FRIACO light and darkness (20 November 2000) Doubtless you will be aware of OFTEL's Consultation on Future Interconnection Arrangements for Dial-Up Internet in the United Kingdom. Or, to quote the Daily Mail's front page of 14 November 2000, '24 HOURS A DAY SURFING FOR £10 A MONTH'. The Consultation, which finally mandates a sustainable version of FRIACO, Single Tandem FRIACO, is enormously complex and we provide a summary of its core findings. We shall see about the '£10 A MONTH' charge, and will not be satisfied until services based on Single Tandem FRIACO are shown to work; it will take a few months for any such services to be offered as OFTEL has set the start date for Single Tandem FRIACO to be 1 February 2001. Even more remarkable than the appearance, after several false starts, of a sustainable unmetered service is what is said about the context of OFTEL's decision. OFTEL appointed an independent panel of experts to study the BT network and how it interconnects with other telecommunications operators' networks, something which has almost certainly never been done before, and the panel's report is embedded within the Consultation. Before continuing, a little terminology; the current BT network is a mixture of PSTN (circuit switching, used for voice calls and some segments of data calls) and IP (packet switching, used for the remaining segments of data calls and for xDSL in toto). IP is universally agreed to be more efficient than PSTN as a means for carrying calls: thus most, if not all, telecommunications networks are slowly moving towards being completely based on IP. The salient points made by the Consultation are: The panel found (Annex 1, 2.1-2.6) that the NTS interconnect model, where metered Internet calls are handled as disguised national calls, so traversing large segments of networks, and which we have argued against for years, is disastrous; the PSTN part of BT's mixed network could be overwhelmed within a few years if metered network traffic keeps increasing at the current exponential rate, and that is without considering unmetered traffic. See Annex 1 2.5 and Annex C. BT's mixed network scales badly - in other words, as it gets bigger and bigger there is a point where capacity starts to drop - and, in the medium term, is considered unviable by the panel (Annex 1, 4.11 and Annex B). OFTEL says straight out (4.50-4.62) that it expects a fundamentally different interconnection model than NTS to be developed for IP networks, which is probably the most far-reaching statement of all. But: A solution has been worked out by all parties involved (4.8-4.38) which, in effect, relieves pressure on BT's mixed network until everything can be switched over to IP, a process given three years at most to take place. What BT is doing at the moment to expand its mixed network is described as 'the least desirable solution' of five considered (2.10) and must change. There are several hints (such as 1.1) that it would be in BT's own interest to roll out xDSL as fast as possible to take Internet traffic off its mixed network and straight onto the IP part. All this is gripping stuff - we would have scarcely believed beforehand that a complex technical document could be such a fascinating read - but is also profoundly disturbing; our rather extraordinary conclusion is that OFTEL could have saved BT from itself by, at last, allowing its mixed network to come under a degree of objective scrutiny which found it grossly inadequate and, in the end, based on faulty assumptions about network geometry and topology. We note that the objective language of the Consultation comes dangerously close to slipping several times, and we fully understand why; there are clearly serious technical difficulties with BT's mixed network and doing nothing is not an option. If the alarm had not been raised here, who knows what would have happened in the future? Perhaps BT was hoping that the panel would find that the only practical way to provide sustainable unmetered narrowband Internet access would be BT SurfTime, which would keep BT firmly in control of the growth and direction of Internet-related services in the United Kingdom. It did not.