The announcement from BT earlier this month that another 100 local exchanges are to be ADSL enabled has no doubt resulted in plenty of people in previously unblessed areas rubbing their hands with glee and thinking 'broadband at last'.
BT is understandably upbeat about the roll out of broadband: 1,115 exchanges have been upgraded with 250,000 lines connected to those exchanges. The company's target is a million lines by this time next year and five million by 2006.
Some of BT's critics, such as the vociferous Broadband4Britain Campaign, say that's nowhere near fast enough and that BT is deliberately dragging its heals to protect its old ISDN revenues.
Of course, BT refutes this. "Broadband is central to BT's strategy and is the future of the company," says Rebecca Webster, head of marketing for BT Wholesale, the part responsible for upgrading exchanges and selling broadband to ISPs. "We've invested millions in providing DSL services at wholesale and retail and in local loop unbundling. The only way we get return on that investment is by driving volume use."
Broadband not quite broadcast
But how big is that volume? When BT says there are 250,000 lines connected to those upgraded exchanges, does that mean that every customer with a telephone line attached to the upgraded exchange is ADSL enabled?
No. According to some engineers not a million miles from the BT camp, when an exchange is upgraded in some areas as few as eight percent of lines are suitable for ADSL. "Some of the exchanges we've upgraded will never be profitable [because of the low number of users]," one said.
For ADSL to work, the signal available on the phone line has to be strong enough. If the loss of signal is greater than 55dB in the loop from exchange to customer's premises and back again, then ADSL won't work. This is why customers are told they must live within a certain radius of the exchange, because the loss is largely a product of the length of the line.
There is also a problem with the age of the copper wires in some places, say the engineers, inducing further losses. Worse still, some of the lines are aluminium, which is even more sensitive to loss. If the customer is on an aluminium line then the signal will fail if they are more than 750m from the exchange, they estimate.
This is a long way from the official BT line (pardon the pun). Webster says on average 95 percent of the lines attached to an upgraded exchange can be ADSL-enabled.
A loss of 55dB used to translate into roughly 3.5km radius from the exchange. But BT has recently begun using rate adaptation technology, extending the distance to 5.5km, although this does mean a loss of performance in upstream speed.
At 3.5km, the downstream speed (ie the speed at which data comes down from the internet to the customer's PC) is 0.5, 1 or 2Mbps (megabits per second), depending on the service level purchased, and 256Kbps upstream (ie the speed at which data passes from the user's PC back to the internet).
At 5.5km, the upstream speed can drop to as low as 64Kbps.
Old copper and aluminium doesn't reduce this distance, says Webster. The 5.5km is based on worse-case lines. "Where there is a new stretch of copper we can often extend over the 5.5km," she says.
This is fine for urban areas where the density of the population dictates that most people live within 5km of their local exchange.
The B4B Campaign reckons it takes just 50 customers — B4B's so-called Flashpoint — from a local exchange to adopt ADSL to make a profit. B4B says it talked to ADSL equipment suppliers and reviewed several service provider business cases to arrive at this figure.
The economic break-even point for ADSL on a given exchange varies considerably, depending on the class of service provided, the technology deployed, whether the service provider is the incumbent telecoms operator (ie BT) or a competitive operator, and the mix of business and residential tariffs.
The most optimistic business case reviewed by B4B indicated a break-even of just 12 business consumers, according to Catch Communications, a Norwegian operator, using xDSL equipment from Net-to-Net technologies. The most optimistic business case reviewed from the UK indicated a break-even of just 25 business consumers; from Easynet, a UK ISP offering synchronous DSL services to businesses.
In a report, DSL: Business Models for Exploiting the Local Loop, telecom analyst firm Ovum concluded that any incumbent telephone company offering ADSL services over its own network should see returns on any local exchange upgrade should 50 residential consumers take services at £30 per month, or 30 business consumers at £75 per month. Competitive operators using the unbundled local loop and co-locating with the incumbent, whilst using managed 'bitstream' services from the incumbent to backhaul the DSL traffic, would break even at around 75 residential consumers or 40 business consumers.
Naturally, BT disputes the figure of 50. "Neither Ovum or B4B have access to the economics of every exchange," says Webster. "For example, if there is no fibre (backbone link to the internet) at the exchange, then upgrading costs millions rather than hundreds of thousands."
ADSL is not the only technological solution available. In many rural areas too few people live 5.5km from their local exchange. So other broadband technologies have to be deployed, such as one- and two-way satellite-based systems, although these are even more expensive.
While BT has been able to apply for funding from the European Union and UK Regional Development Agencies to finance rural broadband roll-out, in the end infrastructure provision comes down to government policy.
BT, like any other public company, is often torn between its duty to its customers and the need to placate its shareholders; the duty to provide an equitable service and the demand to deliver a fat profit.
Although the government is keen to bang the drum for Broadband Britain, it nevertheless remains ideologically committed to relying on companies such as BT to provide the infrastructure. So while short-term profit is in the driving seat, there will be delays on the line.