Final Report Summary - OPTOSAT (DBS LNB-STB optical fibre transmission link)
Executive summary:
Background and objectives
Satellite television (TV) signals are currently carried from the dish to each viewing location in a residence over coaxial cables. With the advent of satellite receivers with multiple tuners this has become problematic since one cable is required per tuner, resulting in many cables needing to be installed, which is costly, time consuming and disruptive. Optical fibres are an attractive alternative to coax owing to their small size, light weight, very low loss, and price. Systems based on these are already being offered by one of the project partners, however whilst suitable for use in larger systems in multidwelling units (MDUs), they are too costly for use in small scale installations such as single family homes (SFHs). The OPTOSAT project was formed to investigate low cost approaches to satellite TV distribution based on optical fibres, which would meet the requirements for these small scale installations.
Outputs / achievements
Within the OPTOSAT project a prototype demonstrator system, based on low cost optoelectronic components, has been designed, built, and tested. Various options for the system architecture were studied, a key requirement being that the overall system cost be minimised in order to make it suitable for use in the SFH scenario. The use of lasers with different wavelengths (colours) to carry the four satellite bands over a single optical fibre was initially considered, however the design study concluded that, whilst technically feasible, the costs involved were prohibitive owing to the high cost and complexity of the required optical components. An alternative architecture was decided upon, which uses four separate fibres within a single cable assembly, each carrying one of the four bands of satellite TV. The benefit of using separate fibres is that identical lasers sources and optical detectors may be used for each, thus avoiding the cost associated with combining the bands onto a single fibre and separating them in the receiver. Furthermore, the project has concluded that cheap digital vertical-cavity surface-emitting laser (VCSEL) lasers and detectors, together with multimode fibre, which are already mass produced for the datacom market, are suitable for an analogue application such as this. This is a key result as it enables a significant reduction in the overall system cost to be achieved. Additional functionality was added to the system by overlaying an optical Ethernet data network onto the same fibres using wavelength-division multiplexing (WDM) techniques. This was achieved using commercially available equipment which operated at a different wavelength from the TV signals. Prototype transmitter and receiver modules were designed, assembled and tested, and additionally the receiver printed circuit board (PCB) assemblies were integrated into a commercial satellite receiver thus producing a set-top box (STB) with an optical-only input. These units have been used to undertake two separate field trials, one in a domestic dwelling in Spain, and the other at another partner offices which was done so as to emulate a domestic installation. The results from both trials were consistent, and demonstrated the system capability to carry the full set of channels from a single dish to multiple locations within the home as well as providing data connectivity between those locations.
Having demonstrated the capabilities of this approach, interest is being sought from major broadcasters, in order to take the next steps towards developing the concept into a commercially available product.
Consortium members
Global Invacom Ltd, Red Embedded Design Ltd, Electronica Seyma SL, Cube Optics AG, UK-ISRI, University of Kent, Centre National de la Recherche Scientifique, Modulight Incorporated.
Further information
For further information on the OPTOSAT project, please visit the projects main website at: http://www.optosat.com
Project context and objectives:
Project context
Over the past two decades there has been an explosion in the adoption of satellite TV with BSkyB now having an installed customer base of over 10 million in the United Kingdom (UK), and although this market will begin to saturate, upgrades and replacement installations are still expected to run at approximately 1 million per annum for the foreseeable future. The total European market is several times this size with approximately 300 million installed STBs, and where it is common for customers to install their own systems. Where a system involves more than two STBs these systems become complex, typically requiring radio-frequency (RF) switches which are costly and require a degree of expertise to design a workable solution. Given this background, there is a clear demand for a new low cost, simple to install, system and it is believed that a solution based on optical fibre technology is the answer.
A satellite TV installation typically comprises a satellite dish and low-noise block downconverter (LNB) to receive the signals, a coaxial cable network to transport the signals to the required viewing locations, and satellite receivers (STB) to decode the signals, and TVs to view the programme content. The bandwidth of the signals received from the satellite at the dish is approximately 4 GHz, which is beyond the capability of coaxial cables to carry over the distances required in SFHs and MDUs. To overcome this problem, within the LNB the content is split into four separate bands approximately 1 GHz wide and located in the frequency band 1 to 2 GHz. This frequency range is within the capabilities of reasonably price coaxial cables, but does mean that each coaxial cable can only carry one band at a time, which is approximately one quarter of the content being transmitted. In a SFH one cable is run to each viewing location, and signalling from the STB enables the LNB to select the appropriate band and transmit it over the coax to the STB. When a channel on another band is required, the STB signals the requirement to the LNB which then routes the appropriate band over the coaxial cable.
Modern satellite receivers now include multiple tuners enabling one or more program to be recorded whist yet another is viewed. In order to support these extra tuners, extra coaxial cables are required, one per tuner, which can require as many as eight or even more cables, the installation of which is costly, time consuming, unsightly and disruptive.
In MDUs the situation is slightly different, but the issue with the profusion of cables is still the same. Whereas in the SFH where up to typically four viewing locations may need to be supported, in an MDU several tens of apartments will need to have services provided from a single dish and LNB, and there may be several view points in each of these. Clearly many tens of coaxial cables cannot run back to a single LNB, so instead a backbone comprising four coaxial cables, one for each satellite frequency band, is installed. Nodes on this backbone, which comprise a splitter and switch (multi-switch) provide connectivity between the apartments and the backbone. The purpose of this multi-switch is to tap off the signals from each band, and to route them as required into the apartments located in the node's vicinity. The switching capability in the node enables the signals from any one of the four cables (bands) to be routed to a STB in the apartment in exactly the same way as with the LNB in the SFM scenario. Once again, one cable from the node is required for each tuner, meaning that there is again a profusion of cables required to support the multiple tuners in the STBs, and the multiple viewing locations is each apartment.
The issue of multiple cables arises because of the limited bandwidth afforded by the coaxial cables. One of the key features of optical fibres is that they have very high bandwidths and in many instances this is limited by the components (lasers, detectors, etc.) to which they are connected rather than the fibre itself. In addition to this, optical distribution of these satellite signals has the advantage of being easier to install (smaller cables), is electrically isolating thereby eliminating electrical shock hazard and the associated requirement for earth bonding, and is immune to EM interference. Global Invacom has developed a range of fibre based products which allow the distribution of satellite TV signals over a passive optical network, where the signals originate at a single location and are split, typically 32 ways, and distributed throughout a building. To achieve this the four satellite bands are first combined into one composite signal by frequency stacking them in the frequency range 1 to 5.5 GHz. This composite signal then modulates a laser to generate the optical signal which is broadcast over the passive optical network. At each receiving point an optical receiver converts the signal back to RF, and then de-stacks the signal, converting each of the four bands back to their original frequencies, thus providing a STB with the RF signals as if it were connected directly to an LNB.
