Commentary on 'Access to Bandwidth - Proposals for Action' The following is our commentary on OFTEL's second consultation document on local loop unbundling, which gives proposals for opening up BT's local loop to competing operators over a two-year timescale. The point of OFTEL's first consultation document, to which the second is a synthesis with proposals for action, was to take soundings on how best to open up ('unbundle') the local loop (the copper wires running from your house or business to the nearest BT exchange) to competition; the local loop is currently owned by BT which has complete control over what can be done with it. Cable operators, in their apportioned fiefdoms, have installed their own local loops: as they currently have only 16% of lines to homes and small businesses OFTEL has left them alone for the moment. Now that large-scale broadband services over copper wires, specifically flavours of xDSL, are technically possible and recognised as vital to the future of the United Kingdom economy the mistakes made with ISDN, for which BT produced a poor service which was too late, too slow and too expensive, could not be repeated. BT wanted what OFTEL called Option 4 in the first consultation paper. What this meant would be that BT would run ADSL over their local loop and give other operators the option to run ADSL in tandem. OFTEL decided that Option 4, although no block to short-term progress as BT are intending to roll out ADSL anyway, was insufficient. They considered that BT would retain control of the technology and thus remain in the lead, in principle, until the next set of regulatory changes; other operators would be forced to follow what BT did first. OFTEL preferred Option 2, which allows other operators to put their equipment on both ends of BT's local loop (in exchanges and in your home or business). This could well be equipment more technically advanced or commercially risky than what BT deploys: for example, flavours of broadband access which provide much higher bandwidth but have a much shorter range. Implementing Option 2 at once implies technical problems. A known shortcoming of broadband access over copper wires, which uses very high frequencies and relatively high signal power compared to telephony over those same wires, is that different flavours of access, which use different frequency ranges to carry data back and forth, could interfere with one another, thus degrading performance. In addition, this 'crosstalk' could result in radio waves of unpredictable frequencies being emitted, possibly disrupting nearby equipment which has nothing to do with telecommunications. As a compromise, OFTEL has proposed that a combination of Option 2 and Option 4 is put in place. With this approach BT has almost two years under Option 4 to get ADSL out as widely as possible without direct competition; on 1 July 2001 Option 2 kicks in and other operators are allowed to run broadband services in competition with BT. Between now and then there are two main activities apart from BT's ADSL rollout: production of a Spectrum Management Plan so that what equipment uses what frequencies, and how the equipment from various suppliers co-exists, can be thrashed out; trials of equipment with a view to having services from competing operators ready to run on 1 July 2001. We believe that this solution, if implemented as written, is as good as could have been come up with at the time. Firstly, the language of Section 2 of the second consultation paper, in particular, demonstrates that OFTEL has listened to what we have to say; the word 'unmetered' is used over and over again. Secondly, the timing is precisely calculated, as BT has as strong an incentive as there could be to get its own broadband access out - the serious commercial risk that competing operators will swamp BT if it does not. Very importantly, broadcast services such as video-on-demand and set-top box access over the local loop start with everyone on an even footing - on 1 July 2001 the current ban on BT offering broadcast services is lifted. Thirdly, there are some very interesting technical points in section 5 of the proposal. What these mean is that the entire frequency spectrum is opened up to competing operators. This means that, as well as competing flavours of ADSL, they could run: basic telephony (which uses low frequencies); data-only lines (as some symmetric variants of xDSL use a very wide frequency spectrum, often including that normally used by voice telephony). That said, we have three main problems with the proposals. Firstly, OFTEL repeatedly asserts that 'demand for advanced services is developing'. From all the feedback we have it is more than developing; it is there, and has been for some time. People are desperate for fast and reliable Internet access with fair pricing. A comment we routinely receive is that data transfer rates, in the first instance, are relatively unimportant: it is the ability to have an unmetered service available at any time that matters. Secondly, how broadband access to rural areas will be made available is fudged with talk of 'virtuous circles'. Because of the technical limitations of xDSL and the limited geographical reach of cable companies it seems that wireless (either fixed radio from operators such as Atlantic Telecommunications and Tele2 or UMTS from mobile operators) will be the only option, but how and when is not clear. Thirdly, our reservations about the probability of wrangles between BT and competing operators stand: such disputes are inevitable with any type of local loop unbundling which leaves the dominant operator still owning the local loop. However, we are generally satisfied with what OFTEL has proposed. The next test will be the pricing of the BT ADSL service, either closely preceded or followed by OFTEL's final ruling on Access to Bandwidth. The first must be affordable; we hope the second will be the contents of this consultation paper, implemented as written. Notes on xDSL Digital Subscriber Line (DSL) technology comes in many flavours: that currently being deployed around the world is usually ADSL, the A standing for asymmetric. What asymmetry means is that the downstream (to you) data transfer speed is higher than that upstream (from you): typically 256Kb/sec to 1Mb/sec downstream and 64Kb/sec to 256Kb/sec upstream. An xDSL connection is always on; there is no need to dial in. Furthermore, the frequencies used by ADSL are much higher than those used by the conventional telephone and do not interfere with it, so you can make a normal phone call and download or upload data at the same time using one line. Symmetric flavours of xDSL may use telephone frequencies as well to transmit data, hence our remark about 'data-only' lines. Faster symmetric flavours of xDSL such as VDSL and HDSL, with equal downstream and upstream data speeds, are already available; the emphasis is on ADSL for the moment because BT has spent about four years procrastinating on deploying it but, under Option 2, there are opportunities for competing operators to leapfrog BT. xDSL is not particularly difficult to implement as it requires no wholesale infrastructure improvements. It is driven by a box at the exchange which connects to the Internet backbone, takes the incoming data stream, breaks it up and sends multiple data streams (typically sixty-four) of high frequency down the copper wire. The xDSL modem attached to your PC, via an Ethernet card, detects the various frequencies of signal coming in and reconstitutes them into a single data stream; as the reverse process is going on at the same time for upstream data transmission the xDSL modem is a powerful piece of kit. A further advantage of xDSL is that, because of the standard Ethernet connection, your PC is configured as though it were attached to an office network. The main, and insoluble, problem with xDSL is that, because high frequencies and relatively high signal power are used to transmit and receive data, it is 'lossy'; the signal only persists for a few kilometres (ADSL) or a kilometre or two (symmetric xDSL) from the exchange before it becomes too degraded to be reconstituted successfully. If you are too far away from an exchange you may not be able to use it; thus our point about rural areas above. Whither Option Zero? The second consultation document doesn't mention our proposals ('Option Zero') written in response to the first, and we didn't expect it to. So why bother writing them? Someone had to 'think the unthinkable' and propose taking the local loop away completely from BT - all of OFTEL's five proposals kept it in BT's possession, thus silently constraining the discussion. As it turns out, about 10% of the responses to the first consultation document mentioned similar solutions to ours, and it's entirely probable that these helped swing the needle towards Option 1 (less BT control) and away from Option 5 (more BT control). Certainly OFTEL's proposals for action were more radical than most, if not all, observers expected. Text by Alastair Scott