Installing fibre optic cables: fast but not soon.
Computers
Fibre moral
by Russell Brown
Don’t count your advanced network chickens until they are hatched.
In the future, there will be bandwidth for everyone: networks that can deliver DVD-quality sound and pictures, for information, entertainment, even telemedicine. You may have read about this, thought about the skinny little copper wire that connects your place to the Internet and wondered exactly how this will come to pass. Fair question.
The news is that New Zealand is beginning to take a first, belated baby step towards that future. Nothing
material has happened yet, but the Ministry of Research, Science and Technology (MORST) recently released a “Discussion Document for a New Zealand Advanced Network”, which advances a proposal in pursuit of a future network.
It won’t be before time. Slovakia, Venezuela, Costa Rica and other countries have parts of an advanced network in place, but we don’t. And addressing our lack is, according the authors of the MORST paper, “imperative to prevent a backward slide” relative to the rest of the developed world.
So, what is all the fuss about? An advanced network in this context is in many ways the same as the fibre-optic networks that already carry our Internet traffic and voice calls around the country and out to the world – the chief difference is that the capacity and, hence, the speed at which it can carry big chunks of data, is vastly greater.
Your ordinary dial-up connection can transfer files at about 50,000 bits per second; your speed-capped Telecom JetStream service at 256,000; and your full-speed JetStream at up to 5,000,000 bits per second. The typical level of service on an advanced network is more like 1,000,000,000 bits (aka a gigabit) per second and it’s anticipated that speeds will reach 40Gbit/s in the new few years.
This doesn’t mean that any single bit travels faster – the laws of physics see to that – just that you can send an awful lot of bits at one time. This is very attractive to researchers (in genetics, for example) who need to exchange very large files – and also, it would seem, to our emerging film post-production industry. The MORST paper considers limiting access to the network to a tightly defined group of users (as has more or less been the case in Britain), but plumps instead for a “widely available” model under which commercial users could become members of the new network from the outset.
Note that word “members”, rather than customers. Under the advanced network model, there would not be a telephone company clipping the ticket in the middle, or setting pricing based on various
maximum data rates. Participants in the new network would, subject to some restrictions on acceptable use, send and receive what they needed, to each other, and to their peers overseas.
The model envisaged would be a “dumb network” – that is, one in which the intelligence rests in devices at its edge, rather than its centre, as is the case with the current, built-for-voice commercial telecommunications infrastructure. It is an array of big data pipes that meet at high-capacity traffic exchange points dubbed “gigapops”. This is essentially the model on which Citylink has built its high-capacity network in central Wellington.
Another potential application for the advanced network is grid computing, which allows processing-intensive tasks, such as genetic sequencing or weather modelling, to be shared among other computers elsewhere on the network that aren’t currently busy. This kind of work is already being done, and it’s quite feasible that researchers around the globe could arrange to share one another’s resources.
It’s all very exciting, until the bill arrives. This kind of development does not suit existing telephone companies, which have to account to their shareholders. Indeed, some people believe that the telco business model militates against these new networks ever being built.
The MORST authors believe that a members’ network would cover its own operating costs, but predict that it will only happen if the New Zealand Government makes a substantial investment in building it – as most other governments have – amounting eventually to about $250 million.
Depending on cost and the development of competition, most of us will eventually take up JetStream-type services over the existing copper network, or new fast wireless services. But none of those will approach the speeds of an advanced fibre-optic network. Telecom will begin to install fibre to homes in the next couple of years, but only where it already has to dig a hole, such as in new subdivisions. The wait for the future may be a long one for most of us.
And yet, there is a strong case for public investment in the advanced network, not only for the benefit of leading-edge researchers and businesses, but also to get the ball rolling. The Internet was pioneered with public investment, via universities, years before the rest of us saw it. If nothing else, it’s a start.