This system is cost effective in the MDU environment, but owing to the way in which the signals are stacked at the head-end (in the bespoke LNB), and then de-stacked in the receiver in the dwelling, the costs are prohibitive for use in a SFH. The component costs associated with the frequency stacking / de-stacking, which requires high-frequency mixers, phase locked loops (PLLs), etc., make it difficult to reduce the product costs below a certain point.
Given the background described in the previous section, there is a clear requirement for a low cost system for the distribution of satellite TV signals from the receiving dish to multiple viewing locations in the home, capable of supporting multiple tuners in a STB, without requiring a profusion of coaxial cables. The solution needs to be backwards compatible with the installed base of STBs, and it is desirable that it is simple to install, ideally by the homeowner themselves, have low-energy consumption, and be immune to the problems associated with electromagnetic interference. In addition to these considerations account needs to be taken of the moves being made by the major broadcasters to facilitate the material recorded and stored on a STB in one location being played back and viewed elsewhere in the home. This requirement means that it would be beneficial if the system designed provided Ethernet connectivity between the different viewing locations / equipment within the home.
Optical-fibre-based solutions solve many of the problems highlighted, however the main stumbling block has been reducing the cost to the point where the solution is economic for SFHs as well as MDUs. The objective of the OPTOSAT project was to investigate alternative approaches to this, studying different architectures, technologies, and components, and to design and construct a demonstration system.
Objectives
As described in the previous section, the OPTOSAT project was set up to investigate the options for reducing the costs of a fibre based system for the distribution of satellite TV signals within the SFH environment.
The fundamental question that required answering was are there alternative approaches to transmitting the four satellite frequency bands over a cable other than the frequency stacking approach already developed. The project was set up to look into the possibility that this could be accomplished using WDM technology. In this instance optical sources (lasers) with different wavelengths (colours) are used to carry the different bands, and because the lasers are at different wavelengths they can be combined onto the same fibre, and then separated at the receiver, using optical filters. Another approach to be investigated was the use of separate fibres for the four bands, which eliminates the requirement for the WDM filters, thereby reducing costs. Whilst at first sight this may appear to conflict with the requirement that the number of cables be minimise, because of the small size of an optical fibre, many fibres may be included in a single cable assembly, and it is straightforward to manufacture a cable carrying four optical fibres and still keep the cable size significantly less than that of a single coaxial cable.
Aside from the question of the systems architecture the project was also set up to investigate the availability of low cost optoelectronic components suited to the transmission of the satellite TV signals. In particular work packages were included investigating the performance of low cost lasers and detectors, and their suitability for use in this application.
The overall target for the project was, having selected the system architecture and identified the optical components to support the design, to produce prototype transmitter and receiver modules and to use these to assemble a demonstration system which would be used to undertake field trials.
Project results:
Description of the main scientific and technological results
System architecture
As explained in the previous sections, the technical objective of the project was to find a cost effective way of transporting the four 1 GHz bands of satellite TV broadcasts from a head-end unit mounted close to the satellite dish, over a fibre network, to multiple viewing locations in the home. Initially the project investigated the use of WDM techniques to combine the outputs from multiple lasers, each carrying one of the satellite four bands, onto a single optical fibre to transmit them to the viewing locations. Investigations carried out early in the project concluded that suitable low cost lasers for this approach are not currently available and are not expected to be for the foreseeable future. Whilst lasers such as coarse wavelength division multiplexing (CWDM) distributed feedback lasers (DFBs) emitting in the 1 310 and 1 550 nm wavelength bands, and their associated multiplex / demultiplex (MUX / DEMUX) components, could achieve the desired functionality, the cost of such sources alone would exceed USD 400 for a single transmitter module, which is more than an order magnitude higher than the target. Optical sources are available at other wavelengths, for example 850, 665 and potentially 520 nm, however not all of these have the required modulation bandwidth or can be used on the same type of fibre. In addition, multiplexing these particular wavelengths onto a common fibre is not done elsewhere, and so suitable multiplexing components are not commercially available and their development and was considered beyond the scope of this project. Given these considerations the project concluded that the use of WDM was not a viable approach and an alternative methodology was sought.
Space division multiplexing (SDM), rather than wavelength division (WDM), is an alternative means of transmitting parallel data streams albeit over a physically separate path. A quarto-LNB, which has four outputs each dedicated to one of the four satellite bands, is connected an optical transmitter module located close to the LNB. Within this transmitter module the signals from each of the LNB outputs are amplified and used to directly modulate a laser, the output from which is then split and routed to the four viewing locations. In this scheme four fibres, each dedicated to one of the satellite bands, are routed to the view locations, however it is worth noting that these four fibres would be housed within a single cable assembly, which is far more compact than a single coaxial cable. The scheme therefore satisfies the criteria that it eases installation and minimises the number of cables used.
At the viewing location a receiver module, containing a photodetector and amplifier for each of the fibres, converts the signals back to the electrical domain, and feeds them to a STB in a form identical to as if it were connected directly to an LNB. This optical receiving circuitry will eventually be built into the satellite receiver itself, however a separate module will be required initially in order to support the existing installed base of legacy STBs.
The principle benefit with this architecture is that there are fewer constraints when selecting the optical source, enabling it to be on the basis of cost and performance only, without the additional constraint imposed by the wavelength multiplexing requirement. This opened up the possibility of using the optoelectronic components used in datacom applications, which are mass manufactured and are available at relatively low cost. Whilst this appeared an attractive approach to take, there was no evidence in the literature to say whether the analogue performance of these components at the frequencies of interest is suitable for the transmission of the DBS signal format used for satellite TV. A major part of the activity was therefore to investigate the performance of these components for this specific application.
Additional functionality may be added to this architecture by overlaying an Ethernet link. The project envisaged achieving this using the 1000BASE-LX standard which employs 1 310 nm wavelength lasers, the modules for which are commercially available. The overlay can either be achieved by adding two extra fibres, one each for the upstream and downstream traffic, or by using WDM techniques to add the signals to two of the fibres already carrying the satellite TV signals. In addition to providing internet connectivity to each viewing location, this functionality also facilitates communication between the different STBs connected, and enables content recorded or being viewed on one unit to also be viewed on another elsewhere in the home. Within the project the WDM approach was taken to demonstrate this capability.
