Q. Dr Paul - I want to deliver video lectures to students at home all over France and have them interact with the lecturer. I’ve followed your earlier advice and I think I can even make money for my university by doing this. But I’ve seen the quality of video over ISDN and it is not nearly good enough. And despite the European talk about fibre, it never seems to happen except in some experiments and the tariffs proposed are far too high for education. What is coming along to solve my problem? Bewildered of Bordeaux
A. Dear Bewildered - Hope is on the way. And you are right, it is nothing much to do with ISDN or fibreing up whole cities. Although the European Commission and much of European industry would like us to believe that fibre optic cable is the only way to deliver good-quality video to homes, it is just not true. Several large US companies including Microsoft are taking key positions in a number of technologies, that will deliver video to the home, none of which include ISDN or fibre.
However, there are generally two main possibilities for delivering good quality video to the home: cable modems and xDSL. Let’s start with cable modems.
The idea behind cable modems is to allow data communications over existing cable TV systems, without requiring a wholesale replacement of the cable TV infrastructure. In reality, some modification of the infrastructure is required – such as for junction boxes – but normally operators can avoid any mass re-laying of cables.
Cable modems in fact can trace their history to the early days of Ethernet, when broadband Ethernets were being developed. These never became popular in the marketplace but the ideas have now been revamped. The original simple ideas about cable modems derived from Ethernet were not much good in reality because the electromagnetic environment within an existing cable TV system is very noisy and there is limited bandwidth available for upstream communications (set-top to head-end). But some expert developers have now got involved, many from backgrounds of experience in radio, cellular or military communications where noise is a fact of life.
The first operational generation of cable modems, such as from Zenith, were of limited throughput - around the 1 megabit/s level - and there was much discussion of how useful they actually were in real cable networks. Since then the speed and the reliability has improved greatly - the current generation of operational modems, such as from Motorola, work at Ethernet speeds - 10 megabit/s. This sounds really fast, but that is ignoring some of the design features of Ethernet.
Ethernet does not give an actual 100% data throughput – the total throughput is only about 30% of 10 megabit/s, that is , 3 megabit/s. This throughput has to be shared between all the users of cable modems on a given cable TV segment, and there might be 100 of these with, say, 10 using it at a time, so one can expect a usable bandwidth downstream (where one needs high bandwidth) of around 300 kilobit/s. That is still pretty fast compared with ISDN, and likely to be a lot cheaper too – when the services are ready, which is not quite yet. But 300 kilobit/s does not allow broadcast-quality video, although it is more than adequate for Internet video-conferencing.
There is another big problem with cable modems for your application – they are restricted to the coverage area of a typical cable TV system, in other words, a city. This does not help you to cover a whole country, or beyond.
Nevertheless, several European R&D projects, mostly in the ACTS area, are planning to use cable modems – including IBCoBN, RENAISSANCE, TOBASCO and CINENET. So are several educational projects including some in the UK "Superhighways for Education" programme.
If you want to cover a country, there is another technology that is relevant. Unfortunately the European telephone operators seem to be keeping it a secret from their customers. (At least in public; in private they are doing a lot more, as you will see.) The reason seems to be that they want to get a return on their investment in ISDN. But the reality is that ISDN is a 20-year old technology and that far higher speeds over copper are now available than are used in ISDN. To see this, just think about Local Area Networks.
For several years, many organisations have been installing Ethernet networks using UTP – Unshielded Twisted Pair, in other words, copper cable – to deliver 10 megabit/s communications, in a wiring layout suspiciously similar to that of a conventional telephone network. And for the last year, vendors have been selling 100 megabit/s Ethernet technology over UTP. Thus it is not surprising that similar developments have been taking place in the Wide Area Network area.
These developments now go under the generic term of xDSL. This term covers a range of technologies delivering speeds from just above ISDN up to around 30 megabit/s. (Yes, three times Ethernet speed over the copper wire coming into your home today.)
Dr Paul is going to break his usual rule of never frightening readers with engineering information in his articles by presenting in table one a summary of the different approaches to xDSL in Europe.
Although there is still much discussion about this, the main variant of xDSL relevant to your needs seems to be ADSL. Engineers believe that only ADSL has the technological and cost parameters to allow the telephone operators of the world to deliver multimedia Internet-based services to the home over their existing copper cabling plant.
