Tuesday, June 17, 2008

Tuesday Column: Why Senator Conroy should take a look at Canberra's Black Mountain Tower

Black Mountain Tower is a monument - but not of the type its creators intended.

Do you ever think about why it’s there? There’s not much reason now.

But back in the early 1970s microwave transmission towers were vital. They were the main means by which telephone calls and television programs were relayed between Sydney and Melbourne. Canberra was an essential stop along the way.

The planners at what is now Telecom could see the need for microwave transmission facilities increasing rapidly and so, thinking big, they planned a monster tower – a design statement that would rise high above the national capital and provide for all of the intercity links that Australia would need.

It was an expensive bet on what seemed at the time to be the technology of the future. (So expensive that a restaurant and viewing platform were included to help fund it.)

Within years of its opening in 1980 it was out of date. Telecom Australia and its predecessors in the Post Master General’s Department had made the wrong bet...

All of Australia’s television operators began sending their programs around the country by satellite. The iconic tower, which was designed to be an essential part of the process of distributing television programs, instead suffered the indignity of having a small satellite-receiving dish placed next to it so that it could receive them for rebroadcasting to a local audience.

And then the laying of huge amounts of optical fibre cable between Sydney and Melbourne made the telephone transmission part of its role irrelevant as well.

Black Mountain Tower stands as a monument to the futility of building really big for the future - one the present Communications Minister Steven Conroy would be well advised to ponder.

The “richest” of the Rudd government Ministers, Senator Conroy has been given the right to spend $4.7 billion of our money building a national fibre-to-the-node broadband network, something he cheerfully says will “rival the Snowy Mountains Scheme in its scale and significance”.

Submissions from companies wanting to help build the network are due by the middle of next week.

Although the Minister has set up a panel of experts to help him assess the submissions he has been quite specific about the technology he wants.

It has to be fibre-to-the-node or fibre-to-the-premises. This means that an optical fibre has to go to within a few hundred metres of each house or office, with the remaining metres able to be covered by the existing copper phone lines.

He has specified a minimum download speed of 12 megabits per second, which is about 12 times what most of us could get a year ago, and said he wants 98 per cent of Australian homes and businesses reached.

Oh, and it all has to be done within five years.

What could possibly go wrong?

A lot.

For a start the ground could move from under him, just as if did from under the planners of Telecom’s Black Mountain Tower.

In fact it is almost certain to. Back in the 1970s and 1980s technological developments weren’t happening that quickly. Right now they are raining down upon us at lightening pace.

Anyone who specifies a particular technology, as the Minister did when he launched Labor’s $4.7 billion fibre-to-the-node promise in March last year, is highly likely to find that it is superceded or made redundant by the time it is built.

Oddly, it looks as if the main competitor to fibre-to-the-node will be copper itself.

The existing wires that connect most of our houses to the exchanges are proving far more useful than ever would have been thought. Three or so years ago they could only carry half a megabit per second (which we used to call broadband). Now they are easily doing 2 megabits and up to 12, and there is talk of them soon carrying 50 – which is far more than is needed for television. It is happening because ever more ingenious ways are being thought up to use radio waves to send vast amounts of data from one end of a copper wire to another and to correctly decode them at the other end.

The technology isn’t a new type of cable. (Copper is turning out to be incredibly versatile.) It is a new way of using the cables we have.

And by comparison with Conroy’s $4.7 billion spending plans it is cheap.

It may be that one day we will reach the limits on the use of the copper that already connects our homes and that one day we will seriously need something faster.

But neither has happened yet. And with technology evolving as quickly it is there is an excellent case for waiting. Anyone considering buying any piece of equipment right now is well advised to wait if they can. It is sound financial advice.

In a report for the Committee for the Economic Development of Australia last year Professor Joshua Gans referred to it as the “late adopter” strategy.

By contrast, the Communications Minister has a “build it and they will come” strategy. His speeches continually refer to future uses for the planned $4.7 billion high-speed broadband network that he says will evolve after it is in place.

Are the Senator’s plans making you feel uneasy? They might if you had heard the Governor of the Reserve Bank Glenn Stevens speak in Melbourne on Friday.

He told a business audience that in order to make room for some of the big infrastructure investments on the horizon he might have to push up interest rates.

As he put it, “in most economies, it is usually not possible, and certainly not prudent, to try to have a consumption boom at the same time as an investment boom”.

The Senator’s $4.7 billion spending program - to be matched or more than matched by a private partner - could be worse than redundant. It could push up inflation and interest rates.

Think of what’s involved. Rewiring the cables going into nine million houses would require enough skilled workers to disconnect and reconnect cables at footpath pillar boxes at the rate of 5,000 per day - every day for five years.

And that’s only part of it. In terms of the strain it would put on our resources it would indeed rival the Snowy Mountains Scheme. But the result might be a more dangerous version of the Black Mountain Tower.