Optical components and system link design
The availability and performance of lasers and detectors capable of carrying the DVB satellite TV signals with frequencies extending up to 2.15 GHz was investigated. All types of semiconductor lasers were considered, Fabry Perot (FP), DFB, and VCSELs, emitting at the common wavelengths of 850, 1 310, and 1 550 nm. Whilst any of these classes of lasers can be procured with the requisite bandwidth capability, it is the cost requirement which proved to be that which determined the optimum choice. Multimode VCSEL lasers, emitting at 850 nm are cheap to fabricate and package, and transmitter optical subassemblies (TOSAs) which use these VCSELs are manufactured in large numbers for the datacom industry, and are as a consequence the most price competitive. Whilst there was little in the literature to say whether they would have the linearity and noise characteristics required for this analogue transmission application, the decision was taken to pursue these as the primary design path, and emphasis in the project was placed on establishing whether their analogue modulation characteristics were satisfactory. The situation with the optical detector was similar to that of the lasers. Gallium arsenide (GaAs) PIN photodiodes are used in high volumes in the datacom market, are cheap and are readily available as receiver optical subassemblies (ROSAs) usually with an integrated front end transimpedance amplifier (TIA). Again the analogue performance is not publicised since their primary application is in the datacom arena which employs digital modulation, and this was also investigated at an early stage.
Laser evaluation
Multimode VCSELs are readily available, and those which appeared best suited to the OPTOSAT project were those aimed at 2.5 and 4.25 Gb / s digital modulation rates in datacom applications. Both of these variants were characterised, and the figures below illustrate the performance characteristics measured for these parts.
Of particular interest was the analogue modulation bandwidth and linearity characteristics of these devices. The results from the frequency responses measured indicate that there is little difference in the modulation bandwidth of the two parts, and that there is a 2 dB roll off in gain across the frequency band of interest which was considered to be manageable.
As it would be expected from the gain compression characteristics for the two parts, the bias conditions influence the 1 dB input compression point, with better linearity being seen at the higher drive levels. Again little difference was seen between the two variants of this device.
Based on these results the 2.5 Gb / s variant was selected for the prototype transmitter build as its performance was adequate and it was the cheapest option.
Detector evaluation
A similar exercise was undertaken for the detector, which are procured as ROSAs. Although these components may be purchased with or without integrated TIA front end amplifiers, since those which include the TIA are manufactured in much higher volumes they are available at significantly lower cost, and hence priority was given to determining whether they were fit for this application.
Versions designed for 2.5 and 4.25 Gb / s datacom systems were again available and their performance was compared in a similar manner as for the VCSELs. The measured frequency responses for the two variants indicate that that there is only a small advantage in using the more expensive 4.25 G ROSA, and that the frequency roll off across the band with either part is manageable. The lower cost 2.5 G part was therefore selected.
Fibre
The cheap TOSA sources and ROSA detectors selected are designed for use with multimode fibre, and whilst this fibre bandwidth and loss characteristics at 850 nm is inferior to that of a single mode fibre when used with 1310nm sources, the manufacturers data suggested that it would be suitable for use over the 50-metre span lengths required for an installation in a SFH. There was a question as to whether modal noise effects, which degrade performance in some multimode systems, would degrade the signals, however no evidence for this was seen in any of the component or system testing undertaken.
Passive components
Passive optical components, optical splitters and wavelength multiplexers, are also required to realise the OPTOSAT system architecture, including the Ethernet overlay. Specifications for these parts were developed as part of the project, and samples manufactured to this specification by one of the partners, Cube Optics.
Two design variants for the transmitter module were considered. In one of these variants the output from each laser is split four ways to provide the signals to be broadcast to each of the four viewing locations the system has been designed to support. To minimise the power requirements from the lasers and to maximise the system margin, it is desirable that these splitters have low loss and uniform outputs from each of the four ports. The results from the measured characteristics for the splitters manufactured show a mean insertion loss of only 6.8 ± 0.6 dB, and a channel imbalance of ± 0.2 dB, which is more than adequate for this application.
WDM multiplexers capable of combining / separating the satellite TV signals transmitted at a wavelength of 850 nm with the Ethernet traffic carried at 1 310 nm were also built. From the insertion loss and isolation characteristics for these devices the average insertion loss was 1 dB, and the maximum value observed was 1.2 dB. The isolation values show that the worst case optical isolation is 44 dB, which in the electrical domain equates to double this, i.e. 88 dB, which is more than adequate.
System link design
Prior to the construction of the prototype modules and system demonstrator a detailed analysis of the system link budget was undertaken. The analysis was undertaken using a combination of datasheet and measured parameter values for the selected components. The system model constructed was capable of predicting the signal level and carrier-to-noise ratio (CNR) at the output of the system where the satellite receiver (STB) would be connected. The assumptions used in the analysis regarding the system parameters and optical component characteristics are considered to be a realistic worst case scenario.
The analysis predicted that an optical power of > - 17 dBm is required at the receiver in order to achieve a CNR of 11 dB at the system output, which would allow error free reception of all current satellite TV broadcasts. With the assumptions for the optically split system architecture, the received optical power would be - 11 dBm, leaving an optical margin of 6 dB thus demonstrating that robust transmission over the link could be achieved.
The results from the impact of laser noise (RIN) and the drive level applied to the laser indicate that if the laser RIN is below - 122 dB / Hz it results in a system penalty of less than 1 dB penalty. The RF drive level applied to the laser is also a critical factor affecting the system output CNR, with high drive levels giving better performance. This is also shown in the figure, where the system margin is plotted as a function of the RF drive, which is indicated as the optical modulation index (OMI). OMI is the RF modulation applied to the laser expressed as a percentage of its dc bias, and the figure suggests that an OMI of at least 0.18 is required to achieve the 3 dB margin required for reliable transmission. The figure suggests that increasing the OMI further and further will improve the link performance, however this improvement will be limited in practice by the intermodulation distortion that occurs within the laser at high modulation indices, an effect which has not been included in this analysis.
To confirm these link performance predictions a system test was undertaken with a VCSEL laser being driven by the signals from an LNB receiver mounted on a satellite dish. The VCSELs output was connected to a photodiode via a variable optical attenuator, and the output from the photodiode connected to satellite TV meter to monitor the signal quality. The results from the modulation error ratio (MER), a parameter closely related to the CNR - which varied as the RF drive to the laser to the laser was varied - show that there is an optimum range between - 35 and - 25 dBm, below which the MER is degraded owing to the low received signal level, and above which is degraded as a result of intermodulation distortion generated in the laser.
In addition to looking at the effect of the RF drive to the laser, the impact of optical loss and laser RIN were also investigated. Figures 10 and 11 confirm that the minimum received optical power of - 17 dBm predicted theoretically is consistent with experiment. All the results obtained were in reasonable agreement with the theoretical predictions and thus validated the system model.
System modules
Having defined the specifications for the optical components, identified suppliers and undertaken initial link tests confirming the viability of the approach, the design, assembly and testing of the transmitter and receiver modules was undertaken.
