鈥淲HILE you were out today, we called . . . 鈥� Although I have no proof that
alien abduction was responsible for my failing to hear the doorbell, but I offer
up the theory for Popperian falsification. Heaven knows, the phenomenon is
repeatable enough. However, it might be more fruitful to ask why British Telecom
needed to get into the house at all.
I can see the quick answer: BT has replaced the cable leading from the
telegraph pole across the street and wants to lead it through my front window
frame into a junction box. The new wire is all coiled up and waiting on my
windowsill. It鈥檚 not even optical fibre, just a bundle of insulated copper that
Alexander Graham Bell would have spliced up in a trice. But why, in this age, do
telephone companies still go around stringing copper wires from poles?
For an answer, I go to Essex to see a man who knows about these things. Kevin
works at the laboratory that invented optical fibre communications back in 1966.
The lab鈥檚 present owner, Nortel, is ploughing billions of dollars into clever
ways of squeezing more bits per second down them. Optical fibres are hot. The
reason is obvious: the Internet has turned upside down the rules of designing
and operating telephone networks.
Advertisement
The straight growth statistics are impressive enough. As recently as 1980,
less than 5 per cent of all telecoms traffic was computer data. In the
mid-1990s, the volume equalled that of voice calls. By 2000, just 20 months
away, voice calls will make up only a few per cent of total traffic. But more
important, Net traffic behaves in a whole new way. Before the Net, telephone
calls followed the 70:30 rule. Seventy per cent of calls originating on an
exchange would terminate there. Links between exchanges were designed
accordingly.
Remarkably, optical fibre technology is keeping pace, with tricks like
wavelength-division multiplexing through all-optical amplifiers. The current
standard for equipment is 10 gigabits a second. By the end of this year, Kevin
says, they鈥檙e talking about 160 gigabits a second through a single fibre.
It鈥檚 an extraordinary wave of innovation, the 1990s equivalent of Victorian
railway engineering or 1950s nuclear power. Kevin鈥檚 boss puts it another way.
鈥淭here鈥檚 a war on out there, and we鈥檙e making the weapons.鈥�
That鈥檚 all very well. But what about my problem, the gap between the
telegraph pole and my desk? This, apparently, is the 鈥渓ast mile鈥� where distance
has not been killed and older economic calculations hold sway. Nortel has one
hope on the horizon, in tune with the fashion for utilities to diversify into
more than one service. Instead of installing a separate string of copper for Net
connections, it has developed a way of sending data down electricity cables.
(Don鈥檛 try it at home: it involves exciting the gap between the outer insulator
and the three power cables.)
But why stop there? Electricity cables are no more necessary than telephone
cables. The only pipe a household really needs is a water pipe, and that would
make an exemplary electrical conductor. OK, wiring up bath taps to the mains is
not ideal from a safety point of view. But no doubt it would be possible to
mandate insulation standards into the building regulations. And what is a water
pipe if not a large-bore optical fibre鈥攁lbeit one with some problematic
refractive qualities? There鈥檚 no fundamental obstacle to using a stream of water
as a waveguide for laser light.
But why stop there? What鈥檚 to stop us transmitting natural gas down the same
pipe? The technology for pumping liquid and gas simultaneously鈥攖he bane of
DIY central heating systems鈥攈as been proven in the offshore oil industry.
Adapting it for domestic use would be an engineering challenge, but hardly a
conceptual breakthrough.
I explain my scheme for a multiutility company offering heat, light, Net and
cable TV down one convenient fit-and-forget pipe. Kevin looks worried. 鈥淚鈥檇
better give you a lift to the station now.鈥� Have I unwittingly stumbled on a
secret master plan?