Andrew Leigh writes:

Public policy by revolving restaurant

Peter Martin has a lovely column in today’s Canberra Times, using a metaphor that will appeal most to his local audience. Pointing out that the Telstra Tower was rendered unnecessary within a few years of its construction, he asks whether Labor’s plan of fibre to the node really is smart public policy.

This reminds me of the Woodrow Wilson School at Princeton, which spent hundreds of thousands of dollars in 2000 outfitting each seat in its classrooms with an ethernet port. A year later, wireless took over, and they issued all their students with a free wireless card. A year after that, all new laptops were wireless capable, and they stopped issuing cards. So far as I know, the ethernet outlets were utilised for about nine months.

Joshua Gans writes:

Unplanned obsolescence

Peter Martin has a terrific piece about Canberra’s Black Mountain Tower — an expensive piece of infrastructure found technologically obsolete within a couple of years as satellite killed the microwave star. It now pretty much a glorified lookout. Martin was drawing an analogy with current broadband plans. The whole example really resonated and towards the end of the article I realised why:

In a report for the Committee for the Economic Development of Australia last year Professor Joshua Gans referred to it as the “late adopter” strategy.

By contrast, the Communications Minister has a “build it and they will come” strategy. His speeches continually refer to future uses for the planned $4.7 billion high-speed broadband network that he says will evolve after the network is in place.

One little correction, believe it or not my report was written and released in 2006. Somewhat amazingly it still doesn’t seem to be obsolete.


Neerav Bhatt said...

"Now they [copper lines] are easily doing 2 megabits and up to 12"

Peter, currently available retail ADSL+ copper line services in Australia are capable of download speeds of 24 Megabits/sec for people who live within 500m-1km of their phone exchange

source: http://www.internode.on.net/residential/internet/home_adsl/extreme/

Anonymous said...

I live near my exchange and only 7Km from the GPo but I cannot get ADSL at home: Telstra is using pair gain at my exchange!

Run the optic fibre!

MkeM said...

neerav is absolutely right. I have clocked our ADSL2+ link at better than 12 Mb/s when downloading software, although there are very few web sites capable of delivering that speed.

I'm inclined to think that downplaying ADSL2+ and pushing fibre to the node was a deliberate strategy by Telstra to replace the copper network with one where they could charge competitors whatever they liked and squeeze them out of business.

And you don't need to run fibre everywhere to get rid of the idiotic crippling pair gain technology. Fibre to the node makes sense in built-up areas that are too far from an exchange to get ADSL2+ over copper. In more dispersed areas wireless is showing growing potential.

Andos said...

This is a very interesting article, Peter. Dealing with current inflation and interest rates while trying to remove the 'infrastructure bottle-necks' that we face is one of the biggest problems facing the Federal and State Governments.

I think that you're interpretation is a tad pessimistic, though, with regards to the possible obsolescence of new fibre optic cable.

If we can create noise filtering and compression software/hardware to get speeds up to 50 MB/s on copper, imagine the speeds we could achieve on comparatively noise-less fibre optic. I can't see fibre-optic being replaced as a means of data transmission soon (although I'm could be wrong).

Think about the applications requiring GB/s of bandwidth over the next few decades; medical, educational, entertainment. I doubt copper could meet this better than fiber-optic.

Anonymous said...

@Andos - the applications that need > 50mbps are not yet important for everyday consumers and not likely to be in the near term; so let's wait and see. This is Telstra / media company boosterism.

@Peter and Neerav: 24mbps is the theroetical maximum. The ACCC has stopped ISPs advertising that amount as they can't deliver. But you are right that Peter has understated current ADSL2+ speeds which in reality deliver something like 10 - 16 maybe even 20 mbps if they are within a few km of the exchange. But that just adds to Peter;s argument.

@Anonymous & Peter - if you take Peter's argument seriously then we should not build fibre to the node where there is existing ADSL2+ and should only now invest in bringing up to speed the 45% or so of the population who are not getting that through a combination of fibre in outer suburban, some rural towns and of course new estates and wireless and satellite elsewhere.

mnot said...

Thanks, Peter. I've worked in the Internet industry for years, but for the life of me cannot figure out why this is necessary, beyond political posturing and making somebody's mates rich. Neither can anyone I know (who isn't working for one of those putting their hat in the ring).

The infrastructure we should be investing in -- public transports, intercity highways and the like -- aren't sexy enough for us to care about.

Anonymous said...

