Chorus, the infrastructure arm of Telecom, is working hard on shortening the local copper loop. Around the country, people are starting to see the large green Whisper cabinets made by Eaton in Christchurch in the streets. Chorus announced on May 3 that the cabinet installation process was half-way through, with 1800 of the planned 3600 cabinets activated.
The cabinets are being rolled out as part of a $1.4 billion programme that forms part of Telecom’s legally binding Undertakings to the government; and the promise is better broadband. However, as the government is presently working on a fibre to the premises broadband rollout, Telecom is undoubtedly nervous about its huge investment being left stranded in the wake.
It is no surprise then that Telecom has gone on the publicity offensive to boost the fortunes of the cabinetisation programme. Analysts and even TelstraClear CEO Dr Alan Freeth have said the cabinets are the natural way to deliver the government’s ultra-fast broadband to customers.
And, it is true that the cabinets represent the state of art in copper-based broadband technology; what is more, they are not “vapour ware” and will be rolled out by the end of next year. As well, the government’s UFB is still in the planning stage.
That it is too soon to write off copper as a physical medium for broadband distribution was demonstrated by Alcatel-Lucent recently. The giant telco industry supplier is presently boasting about lab experiments that it says have achieved 300Mbit/s speeds over 400 metres and 100Mbit/s a kilometre away.
This is achieved by bonding several lines together, as Ericsson did last year, and push 500Mbit/s through with six copper pairs forming one broadband conduit.
Unfortunately, these reports on Alcatel-Lucent and Ericsson’s achievements focus only on the fast headline speeds and not on whether it is practical to use several copper pairs to reach, with some difficulty, speeds that fibre-optic cable can do with a single strand.
There are also questions around how successfully increased signal interference and power requirements that come with bonded DSL can be managed, and customer premises equipment that supports multiple lines is still rare.
Government goals specify that the UFB should provide people with 100/50Mbit/s dedicated bandwidth. DSL today is typically shared bandwidth, with customers getting a slice of what is available at any given time. It can be a very small slice too, meaning your multi-megabits per second connection crawls along when many users are active. Cabinetisation provides multiplexed or shared bandwidth and the question is if Telecom would provision sufficient amounts to keep up with future demands from users wanting to exchange increasingly large amounts of data.
Telecom hasn’t yet explained how it intends to provide voice service from the cabinets either. Presently, telephone lines are multiplexed in the cabinets over its own bearer in the cabinets, and switched at the exchange on first-generation digital equipment that is decades old. This is hardly the NGN Telecom promised the government.
Although Voice over IP is a possibility with the cabinets, Telecom has little capability in this area, surprisingly enough despite being a telco. Instead, Telecom works with WorldxChange, which specialises in VoIP, but the collaboration so far extends to a few greenfield fibre deployments only.
Following our regulator’s decision, access seekers can service customers from cabinets, as they can from exchanges. However, unbundled cabinets present a tougher business case than unbundled exchanges.
While Telecom has made space available inside the Eaton cabinets, and also through a plinth that can fit a secondary enclosure, the roadside units serve hundreds of customers and not thousands like the exchanges do.
As the telco business is a low-margin game with high capital expenditure it follows that large customer volumes are required to succeed. Your average DSL customer makes $60 dollars a year or less for a provider. With that amount of money involved investing in sub-loop unbundling, with backhaul as well as business and operations systems, is challenging, to put it mildly.
With all these limitations, it is hard to see how sub-local loop unbundling could be as attractive as exchange-based unbundling. According to Chorus, out of 1.8 million copper lines from the exchanges, a mere 47,000 have been unbundled.
That the cabinets make business tougher for competitors could ultimately doom the programme. The government is adamant that it wants a competitive telco market, without repeating past mistakes that have left New Zealand with a single dominant player requiring regulation to rein it in.
Unless Telecom can come up with a proposition for Chorus that enables competition in the cabinets, it is doubtful the government will listen and instead prefer to start from scratch with a fibre to the premises network.