There are nearly 600 million telephones in the world served by copper wire. By the end of 1997 it is unlikely that more than 3% of these will be served by fibre or coaxial cable systems. The recent explosion of interest in multimedia Internet access across the world and the resounding lack of interest from consumers in paying for Video on Demand suggests that ADSL could be a winning technology. One of the leading ADSL developers (so treat the figure with caution) predicts that there could be 100 million users of ADSL technology around the world by the year 2000.
Around 40 telephone operators world-wide have now announced plans to test ADSL. However, despite the engineering evidence and the buy-in from the computer industry (Microsoft and the like) not many of these have yet announced plans to provide operational service. In the US, Bell Atlantic is starting trials with ADSL in the Washington DC area. GTE are conducting a small field trial in Dallas, Texas, including at two libraries. Ameritech has conducted field tests with Andersen Consulting to provide ADSL for distance learning in Michigan schools. The world-wide UUNET Internet provider has just announced a trial in Canada of ADSL for Internet access. US West claims that it will start commercial service late in 1996.
One key issue is what the tariffs will be. Much expert opinion predicts a US figure of $50 per month after the first year or two of operation. This is a very cheap price for the equivalent of an 2 megabit/s connection into one’s home! Such prices, if sustainable, should make many telephone operators dump their ISDN plans, especially those who have been late adopters of ISDN.
Despite relatively few public announcements, a surprising number of European telephone operators are interested now in xDSL. The most active include BT (British Telecom), Telecom Italia and PTT Nederlands. Less active ones include the PTOs of Belgium, Denmark, Finland, France, Norway, Spain, and Sweden, and some smaller, innovative operators in UK and Finland.
European companies quietly interested include not just Nokia but Alcatel and Siemens.
The main trials of xDSL in Europe are as follows:
BT’s trial of ADSL in their large Video on Demand trials in the east of England
Helsinki Telephone Company have run a small trial in Helsinki
Telenor have run a small trial (around 30 users) using Westell ADSL modems.
Kingston Communications (in Hull) are planning a trial
Trials will start soon in several cities in Italy
There are rumours of a trial in the Netherlands soon, based on AT&T Paradyne technology.
Close to Europe, Israel Telecom have recently completed a 100-user trial of ADSL.
Several commentators would point to BT having the right parameters to be among the first to deploy an operational ADSL service. For example, they are innovative, highly commercial, understand the Internet and have been cautious over ISDN. One of their Web sites (I will not tell you which) gives some teasing indications. However, many people doubt that they will do anything about ADSL before the UK elections.
I am afraid that I would not put France Telecom among an early adopter of ADSL, but one never knows these days. They are becoming more commercial and should respond to customer pressure. Such as from you and your colleagues. Good luck!
While you are waiting for France Telecom, you can be trying out Video on Demand over your Local Area Network, or perhaps over a wider Intranet, using some friendly local corporate organisation. There is no major problem nowadays in finding a Video Server system to store video and make it available on demand over the Internet – there are several well-known systems such as Xing Streamworks (which is used in the Internet/ATM Video on Demand trials in Finland) and the newer VDOLive system. But you will need, and have to pay a lot of money for, fast and large disc storage!
If you have sufficient Internet bandwidth, you could also try one of the more or less commercial Internet video-conferencing products; such as the venerable CU-SeeMe or a more modern offering; or one of the many quasi-products from research labs.
You will look in vain at present for Web entries on "ADSL in education". Most of the material appears to be unrelieved marketing or technical; but fortunately much of the technical material is written for the relatively non-technical user.
The main sources of information on ADSL are:
The ADSL Forum pages: http://www.sbexpos.com/sbexpos/associations/adsl/home.html which contain a useful tutorial on ASDL and on related technologies.
The Telechoice Report on ADSL, a cornucopia of information of all versions of xDSL: http://www.telechoice.com/xdslnewz/.
Dan Kegel’s ADSL page: http://www.alumni.caltech.edu/~dank/idsn/adsl.html and his related ISDN page.
Analysis maintain a broadband information page: http://www.analysys.co.uk/vlib/bband.htm and there is a constant stream of information on the ACTS INFOWIN pages: http://www.infowin.org for an entry point to this.
Just to prove that ADSL is not just an American scare story, there are now some European tutorials on ADSL – including one from Kimmo Saarela at Tampere University: http://www.cs.tut.fi/tlt/stuff/adsl/pt_adsl.html and another from Norway: http://www.nta.no/telektronikk/4.93.dir/adsl.html.