less than a handful of 'spots' can be used to cover an entire country and the signal need not be broadcast to a whole continent. Additionally, spot coverage reduces the need to cater for continental-wide conditions but instead cope with weather in a much more localised fashion.
The higher frequency of Ka-band means that not only is their more bandwidth but there is also less existing transmission 'traffic' and thus less potential congestion. Security is a secondary benefit of Ka-band. Being only recently used commercially, the availability of dish hardware is at present not in the domestic environment but only the professional. While this may seem to defeat the earlier argument of production volume savings the only limited production items for Ka-band are the antenna systems, as the receiver hardware is the same as other bands and still utilising the volume production. Because of the lack of popularity of Ka-band hardware in the domestic arena there is added security from there being a limited potential for undesirable observers of the signal, even before additional security is added.
Conclusions
The ideal would be that each cinema be provisioned with a terminal which implements DVB-S2 reception, IP data handling and the DVB-RCS (return channel over satellite) systems. While Ku-band is most common, in Europe there is a rise in the availability of Ka-band transponders. Given the potential of this combination for high speed data delivery there will doubtless be operators willing to exploit this. Each terminal would be linked though the satellite and back to the hub teleport. The system would be budgeted for a highly dense data transmission with low level of redundancy which would be able to fall-back to a much more reliable transmission as dictated by the conditions. The focused nature of Ka-band means that delivery can be controlled regionally and the signal could be adapted according to the needs of that region.
The service can achieve reduced costing because these parameters need not be dedicated to this service, thus the service provider can utilise the periods when data is not being transmitted for d-Cinema to service other customers. If appropriate for maximum value cinema content could be delivered as a low priority modest trickle during the day and flood through at night when other commercial customers have significantly less demand. The delivery of the content can be managed by the service provider so this content which needs quick delivery can be charged at peak rate and other content which has a longer time-frame for delivery charged accordingly.
Hardware which previously would have cost tens of thousands of dollars can now be produced for a couple of thousand. The modern satellite market, as now stands, is mature. The rise of terrestrial connectivity has eaten at the traditional markets for satellites and left hardware assets with spare capacity but already depreciated. The satellite business as a whole is by no means in difficulty. Companies are adapting to the challenge of fibre optics and playing to satellite's strengths. Fibre is a point-to-point technology which cannot achieve the geographic diversity of satellite and when the fragile fibre breaks the failure is singularly catastrophic. But satellites are heavily resilient devices built to the highest standards to withstand nearly two decades in the harshest environments. Yet basic maintenance repairs to the terminals in such a system can be done by most professional domestic antenna installers with only minimal additional training. Most satellite operators have plans for re-provisioning a satellite's services in the event it should suffer from the very unlikely occurrence of a catastrophic failure.
Satellite broadcasting is addressing the challenges of the current age. Taking advantage of the global interest in wireless technologies and adapting to the growth in the use of fibre connectivity. No other technology can delivery such high volumes of information to such a wide geographic area and especially with increasing densities of data as well as improved error protection this method of communication can be exploited yet further. The next generation of cinema has much to benefit from the next phase of the satellite business, both could work together to deliver an exciting future for our viewing pleasure.
[1] While the ATSC standard is the US domestic distribution equivalent, often professional ATSC equipment is DVB compatible and most US satellite teleports are happy to work with DVB traffic.