Australia needs a wireless broadband network, not expensive fibre-to-the-node: Kohler

Communications Minister Stephen Conroy should think again before demanding that the National Broadband Network be all fibre-to-nodes and that it should be built all at once.

It makes no sense for regional Australia to be covered in expensive fibre. Wireless networks using low frequencies would be perfectly good at a fraction of the cost.

The only reason for turning his back on a wireless solution for regional areas would be because it would make that part of his whiz-bang fibre-to-the-node (FTTN) broadband plan look uncomfortably like that of his predecessor Helen Coonan, which Conroy cancelled at the beginning of April with a satisfying saving of $958 million.

It also makes no sense to fibre-up all the cities immediately – whether Telstra or Terria does it.

ADSL2+ covers 85% of the urban population with up to 24 megabits per second. It will be years before that is too slow.

The point about FTTN is to bring “exchanges” (that is, nodes) closer to customer premises so the copper-based ADSL signals have a shorter distance (the “last mile”) to travel and the data speed does not get eroded too much.

As I understand it, 85% of homes are already close enough for 24Mbps – so why pay for nodes to guarantee them 12Mbps?

The main consequence of that could be the removal of ADSL competition, because Telstra is arguing that there must be full switchover of ADSL services to a monopoly FTTN (which is why, of course, Telstra is very keen to do it).

So to fix a data speed problem for 15% of the population Conroy would do irreparable, permanent damage to telecommunications competition.

As for the bush, there is definitely slow internet due to market failure and it needs Government money to fix it, but that doesn’t require fibre and nodes in country towns.

Fibre is needed to get the data to country towns, but once it is there wireless can deliver it to homes as quickly as anyone needs.

The only question is what spectrum to use. The lower the frequency, the more easily the waves go through buildings. The 5.8MHz that has been used for much wireless data up to now needs an aerial on the roof.

The OPEL consortium of Optus and Elders that Conroy kyboshed in April had been planning to buy the 2.5MHz spectrum owned by Austar for $65 million and use that in country towns. Austar still has that spectrum, and it’s still for sale.

One.Tel’s main asset before it went broke was a national 1.8MHz spectrum, which would also be useful for high-speed data as part of a national broadband network. The liquidator, Steve Sherman, sold it to all the various state rail corporations to be used for digital train radios; as I understand it, these things are still on the drawing boards.

The main problem with regional broadband is not the customer access network in each town, which can be cheaply supplied by wireless spectrum, or even getting it to really remote areas, because they can be supplied with satellite. The problem is getting the data out to the regional centres and back again.

That has to be done with fibre cables that are called “backhaul”. At the moment Telstra is the only firm with a national fibre backhaul network, but that is changing. Soul Telecommunications, a member of the Terria consortium, has built a lot of fibre backhaul, and Optus is also building it for 3G mobile.

Optus says it expects to cover 96% of the population with 3G, including fibre backhaul, by the end of this year, and 98% by the middle of next year.

A few other fibre cables have been laid as well. For example the Victorian Government spent $21.5 million installing a fibre network along fast rail tracks to Geelong, Bendigo, Ballarat and the LaTrobe Valley.

The Melbourne Herald Sun carried an article last week based on Freedom of Information documents showing that after three years only three companies are actually using this fibre, and are paying a total of $85,700 in rent. That comes nowhere near the annual maintenance cost of $700,000, let alone the capital cost.

The Victrack network, as it’s called, contains 30 fibres and is capable of delivering 50,000 TV channels, according to the Herald Sun’s documents.

So it would be pretty stupid to duplicate that network with more fibre backhaul as part of the new national broadband network, and it would be equally stupid to duplicate the existing Optus and Soul fibre elsewhere in the country.

And it would also be stupid, in my view, not to use the Austar wireless spectrum – a snip at $65 million – or perhaps the old One.Tel spectrum if the rail corporations can spare it, to deliver high-speed broadband cheaply to regional cities and towns.

And then once that is all done, and the bush has a faster, more reliable, broadband network than the cities, then the ADSL in the cities could be gradually upgraded to fibre.

By that time, perhaps, 24Mbps will be getting a bit slow, and by then fibre to the premises might be viable.

The most important thing now is for the Communications Minister, Stephen Conroy, to not be seduced by Telstra’s claim to be the only viable builder of a national broadband network, because it will cost $25 billion and be massively complex.

Yes, it would be a nice big announcement for the Minister, or rather, I’m sure, the Prime Minister: Australia’s magnificent broadband future, made possible by the far-seeing Rudd Government.

In fact, it’s mostly built already if you count wireless spectrum and existing backhaul, and finishing it off shouldn’t cost too much if common sense prevails.

This article first appeared on Business Spectator