Transmitter and receiver modules
Two approaches were considered for the design of the transmitter module. One of these employs RF splitting to provide the RF drive signal to 16 lasers, each of which connects directly to one of the four viewing locations, thereby providing the four bands to the four viewing locations. The alternative approach uses four lasers, with the output of each of these split four ways using a 1 x 4 optical splitter, to produce the required 16 outputs necessary to deliver the four bands to the four viewing locations. The figure shows the transmitter module signal path for just one of the four satellite bands, and so would be replicated four times in order to provide a fully functional module. The RF split version of the module therefore require 16 VCSEL lasers, whereas the optical split version only requires 4 lasers together with four 1 x 4 optical splitters. At the time this work was undertaken it was unclear as to which approach would perform best and offer the lowest cost, however subsequently it was concluded that the cost of the four optical splitters outweighed that of the extra lasers, and the RF splitting approach is now considered to be the favoured option. Prototypes of both designs were fabricated and used in the system demonstrators and field trials.
Two prototype transmitter modules, one using optical splitting, the other RF splitting, and of one receiver module were designed with a view to easing the tasks of modifying the PCBs during the debug and optimisation phase, rather than to minimise their size and cost as would be the case for the final product. There is therefore considerable scope for size reduction through the optimisation of the PCB layout, routing of the optics, and choice of housing, however the units produced were entirely suitable for evaluating the functionality and viability of the concept and the components used. This figure shows a measurement of the system frequency response using the prototype modules connected via a length of multimode fibre. The overall gain is close to unity, so any receiver connected to the system will receive signal levels as if it were connected directly to the LNB. Considerable gain slope is present across the band, and this was subsequently found to be caused by parasitics associated with manner in which the lasers and detectors were mounted on the PCB. This was considerably improved later by mounting these parts in a more optimal manner, and is not considered to be a limitation in the system overall capability.
In total, four transmitters and receiver modules have been fabricated, and these have been used in the system demonstration and field trials described later.
In addition to the standalone transmitter and receiver modules described above, the receiver circuitry was also integrated into a commercially available STB, a twin tuner EchoStar HDS 600RS, thus converting the unit into a fully functioning OPTOSAT system receiver. A figure shows the architecture and photographs of the OPTOSAT PCBs mounted in the STB. The scheme also required the inclusion of an IP media converter to interface the OPTOSAT receivers IP optical output to the STB electrical input. The unit was fully functional, and was used in the system demonstrator and field trial.
System demonstrator and field trials
A demonstration system was assembled in one of the project partner offices, using these prototype units, and emulating what was considered to be a likely installation in a SFH. The set-up comprised a quarto-LNB mounted on a dish, connected to the OPTOSAT transmitter module via four coaxial cables. The actual cable run used in this demonstrator was significantly longer than would normally be the case and some slope compensation was included to allow for the greater attenuation at higher frequencies, which was significant for this cable length. In a practical SFH installation the coaxial cable run to the transmitter module would only be a few metres and slope compensation would not be required.
In addition to viewing the TV channels broadcast from the satellite, a TV meter could be connected at any of the viewing locations, which allowed a quantitative assessment of the of the signal quality. To assess the overall performance of the system the signal quality was first assessed at the satellite dish before transmission over any coaxial cables or the OPTOSAT system. These measurements were then repeated at the viewing locations, and the results compared with those taken at the dish to evaluate the impact of the system on the signal quality. A commonly used figure of merit for the quality of a broadcast digital TV signal is its MER, which is closely related to the CNR and is one of the parameters reported by the TV meter.
The results from the comparison between the MERs measured in zone 1 and those measured directly from the LNB show that there is little difference between the measurements, the exceptions being at the highest frequencies within some of the bands, and this attributed to the roll of in the systems frequency response at these frequencies. As discussed above, this roll off is associated with mounting of the lasers and detectors within the modules and will be straightforward to rectify when developing the final product, and is not considered to be a limitation in the systems capabilities.
From the comparison of the MERs obtained after transmission over 10 and 70 metre lengths of fibre it is expected there is no significant difference in the performance seen, thus demonstrating that the long lengths of fibre may be used without any discernible effect on system performance.
In addition to the testing undertaken on the satellite TV reception, the Ethernet data connectivity over the system was demonstrated by connecting a PC to the OPTOSAT system at one location and using it to access the internet, and also by streaming content from the LaCie Cinema to one of the other zones.
Having assessed the performance of this demonstrator, a field trial was undertaken using these prototype OPTOSAT modules. The system was installed in a domestic residence in Spain, and the overall performance of the system was found to be very similar to that seen in the demonstration set up.
Conclusion
This project has succeeded in identifying a low cost architecture for the distribution of DVB satellite TV signals based on optical fibre technology, using low cost optical components and multimode optical fibre, all of which are mass produced for the datacom market. The project has shown that these components, which are designed for use in digital systems, are also suitable for use in analogue applications such as this, thus enabling a cost reduced solution to be realised. The project has built prototype transmitter and receiver units, also integrating this receiver circuitry into a commercial satellite receiver, and has used these to build a system demonstrator and successfully undertake field trials. In addition to this, the project has demonstrated that the approach can be extended to provide IP data connectivity between the viewing locations in the home, which apart from providing internet access, would also allow play back of material stored on one STB by streaming it over this data connection to another viewing location in the home.
Potential impact:
The potential impact and the main dissemination activities and exploitation of results
The impact and results of the project will be exploited by generating a new range of products based on the results of the research with the product form, function and price being led by market need. It is proposed that these new products are an extension to the existing Global Invacom Fibre MDU Range to cover some of the areas of the market that are not addressed by the current product range.
GIL was the first company in the world to develop a range of low cost Optical Fibre based products for the high volume DBS market, these products have been very successful for the company and have generated significant year on year sales growth and revenues.
In Europe the MDU market only represents around 10 % of the overall DBS market, it is the other 90 % that the OPTOSAT project was aimed at. These new products will utilise components based on the specifications generated within the project in order to generate a means for each of the industrial partners in the project to benefit from the potential supply of these parts. With an annual market size of over 30 million units for Europe alone this presents a major market for the technology generated within the project.
The sales of the finished products by GIL to its existing customer base and the sales of parts to GIL by the consortia partners will generate new sources of revenues for each of the partners and countries involved in the project. These sales will continue on an annual basis for three to four years, before a new generation of products and components are required.
The impact of this new technology on the wider European Community will be to enable a wider deployment of satellite TV technology into new areas. One of the main limitations in the past that has restricted the general deployment of satellite TV has been the complexity of satellite TV installations in homes requiring more than one STB. This often results in most homes only using satellite TV in one viewing location and relying on terrestrial TV for the second viewing location. With OPTOSAT this limitation is reduced as the installation of the second third and fourth box installation is much easier using this new technology than traditional integrated-frequency (IF) switch systems. This will enable countries to transfer more of its TV services to satellite, freeing up the valuable terrestrial spectrum for other new services such as Long-Term Evolution (LTE).
At the end of the project the demonstration system was presented to the team in a major European broadcaster responsible for introducing new technology. The demonstration was successful and well received as being in line with the needs of the broadcaster, as a result of this a follow up demonstration has been requested in order to present the technology to a wider audience, including the engineering team. We are currently in the process of moving the demo system to the Global Invacom office in Stevenage in order to facilitate this demonstration.
