Video-on-demand trials abound but even those companies taking part say the lack of decent communications into most people’s homes is the biggest barrier facing them. In the US, 94% of homes are passed by televison cable and 23 in every 100 people subscribe, but in the UK by 1993 only 12% of homes were passed by cable and one in 100 people subscribed. It could easily be 10 years before the UK has similar coverage to the US. Philip Martin, business development manager at Concurrent Computer Corp believes the first people to take up the service will be those that use cable television already. If people in the UK are not used to cable services they could be less inclined to take cable-based video-on-demand services.
Minor changes
The actual cost to the subscriber of the link from the cable in the street into the home in the UK is not high – around ú10 or ú20. If the subscriber already has cable no extra wiring is necessary into the home for video-on-demand services, except maybe minor changes, such as a new coaxial cable from the interface box to the set-top, all within the home. Another way of sending data into the home is down standard copper telephone cable using a technology called ADSL Asymmetric Digital Subscriber Loop. This is one approach being used by the group carrying out the British Telecommunications Plc video-on-demand trial in Ipswich, Suffolk. It will also test fibre optic links as an alternative to copper. ADSL is nice because it can run data directly from the server to the set-top box across existing phone lines without needing to dig up the road to lay cable. In addition, in the UK many more people have access to telephone lines than cable. But the bandwidth using ADSL is limited at the moment to some 6Mbps to 8Mbps. One expert believes that figure is not likely to increase and the technology is probably already hitting its bandwidth limit. The Cambridge, UK-based Interactive TV Trial is using Asynchronous Transfer Mode digital data packet switching to send the data into the home down copper cable. The Cambridge Trial is going to run a variety of bit rates of Asynchronous Transfer Mode. At the moment it is looking to run 2Mbps, although it is possible to run up to 25Mbps of bandwidth down the existing cable. The 2Mbps rate is enough to support one stream of data into the home to one set-top box. If more than one person wants to watch videos on demand into the same home, then an additional 2Mbps will be needed for each set-top box. The possibility of doing this is something those working on the Cambridge trial will look into at a later date. The cost of providing the cable infrastructure in the UK is very high. Hugo Davenport, chairman of Cambridge Cable Ltd, which is providing the cable link service for the 250 homes in the trial at Cambridge, said if you can’t provide a service people can afford, you haven’t got a business. The problem is, at present video-on-demand costs thousands of pounds per subscriber. Cambridge Cable is having to lay a mini fibre network that can carry up to 100 Mbits per second of data bandwidth as part of the trial in addition to the existing copper cable that carries up to 25Mbits per second of data. The fibre network carries the data from the home to the kerbside. A coaxial cable then takes the signal into the home, because it is not economically viable to take fibre right into the home.
By Abigail Waraker
But he said the network costs will come down rapidly within the next five years to something quite economically viable so service providers will be able to start reducing charges and this will increase demand. Charges will also be affected by the cost of sending the data to the home. British Telecom defines the cost per stream as the cost of sending the information down the line to the customer. Martin argued that a cost per stream of about ú1,000 is essential to be competitive. British Telecom is providing the service into the home at the Ipswich, Suffolk video-on-demand trial due to start this summer. The company is keeping the fig
ure for its cost per stream under wraps. It said the equipment is not yet at production level and economies of scale will com e in as more customers are linked up, so the cost per stream will come down. It has been estimated by those outside British Telecom that its cost per stream could be something like ú5,000 at this stage. Part of the goal of the Ipswich trial is to determine how the cost per stream will be paid for – whether by the customer or service provider. Methods of payment in the Ipswich trial are to be announced shortly. The Cambridge trial is planning to charge people using ‘Monopoly money’ for the time being. The Cambridge consortium is still in discussion to determine the charging policy. With so many people jumping on the video-on-demand bandwaggon, discussions about standards becomes important, otherwise the spread of the service might be limited. Hewlett-Packard Co is very optimistic that the current lack of standards will not be a problem and is participating in DAVIC. DAVIC, the Digital Audio Visual Interoperability Council, is a voluntary group set up last March to establish standards for video- and near-video- on-demand services from the server end right through to the set-top box in the home. It has 160 member companies including the likes of Digital Equipment Corp, although most support has come from the cable and telephone industry. This is the group that is going to do it. There are enough people that want a solution. We are continuing to work on it, said Al Kovalick, vice-chairman of the DAVIC management committee and principal engineer at Hewlett-Packard.
Commercial servicee
It hopes to have DAVIC 1.0 ready by December. The standard aims to encompass the technology that already exists and define a server interface to that technology. In particular standards are needed for the Remote Procedure Call software layer between the set-top and the server. This RPC standard is particularly difficult because it is not realistic to have lots of memory on the set-top box because of the cost, and because there are so many RPC options to chose from. A second issue is the need for a common interface to all set-top boxes. A common applications programming interface to the set-top box is also need to keep independence from the different operating systems that they use. Faced with these challenges, it might seem that video-on-demand is a long way off in the future. But one consortium has decided to take the bull by the horns and go ahead with a commercial service without a prior trial. DEC argues that trials cost as much as the real deployment of a video-on-demand service, so deployment makes more sense. It feels the interactive television market is well known enough to install without trials. Chicago, Illinois-based regional phone company Ameritech Corp awarded a $475m contract jointly to DEC, Scientific Atlanta Inc and ADC Telecommunications Inc in March. Ameritech plans to provide the service to 300,000 homes in its area this year and another 1m homes a year to the year 2001, making a total of 6m homes. Whether the DEC is right and customers will make use of the service remains be be seen.
