Why I'm voting for the NBN

I started to write this as I grew progressively more tired listening to arguments against the NBN. The first version (being that I was angry and such) became rather a rant. This, hopefully, is closer to objectivity.

First. Allow me to direct you to Mike Quigley CEO NBN Co's speech from Aug 18, on why it is better to invest $27b rather than spend $6b. I highly commend it. So there would seem to be a few arguments to draw..


Why we need a screamingly fast internet right now
Australia's economy can make a step change from primary and limited secondary industry into high-end tertiary industry. Services. Analysis. Home business for ICT in health, or large scale data management, international teaching exports without moving people, the list is as long as your own imagination. This could easily move Australia from dig-it-up-and-ship-it-over-there to self-sustaining farming and manufacturing, and earning export dollars through our smarts. We need to do this right now for two reasons

1. Many countries in South East Asia (eg. Singapore, South Korea) have no other option but to do this as well - so the competition is already fierce
2. We are reaching a point where growth in economic prosperity through expansion of status quo - ie, more farms, more holes in the ground, will not be possible without irreversible environmental damage. In order to survive we need to compete on new turf. That turf is smarts.

Of course, we can sit back, and hope all goes well. I believe, that if we miss this chance to step ahead of the world, we move down the Argentinian* path. We can choose to limit ourselves to "as we all are now" which means that we lock ourselves in a time capsule, let the most wealthy do as they wish, and the rest of the nation fend for themselves as they reduce to a basic labour force for a commodity export economy.

Or we make the step change. Now.

In order to achieve this step change we need the nation to be ubiquitously connected, and connected at gigabits-per-second speed. Those parts which are not connected will be left behind.

Why wireless now is not an alternative for fibre.
The physics of optic fibre - a glass tube, with electonics at either end - means that the fibre itself has almost unlimited bandwidth. Limiting factors (and hence upgrades) occur at either end of the fibre. So you only have to upgrade the access points, not the whole fibre.

Wireless speeds are (for WiMax) close to 10Mbps for fixed access now. This means that a 12Mbps network could be rolled out with small efforts from existing providers.

The air is shared, fibre is not.
Any bandwidth for wireless is divided by the load, and the distance. This means either lots of towers or lots less bandwidth. For fibre, there is no sharing. One fibre to one house. When there are lots of users, then lots of towers are demanded. However, users don't stay still -- consider children with laptops at school all trying to use the school's one mobile phone tower. Alternately, the children could all share 2 or 3 spare fibres.

Wireless needs fibre, not the other way around
The mobile phone towers which will carry any new wireless technology will need to be placed close enough to the houses to allow the broadband signal -- most broadband wireless has a range of a few tens of kilometers. The towers themselves are connected to a fibre backhaul - because the spectrum is shared. This means fibre to within a few km of every population centre in Australia. The extra length to the house seems a small addition.

Why we do not need to fear "the next generation"
10Mbps vs 1Gbps. How long till the next generation? Typical "next gen" wireless is 10 times faster than the previous generation, and Cooper's law "states that every 30 months the amount of information that can be transmitted over the useful radio spectrum double" That is, if you've bought a 10Mbps wireless network, then after 3x30 = 7 years you've got a new generation of products floating around which are ten times faster.

So, 100Mbps in another 7 years. And 1Gbps, 7 years later. This matches Lawrence G. Roberts' view (Lawrence is credited with initiating the internet). That's when the technology will be invented. Not when it will be rolled out in Australia.

The fibre connections will be growing at a similar rate -- order of magnitude faster every 7 years. So wireless will never catch fibre. If we invest in fibre now, and then do nothing until the most advanced wireless technology looks close, we will have waited nearly 30 years. At that point the single upgrade across the fibre network will be much like the upgrade from copper today.

Why we cannot let the existing telco's build mini-networks (wireless or otherwise)
The electricity network was originally multiple private systems. When the crunch came, one little network failed, and all the others could not cope. A state-based system was needed to meet growth and the allow for the large capital investment in large-scale generation which was more efficient, and more reliable.

A nationalized wholesale fibre network is the same. Without a national wholesale, customers will be forced to pay transfer fees between otherwise disconnected networks (as the data moves between various patchworks). A single inter-connect is need to ensure sufficient investment, and to prevent unfair gouging. Electricity, gas and water utilities operate on the same basis. Whilst it is true that almost always such entities are "sold off" by governments, this is usually only after they have raised substantial return and when the governments are effectively seeking collateral for new works. Relying on subsidized private infrastructure to ensure that the nation's economy can leap-frog our nearest competitors is too great a risk.

For these reasons, and for the fact that such a network represents to greatest chance to see a truly economically and environmentally sustainable Australia, I will vote for the NBN.