List of websites: http://www.optosat.com
Background and objectives
Satellite television (TV) signals are currently carried from the dish to each viewing location in a residence over coaxial cables. With the advent of satellite receivers with multiple tuners this has become problematic since one cable is required per tuner, resulting in many cables needing to be installed, which is costly, time consuming and disruptive. Optical fibres are an attractive alternative to coax owing to their small size, light weight, very low loss, and price. Systems based on these are already being offered by one of the project partners, however whilst suitable for use in larger systems in multidwelling units (MDUs), they are too costly for use in small scale installations such as single family homes (SFHs). The OPTOSAT project was formed to investigate low cost approaches to satellite TV distribution based on optical fibres, which would meet the requirements for these small scale installations.
Outputs / achievements
Within the OPTOSAT project a prototype demonstrator system, based on low cost optoelectronic components, has been designed, built, and tested. Various options for the system architecture were studied, a key requirement being that the overall system cost be minimised in order to make it suitable for use in the SFH scenario. The use of lasers with different wavelengths (colours) to carry the four satellite bands over a single optical fibre was initially considered, however the design study concluded that, whilst technically feasible, the costs involved were prohibitive owing to the high cost and complexity of the required optical components. An alternative architecture was decided upon, which uses four separate fibres within a single cable assembly, each carrying one of the four bands of satellite TV. The benefit of using separate fibres is that identical lasers sources and optical detectors may be used for each, thus avoiding the cost associated with combining the bands onto a single fibre and separating them in the receiver. Furthermore, the project has concluded that cheap digital vertical-cavity surface-emitting laser (VCSEL) lasers and detectors, together with multimode fibre, which are already mass produced for the datacom market, are suitable for an analogue application such as this. This is a key result as it enables a significant reduction in the overall system cost to be achieved. Additional functionality was added to the system by overlaying an optical Ethernet data network onto the same fibres using wavelength-division multiplexing (WDM) techniques. This was achieved using commercially available equipment which operated at a different wavelength from the TV signals. Prototype transmitter and receiver modules were designed, assembled and tested, and additionally the receiver printed circuit board (PCB) assemblies were integrated into a commercial satellite receiver thus producing a set-top box (STB) with an optical-only input. These units have been used to undertake two separate field trials, one in a domestic dwelling in Spain, and the other at another partner offices which was done so as to emulate a domestic installation. The results from both trials were consistent, and demonstrated the system capability to carry the full set of channels from a single dish to multiple locations within the home as well as providing data connectivity between those locations.
Having demonstrated the capabilities of this approach, interest is being sought from major broadcasters, in order to take the next steps towards developing the concept into a commercially available product.
Consortium members
Global Invacom Ltd, Red Embedded Design Ltd, Electronica Seyma SL, Cube Optics AG, UK-ISRI, University of Kent, Centre National de la Recherche Scientifique, Modulight Incorporated.
Further information
For further information on the OPTOSAT project, please visit the projects main website at: http://www.optosat.com
Project context and objectives:
Project context
Over the past two decades there has been an explosion in the adoption of satellite TV with BSkyB now having an installed customer base of over 10 million in the United Kingdom (UK), and although this market will begin to saturate, upgrades and replacement installations are still expected to run at approximately 1 million per annum for the foreseeable future. The total European market is several times this size with approximately 300 million installed STBs, and where it is common for customers to install their own systems. Where a system involves more than two STBs these systems become complex, typically requiring radio-frequency (RF) switches which are costly and require a degree of expertise to design a workable solution. Given this background, there is a clear demand for a new low cost, simple to install, system and it is believed that a solution based on optical fibre technology is the answer.
A satellite TV installation typically comprises a satellite dish and low-noise block downconverter (LNB) to receive the signals, a coaxial cable network to transport the signals to the required viewing locations, and satellite receivers (STB) to decode the signals, and TVs to view the programme content. The bandwidth of the signals received from the satellite at the dish is approximately 4 GHz, which is beyond the capability of coaxial cables to carry over the distances required in SFHs and MDUs. To overcome this problem, within the LNB the content is split into four separate bands approximately 1 GHz wide and located in the frequency band 1 to 2 GHz. This frequency range is within the capabilities of reasonably price coaxial cables, but does mean that each coaxial cable can only carry one band at a time, which is approximately one quarter of the content being transmitted. In a SFH one cable is run to each viewing location, and signalling from the STB enables the LNB to select the appropriate band and transmit it over the coax to the STB. When a channel on another band is required, the STB signals the requirement to the LNB which then routes the appropriate band over the coaxial cable.
Modern satellite receivers now include multiple tuners enabling one or more program to be recorded whist yet another is viewed. In order to support these extra tuners, extra coaxial cables are required, one per tuner, which can require as many as eight or even more cables, the installation of which is costly, time consuming, unsightly and disruptive.
In MDUs the situation is slightly different, but the issue with the profusion of cables is still the same. Whereas in the SFH where up to typically four viewing locations may need to be supported, in an MDU several tens of apartments will need to have services provided from a single dish and LNB, and there may be several view points in each of these. Clearly many tens of coaxial cables cannot run back to a single LNB, so instead a backbone comprising four coaxial cables, one for each satellite frequency band, is installed. Nodes on this backbone, which comprise a splitter and switch (multi-switch) provide connectivity between the apartments and the backbone. The purpose of this multi-switch is to tap off the signals from each band, and to route them as required into the apartments located in the node's vicinity. The switching capability in the node enables the signals from any one of the four cables (bands) to be routed to a STB in the apartment in exactly the same way as with the LNB in the SFM scenario. Once again, one cable from the node is required for each tuner, meaning that there is again a profusion of cables required to support the multiple tuners in the STBs, and the multiple viewing locations is each apartment.
The issue of multiple cables arises because of the limited bandwidth afforded by the coaxial cables. One of the key features of optical fibres is that they have very high bandwidths and in many instances this is limited by the components (lasers, detectors, etc.) to which they are connected rather than the fibre itself. In addition to this, optical distribution of these satellite signals has the advantage of being easier to install (smaller cables), is electrically isolating thereby eliminating electrical shock hazard and the associated requirement for earth bonding, and is immune to EM interference. Global Invacom has developed a range of fibre based products which allow the distribution of satellite TV signals over a passive optical network, where the signals originate at a single location and are split, typically 32 ways, and distributed throughout a building. To achieve this the four satellite bands are first combined into one composite signal by frequency stacking them in the frequency range 1 to 5.5 GHz. This composite signal then modulates a laser to generate the optical signal which is broadcast over the passive optical network. At each receiving point an optical receiver converts the signal back to RF, and then de-stacks the signal, converting each of the four bands back to their original frequencies, thus providing a STB with the RF signals as if it were connected directly to an LNB.