Well, I hope they can make better use of my phone line in the future. I live 3.5 KM from my local exchange, but my phone line is so poor quality (I have been told by Telstra it is the lowest grade copper they ever used) that I can't get high speed ADSL. When I applied for ADSL 4 years ago I was told it would not work on my line. So I had to get cable internet that is sent down the Foxtel cable. Of course this limits my ISP options to 1 - Telstra Bigpond. This means I am subject to monopoly pricing - hence cable internet is far more expensive than ADSL. A year ago I got in touch with an ISP to see if it was now possible for me to get ADSL. They said that it was thanks to improvements in the technology. However, since my line was so poor, the fastest connection they could _guarantee_ was 512K, which is about 20% of the speed I get from cable. I hope utilisation of the phone line improves at the rate you suggest it will. But over the last 4 years the technology for my phone has gone from a speed of 0 to a speed of 512K. My cable connection is fast, but there is no competition, so it effectively costs double what an equivalent ADSL plan would cost. Why can't we have a system where any ISP can have access to the Foxtel cable for people that can't get ADSL? Surely that would push prices down.

Letter said...

Letter to Editor June 20, 2008

Peter Martin (''Bets on the future are risky'', June 17, p11) on Black Mountain Tower is probably too young to rememberthe real reason for the tower.

It replaced two ugly orange and black smaller steel towers one for ABC and the other for commercial TV.

Each tower interfered with the signal of the other, causing black spots.

The higher Telecom tower carried all the antennas for ABC and commercial television and was to provide better signals to the developing southern and northern suburbs.

The hills in the south caused multiple ghosting images on TVs and there were constant complaints about the quality of the signal.

Eventually, Canberra grew too big and repeater stations had to be built on various hills.

For those people who still get ''free to air'' TV, Black Mountain Tower is essential.

The microwave dishes on Black Mountain Tower were incidental to the main purpose of TV transmission.

Personally, I would rather have signals in the ultra-high frequency than signals in the microwave frequency whose health effects are yet to be determined.

David Roberts, Dickson

JM said...

12Mbit? A phrase comes to mind - "poverty of ambition"

Japan (and I think Korea as well) already have 40Mbit and are upgrading to 100Mbit using fibre to the premises. The fibre is already there, they're just upgrading the termination equipment.

As rapidly as copper gets faster, fibre will always be faster still.

Fibre also has an economic advantage as its bandwidth is intrinsically much higher than copper in all circumstances. The major cost with both copper and fibre is pulling the cable, not the material cost (although copper is likely to become more expensive over time).

For any given cable run, the cost of pulling copper and fibre is pretty much the same. But the capacity of fibre is much, much greater.

Fibre wins no matter how much performance is squeezed out of the buggy whip of copper.

Marek said...

Do new housing estate's get connected with copper or fibre?

Letter said...

Letter to Editor, June 20, 2008

Peter Martin's opinion piece (''Bets on the future are risky'', June 17, p1) and David Roberts's subsequent letter (June20) don't tell the entire story.

Canberra Television (CTC7) began building its original studio and transmitterfacility on Black Mountain in September 1961. They began test transmissions on April 7, 1962. The ABC also had its studio and transmitter on the mountain, each station having its own guyed lattice steel towers.

The proposal for a communications tower was first made by CTC. As there were technical issues affecting both broadcasters caused by their proximity to each other.

Due in part to political reasons, the tower was eventually built by the Postmaster General's Department. It subsequently changed to Telecomand then Telstra. Thus it was built both for telecommunications and broadcasting.

As a broadcast facility it is successful: there are five analogue and five digital TV services. In addition there are four community, two commercialand five national FM services. There are also mobile phone and many private data services.

Certainly, with the closing down of Telstra's microwave network, the facility is probably not an asset for Telstra. But that doesn't mean that the tower is no longer useful.

Rodger Bean, Watson

Letter said...

Letter to Editor, June 20, 2008

Peter Martin's exposition on the risks of investing too early in major projects, with particular reference to a national broadband system, is worthy of the debate that, unfortunately, it may not provoke (''Bets on the future are risky'', June 17, p11).

Here are a few hopefully contentious comments. Should governments invest only in projects that are risk-free? Mostly, but not always.

Broadband in Australia also demands government involvement since commercial firms will otherwise shun the bush. And parity of service in rural areas with urban areas will only be assured if government dictates what has to be done everywhere.

Is the Government right in specifying a national structure comprising, one, fibre to the node (optical fibre cables going to a telephone exchange or to street junction boxes, with copper wire connections thereafter), and, two, fibre to individual premises, both with minimumdownload speed of 12 megabits per second?

At present Martin would have them do nothing, just in case what they do today might turn out to be a white elephant tomorrow. But many will not disagree with their proceeding as they are.

I regard the specified speed as adequate for most purposes but in sparsely populated areas I would have the nodes as terminal boxes with wireless transmission therefrom.

When is it ''too early'' to take the plunge? No one can say. Martin's risk will always be there no matter how long you wait.

I'll be hot-footing it to the end of the next rainbow when copper wires are shown to support 50 megabits per second on a multi-subscriber service at 10km range.

Jack Lonergan, Isaacs

Post a Comment