Gary Miller of BT Openreach outlined the activities that his company was employing to satisfy the government's Broadband Delivery UK (BDUK) programme.
He explained that there had been an EU target to deliver broadband to 100% of the population by 2013 and a 30 Mps service by 2020, with 50% of the population receiving 100Mbs. Clearly the topography of the various European countries and their population densities vary and this determines the most economic method of delivery and the ease of roll-out. As an example he compared the ‘feudal’ English village based around a central crossroad surrounded by fields with the French equivalent with each homestead standing in its own fields. Unfortunately while this made it easier for BT to provide a high-speed service to an English village in did present some difficulties for the speaker in defining ‘rural’ to the satisfaction of ‘not-spot’ customers in the audience! Indeed he fielded an exceptional large barrage of questions but was rarely stumped for an answer.
BT Openreach has been awarded most of the contracts under BDUK programme. This was possibly because the broadband infrastructure had to be made available to all Internet Service Providers (ISP) and as BT Openreach was specifically created as a wholesaler of services it was better fitted than most.
The purpose of the BDUK was to provide funding in areas where there were no commercial plans to improve the service. The aim was to fund the ‘gap’ between the payback of a commercially viable service and that of the rural services.
At the moment the commonest method of delivering fixed-link broadband is using Asymmetric Digital Subscriber Line (ADSL), [Asymmetric – downlink is faster than uplink], over the copper wires used for telephone service from the exchange to the home. Often the cables from the exchange are connected via distribution cabinets at the side of the road, (painted green, but often grey in Bath). New FTTCs (Fibre To The Cabinet) were being fitted within 50 m of the existing cabinets. These cabinets are linked by a fibre optic, (actually a bundle of four fibres), to the exchange and linked by copper wires to the standard distribution cabinet. The new cabinets contained the electronics and a UPS (Uninterruptible Power Supply). The intention is that the new cabinets remain locked while patching in the existing cabinet can be carried out when the ‘fibre’ service is enabled. At the present time some 5,600 exchanges are equipped to feed 90,000 cabinets potentially supplying 26 million customers.
In some instances a FTTP (Fibre To The Premises) service is offered where customers need a very fast connection or where there are many customers in close proximity, such as in a block of flats.
The speaker was questioned as to why FTTC seemed to be BT’s preferred solution, wouldn’t FTTP be ‘future proof’? It was explained that it was very much a matter of economics, running fibre to thousands of distribution cabinets was a lot cheaper than running it to millions of homes. Even so the payback period for FTTC was of the order of ten years and that of FTTP would be expected to be very much longer.
The cost of the FTTC programme was estimated at between £5-£6B, while a FTTP solution would cost around £25-30B. Most of this money had to be obtained on the commercial market as government was only putting forward £0.5-1B. There had to be a commercial judgement as to the premium that a customer was prepared to pay for a faster service. A working estimate had been that only 20% of customers would take up the service at the present prices, fortunately this appeared to be an underestimate, perhaps nearer to 30%. This meant that now a £129M return on capital would be available for re-investment.
Currently the plan agreed with government and the local authorities was to deliver fast broadband to 90% of customers and this had nearly been achieved. The next stage was to achieve 95% by 2017. Some people have suggested that the UK was lagging behind other countries in Europe, but this was disputed, in fact the UK attributed a contribution of some £3.6 billion of GDP to the use of broadband.
For parts of the country where fixed links were difficult or not economic it was possible to use satellite or mobile telephone links. Unfortunately both these methods, while capable of providing high peak data rates, were unable to meet the high continuous demands of gaming or video streaming that customers were increasingly demanding. Sometimes it was possible to use a microwave link to a FTTC, which viable as long as there were sufficient customers close to the cabinet. Even within densely populated areas there were places that were not easy to serve. These included areas such as London’s docklands, where cable ducts went around docks that no longer existed and more modern areas where cables run directly from premises to the exchange. In the later case new intermediate distribution cabinets were being installed where practicable.
There were some questions raised as to what service level people actually required, this had been addressed by a modelling process taking account of some 13 demographic groupings, 4 type of user and 3 classes of TV user, giving 156 household types.
Finally there was a brief description of proposal for even faster broadband (800Mps), (G.fast), to be achieved by fitting new equipment in the fibre cabinets or in mini-cabinets between the fibre cabinet and the home. This however would only benefit those already getting the better services.
All told this was a marathon performance that at times became more of a dialogue with the audience as questions and answers flew fast and furiously!
It remarkable to realise what has been achieved in such a short time and how a network designed for telephony (circa. 4kHz bandwidth) has been able to deliver data rates of 80Mbs and above. I do though wonder why there should be an expectation that super-fast broadband should be available everywhere – if one chooses to live ‘off the data grid’ there are consequences, just as living far away from the motorway network must restrict transport options. These are the reasons, after all, for the historical move of people from the country to the city. Also why do businesses ‘in the country’ need huge data rates? The bulk of the data would be better and safer on a remote server with the link to the business premises being restricted to email and SQL transactions. Why indeed does anyone ‘need’ these high-speed networks? There was a time when people wrote Nobel prize-winning papers on computers with 360k operating systems! Not that long ago I was involved in a huge data migration project. We were shocked to find some users had single-page text documents that were 10Mb in size. This was because they had included a very high-resolution copy of the organisation’s logo on every page. The logo wasn’t even a requirement, ‘it just looked nice’. A single drunken evening can generate more ephemeral embarrassing data, (eternally retained by Facebook), than a company used to create in a year. Surely we can’t continue to abuse this marvellous facility indefinitely? Some of us will recall that Professor Ian Bitterlin touched on this during his talk ‘Green Data Centres’ in 2013. He has kindly provided an updated version here (PowerPoint) and the original here (7Mb Adobe Acrobat).
Thinkbroadband