This system is cost effective in the MDU environment, but owing to the way in which the signals are stacked at the head-end (in the bespoke LNB), and then de-stacked in the receiver in the dwelling, the costs are prohibitive for use in a SFH. The component costs associated with the frequency stacking / de-stacking, which requires high-frequency mixers, phase locked loops (PLLs), etc., make it difficult to reduce the product costs below a certain point.
Given the background described in the previous section, there is a clear requirement for a low cost system for the distribution of satellite TV signals from the receiving dish to multiple viewing locations in the home, capable of supporting multiple tuners in a STB, without requiring a profusion of coaxial cables. The solution needs to be backwards compatible with the installed base of STBs, and it is desirable that it is simple to install, ideally by the homeowner themselves, have low-energy consumption, and be immune to the problems associated with electromagnetic interference. In addition to these considerations account needs to be taken of the moves being made by the major broadcasters to facilitate the material recorded and stored on a STB in one location being played back and viewed elsewhere in the home. This requirement means that it would be beneficial if the system designed provided Ethernet connectivity between the different viewing locations / equipment within the home.
Optical-fibre-based solutions solve many of the problems highlighted, however the main stumbling block has been reducing the cost to the point where the solution is economic for SFHs as well as MDUs. The objective of the OPTOSAT project was to investigate alternative approaches to this, studying different architectures, technologies, and components, and to design and construct a demonstration system.
Objectives
As described in the previous section, the OPTOSAT project was set up to investigate the options for reducing the costs of a fibre based system for the distribution of satellite TV signals within the SFH environment.
The fundamental question that required answering was are there alternative approaches to transmitting the four satellite frequency bands over a cable other than the frequency stacking approach already developed. The project was set up to look into the possibility that this could be accomplished using WDM technology. In this instance optical sources (lasers) with different wavelengths (colours) are used to carry the different bands, and because the lasers are at different wavelengths they can be combined onto the same fibre, and then separated at the receiver, using optical filters. Another approach to be investigated was the use of separate fibres for the four bands, which eliminates the requirement for the WDM filters, thereby reducing costs. Whilst at first sight this may appear to conflict with the requirement that the number of cables be minimise, because of the small size of an optical fibre, many fibres may be included in a single cable assembly, and it is straightforward to manufacture a cable carrying four optical fibres and still keep the cable size significantly less than that of a single coaxial cable.
Aside from the question of the systems architecture the project was also set up to investigate the availability of low cost optoelectronic components suited to the transmission of the satellite TV signals. In particular work packages were included investigating the performance of low cost lasers and detectors, and their suitability for use in this application.
The overall target for the project was, having selected the system architecture and identified the optical components to support the design, to produce prototype transmitter and receiver modules and to use these to assemble a demonstration system which would be used to undertake field trials.
Project results:
Description of the main scientific and technological results
System architecture
As explained in the previous sections, the technical objective of the project was to find a cost effective way of transporting the four 1 GHz bands of satellite TV broadcasts from a head-end unit mounted close to the satellite dish, over a fibre network, to multiple viewing locations in the home. Initially the project investigated the use of WDM techniques to combine the outputs from multiple lasers, each carrying one of the satellite four bands, onto a single optical fibre to transmit them to the viewing locations. Investigations carried out early in the project concluded that suitable low cost lasers for this approach are not currently available and are not expected to be for the foreseeable future. Whilst lasers such as coarse wavelength division multiplexing (CWDM) distributed feedback lasers (DFBs) emitting in the 1 310 and 1 550 nm wavelength bands, and their associated multiplex / demultiplex (MUX / DEMUX) components, could achieve the desired functionality, the cost of such sources alone would exceed USD 400 for a single transmitter module, which is more than an order magnitude higher than the target. Optical sources are available at other wavelengths, for example 850, 665 and potentially 520 nm, however not all of these have the required modulation bandwidth or can be used on the same type of fibre. In addition, multiplexing these particular wavelengths onto a common fibre is not done elsewhere, and so suitable multiplexing components are not commercially available and their development and was considered beyond the scope of this project. Given these considerations the project concluded that the use of WDM was not a viable approach and an alternative methodology was sought.
Space division multiplexing (SDM), rather than wavelength division (WDM), is an alternative means of transmitting parallel data streams albeit over a physically separate path. A quarto-LNB, which has four outputs each dedicated to one of the four satellite bands, is connected an optical transmitter module located close to the LNB. Within this transmitter module the signals from each of the LNB outputs are amplified and used to directly modulate a laser, the output from which is then split and routed to the four viewing locations. In this scheme four fibres, each dedicated to one of the satellite bands, are routed to the view locations, however it is worth noting that these four fibres would be housed within a single cable assembly, which is far more compact than a single coaxial cable. The scheme therefore satisfies the criteria that it eases installation and minimises the number of cables used.
At the viewing location a receiver module, containing a photodetector and amplifier for each of the fibres, converts the signals back to the electrical domain, and feeds them to a STB in a form identical to as if it were connected directly to an LNB. This optical receiving circuitry will eventually be built into the satellite receiver itself, however a separate module will be required initially in order to support the existing installed base of legacy STBs.
The principle benefit with this architecture is that there are fewer constraints when selecting the optical source, enabling it to be on the basis of cost and performance only, without the additional constraint imposed by the wavelength multiplexing requirement. This opened up the possibility of using the optoelectronic components used in datacom applications, which are mass manufactured and are available at relatively low cost. Whilst this appeared an attractive approach to take, there was no evidence in the literature to say whether the analogue performance of these components at the frequencies of interest is suitable for the transmission of the DBS signal format used for satellite TV. A major part of the activity was therefore to investigate the performance of these components for this specific application.
Additional functionality may be added to this architecture by overlaying an Ethernet link. The project envisaged achieving this using the 1000BASE-LX standard which employs 1 310 nm wavelength lasers, the modules for which are commercially available. The overlay can either be achieved by adding two extra fibres, one each for the upstream and downstream traffic, or by using WDM techniques to add the signals to two of the fibres already carrying the satellite TV signals. In addition to providing internet connectivity to each viewing location, this functionality also facilitates communication between the different STBs connected, and enables content recorded or being viewed on one unit to also be viewed on another elsewhere in the home. Within the project the WDM approach was taken to demonstrate this capability.
Optical components and system link design
The availability and performance of lasers and detectors capable of carrying the DVB satellite TV signals with frequencies extending up to 2.15 GHz was investigated. All types of semiconductor lasers were considered, Fabry Perot (FP), DFB, and VCSELs, emitting at the common wavelengths of 850, 1 310, and 1 550 nm. Whilst any of these classes of lasers can be procured with the requisite bandwidth capability, it is the cost requirement which proved to be that which determined the optimum choice. Multimode VCSEL lasers, emitting at 850 nm are cheap to fabricate and package, and transmitter optical subassemblies (TOSAs) which use these VCSELs are manufactured in large numbers for the datacom industry, and are as a consequence the most price competitive. Whilst there was little in the literature to say whether they would have the linearity and noise characteristics required for this analogue transmission application, the decision was taken to pursue these as the primary design path, and emphasis in the project was placed on establishing whether their analogue modulation characteristics were satisfactory. The situation with the optical detector was similar to that of the lasers. Gallium arsenide (GaAs) PIN photodiodes are used in high volumes in the datacom market, are cheap and are readily available as receiver optical subassemblies (ROSAs) usually with an integrated front end transimpedance amplifier (TIA). Again the analogue performance is not publicised since their primary application is in the datacom arena which employs digital modulation, and this was also investigated at an early stage.
Laser evaluation
Multimode VCSELs are readily available, and those which appeared best suited to the OPTOSAT project were those aimed at 2.5 and 4.25 Gb / s digital modulation rates in datacom applications. Both of these variants were characterised, and the figures below illustrate the performance characteristics measured for these parts.
Of particular interest was the analogue modulation bandwidth and linearity characteristics of these devices. The results from the frequency responses measured indicate that there is little difference in the modulation bandwidth of the two parts, and that there is a 2 dB roll off in gain across the frequency band of interest which was considered to be manageable.
As it would be expected from the gain compression characteristics for the two parts, the bias conditions influence the 1 dB input compression point, with better linearity being seen at the higher drive levels. Again little difference was seen between the two variants of this device.
Based on these results the 2.5 Gb / s variant was selected for the prototype transmitter build as its performance was adequate and it was the cheapest option.
Detector evaluation
A similar exercise was undertaken for the detector, which are procured as ROSAs. Although these components may be purchased with or without integrated TIA front end amplifiers, since those which include the TIA are manufactured in much higher volumes they are available at significantly lower cost, and hence priority was given to determining whether they were fit for this application.
Versions designed for 2.5 and 4.25 Gb / s datacom systems were again available and their performance was compared in a similar manner as for the VCSELs. The measured frequency responses for the two variants indicate that that there is only a small advantage in using the more expensive 4.25 G ROSA, and that the frequency roll off across the band with either part is manageable. The lower cost 2.5 G part was therefore selected.
Fibre
The cheap TOSA sources and ROSA detectors selected are designed for use with multimode fibre, and whilst this fibre bandwidth and loss characteristics at 850 nm is inferior to that of a single mode fibre when used with 1310nm sources, the manufacturers data suggested that it would be suitable for use over the 50-metre span lengths required for an installation in a SFH. There was a question as to whether modal noise effects, which degrade performance in some multimode systems, would degrade the signals, however no evidence for this was seen in any of the component or system testing undertaken.
Passive components
Passive optical components, optical splitters and wavelength multiplexers, are also required to realise the OPTOSAT system architecture, including the Ethernet overlay. Specifications for these parts were developed as part of the project, and samples manufactured to this specification by one of the partners, Cube Optics.
Two design variants for the transmitter module were considered. In one of these variants the output from each laser is split four ways to provide the signals to be broadcast to each of the four viewing locations the system has been designed to support. To minimise the power requirements from the lasers and to maximise the system margin, it is desirable that these splitters have low loss and uniform outputs from each of the four ports. The results from the measured characteristics for the splitters manufactured show a mean insertion loss of only 6.8 ± 0.6 dB, and a channel imbalance of ± 0.2 dB, which is more than adequate for this application.
WDM multiplexers capable of combining / separating the satellite TV signals transmitted at a wavelength of 850 nm with the Ethernet traffic carried at 1 310 nm were also built. From the insertion loss and isolation characteristics for these devices the average insertion loss was 1 dB, and the maximum value observed was 1.2 dB. The isolation values show that the worst case optical isolation is 44 dB, which in the electrical domain equates to double this, i.e. 88 dB, which is more than adequate.
System link design
Prior to the construction of the prototype modules and system demonstrator a detailed analysis of the system link budget was undertaken. The analysis was undertaken using a combination of datasheet and measured parameter values for the selected components. The system model constructed was capable of predicting the signal level and carrier-to-noise ratio (CNR) at the output of the system where the satellite receiver (STB) would be connected. The assumptions used in the analysis regarding the system parameters and optical component characteristics are considered to be a realistic worst case scenario.
The analysis predicted that an optical power of > - 17 dBm is required at the receiver in order to achieve a CNR of 11 dB at the system output, which would allow error free reception of all current satellite TV broadcasts. With the assumptions for the optically split system architecture, the received optical power would be - 11 dBm, leaving an optical margin of 6 dB thus demonstrating that robust transmission over the link could be achieved.
The results from the impact of laser noise (RIN) and the drive level applied to the laser indicate that if the laser RIN is below - 122 dB / Hz it results in a system penalty of less than 1 dB penalty. The RF drive level applied to the laser is also a critical factor affecting the system output CNR, with high drive levels giving better performance. This is also shown in the figure, where the system margin is plotted as a function of the RF drive, which is indicated as the optical modulation index (OMI). OMI is the RF modulation applied to the laser expressed as a percentage of its dc bias, and the figure suggests that an OMI of at least 0.18 is required to achieve the 3 dB margin required for reliable transmission. The figure suggests that increasing the OMI further and further will improve the link performance, however this improvement will be limited in practice by the intermodulation distortion that occurs within the laser at high modulation indices, an effect which has not been included in this analysis.
To confirm these link performance predictions a system test was undertaken with a VCSEL laser being driven by the signals from an LNB receiver mounted on a satellite dish. The VCSELs output was connected to a photodiode via a variable optical attenuator, and the output from the photodiode connected to satellite TV meter to monitor the signal quality. The results from the modulation error ratio (MER), a parameter closely related to the CNR - which varied as the RF drive to the laser to the laser was varied - show that there is an optimum range between - 35 and - 25 dBm, below which the MER is degraded owing to the low received signal level, and above which is degraded as a result of intermodulation distortion generated in the laser.
In addition to looking at the effect of the RF drive to the laser, the impact of optical loss and laser RIN were also investigated. Figures 10 and 11 confirm that the minimum received optical power of - 17 dBm predicted theoretically is consistent with experiment. All the results obtained were in reasonable agreement with the theoretical predictions and thus validated the system model.
System modules
Having defined the specifications for the optical components, identified suppliers and undertaken initial link tests confirming the viability of the approach, the design, assembly and testing of the transmitter and receiver modules was undertaken.
Transmitter and receiver modules
Two approaches were considered for the design of the transmitter module. One of these employs RF splitting to provide the RF drive signal to 16 lasers, each of which connects directly to one of the four viewing locations, thereby providing the four bands to the four viewing locations. The alternative approach uses four lasers, with the output of each of these split four ways using a 1 x 4 optical splitter, to produce the required 16 outputs necessary to deliver the four bands to the four viewing locations. The figure shows the transmitter module signal path for just one of the four satellite bands, and so would be replicated four times in order to provide a fully functional module. The RF split version of the module therefore require 16 VCSEL lasers, whereas the optical split version only requires 4 lasers together with four 1 x 4 optical splitters. At the time this work was undertaken it was unclear as to which approach would perform best and offer the lowest cost, however subsequently it was concluded that the cost of the four optical splitters outweighed that of the extra lasers, and the RF splitting approach is now considered to be the favoured option. Prototypes of both designs were fabricated and used in the system demonstrators and field trials.
Two prototype transmitter modules, one using optical splitting, the other RF splitting, and of one receiver module were designed with a view to easing the tasks of modifying the PCBs during the debug and optimisation phase, rather than to minimise their size and cost as would be the case for the final product. There is therefore considerable scope for size reduction through the optimisation of the PCB layout, routing of the optics, and choice of housing, however the units produced were entirely suitable for evaluating the functionality and viability of the concept and the components used. This figure shows a measurement of the system frequency response using the prototype modules connected via a length of multimode fibre. The overall gain is close to unity, so any receiver connected to the system will receive signal levels as if it were connected directly to the LNB. Considerable gain slope is present across the band, and this was subsequently found to be caused by parasitics associated with manner in which the lasers and detectors were mounted on the PCB. This was considerably improved later by mounting these parts in a more optimal manner, and is not considered to be a limitation in the system overall capability.
In total, four transmitters and receiver modules have been fabricated, and these have been used in the system demonstration and field trials described later.
In addition to the standalone transmitter and receiver modules described above, the receiver circuitry was also integrated into a commercially available STB, a twin tuner EchoStar HDS 600RS, thus converting the unit into a fully functioning OPTOSAT system receiver. A figure shows the architecture and photographs of the OPTOSAT PCBs mounted in the STB. The scheme also required the inclusion of an IP media converter to interface the OPTOSAT receivers IP optical output to the STB electrical input. The unit was fully functional, and was used in the system demonstrator and field trial.
System demonstrator and field trials
A demonstration system was assembled in one of the project partner offices, using these prototype units, and emulating what was considered to be a likely installation in a SFH. The set-up comprised a quarto-LNB mounted on a dish, connected to the OPTOSAT transmitter module via four coaxial cables. The actual cable run used in this demonstrator was significantly longer than would normally be the case and some slope compensation was included to allow for the greater attenuation at higher frequencies, which was significant for this cable length. In a practical SFH installation the coaxial cable run to the transmitter module would only be a few metres and slope compensation would not be required.
In addition to viewing the TV channels broadcast from the satellite, a TV meter could be connected at any of the viewing locations, which allowed a quantitative assessment of the of the signal quality. To assess the overall performance of the system the signal quality was first assessed at the satellite dish before transmission over any coaxial cables or the OPTOSAT system. These measurements were then repeated at the viewing locations, and the results compared with those taken at the dish to evaluate the impact of the system on the signal quality. A commonly used figure of merit for the quality of a broadcast digital TV signal is its MER, which is closely related to the CNR and is one of the parameters reported by the TV meter.
The results from the comparison between the MERs measured in zone 1 and those measured directly from the LNB show that there is little difference between the measurements, the exceptions being at the highest frequencies within some of the bands, and this attributed to the roll of in the systems frequency response at these frequencies. As discussed above, this roll off is associated with mounting of the lasers and detectors within the modules and will be straightforward to rectify when developing the final product, and is not considered to be a limitation in the systems capabilities.
From the comparison of the MERs obtained after transmission over 10 and 70 metre lengths of fibre it is expected there is no significant difference in the performance seen, thus demonstrating that the long lengths of fibre may be used without any discernible effect on system performance.
In addition to the testing undertaken on the satellite TV reception, the Ethernet data connectivity over the system was demonstrated by connecting a PC to the OPTOSAT system at one location and using it to access the internet, and also by streaming content from the LaCie Cinema to one of the other zones.
Having assessed the performance of this demonstrator, a field trial was undertaken using these prototype OPTOSAT modules. The system was installed in a domestic residence in Spain, and the overall performance of the system was found to be very similar to that seen in the demonstration set up.
Conclusion
This project has succeeded in identifying a low cost architecture for the distribution of DVB satellite TV signals based on optical fibre technology, using low cost optical components and multimode optical fibre, all of which are mass produced for the datacom market. The project has shown that these components, which are designed for use in digital systems, are also suitable for use in analogue applications such as this, thus enabling a cost reduced solution to be realised. The project has built prototype transmitter and receiver units, also integrating this receiver circuitry into a commercial satellite receiver, and has used these to build a system demonstrator and successfully undertake field trials. In addition to this, the project has demonstrated that the approach can be extended to provide IP data connectivity between the viewing locations in the home, which apart from providing internet access, would also allow play back of material stored on one STB by streaming it over this data connection to another viewing location in the home.
Potential impact:
The potential impact and the main dissemination activities and exploitation of results
The impact and results of the project will be exploited by generating a new range of products based on the results of the research with the product form, function and price being led by market need. It is proposed that these new products are an extension to the existing Global Invacom Fibre MDU Range to cover some of the areas of the market that are not addressed by the current product range.
GIL was the first company in the world to develop a range of low cost Optical Fibre based products for the high volume DBS market, these products have been very successful for the company and have generated significant year on year sales growth and revenues.
In Europe the MDU market only represents around 10 % of the overall DBS market, it is the other 90 % that the OPTOSAT project was aimed at. These new products will utilise components based on the specifications generated within the project in order to generate a means for each of the industrial partners in the project to benefit from the potential supply of these parts. With an annual market size of over 30 million units for Europe alone this presents a major market for the technology generated within the project.
The sales of the finished products by GIL to its existing customer base and the sales of parts to GIL by the consortia partners will generate new sources of revenues for each of the partners and countries involved in the project. These sales will continue on an annual basis for three to four years, before a new generation of products and components are required.
The impact of this new technology on the wider European Community will be to enable a wider deployment of satellite TV technology into new areas. One of the main limitations in the past that has restricted the general deployment of satellite TV has been the complexity of satellite TV installations in homes requiring more than one STB. This often results in most homes only using satellite TV in one viewing location and relying on terrestrial TV for the second viewing location. With OPTOSAT this limitation is reduced as the installation of the second third and fourth box installation is much easier using this new technology than traditional integrated-frequency (IF) switch systems. This will enable countries to transfer more of its TV services to satellite, freeing up the valuable terrestrial spectrum for other new services such as Long-Term Evolution (LTE).
At the end of the project the demonstration system was presented to the team in a major European broadcaster responsible for introducing new technology. The demonstration was successful and well received as being in line with the needs of the broadcaster, as a result of this a follow up demonstration has been requested in order to present the technology to a wider audience, including the engineering team. We are currently in the process of moving the demo system to the Global Invacom office in Stevenage in order to facilitate this demonstration.
List of websites: http://www.optosat.com
