@Shadow Minister: NOT as you claim: "itsy bitsy data allowance with their shiny new iPhone's, and these are counted as broadband connections."

The constant growth you are harping on about is in connections. That would be Mobile Phone Handset data bundles, which grew by 66% in one year. They downloaded an average of 4029 [TBytes/q] / 3.1 [Musers] users / 3 [month/q] = 433 MBytes/month. You are right in saying they are well on their way to exceeding fixed line users. But the characterisation of their downloads as itsy bitsy sounds about right.

There are also the mobile broadband users. These are the ones you are claiming are going to undermine the NBN. We don't know how they are growing in numbers (last years column has "n/a"), but we know their data usage grew in total by 19% - ie not much. Their downloads were 16990 [TBytes/q] / 3.7 [Musers] / 3 [mo/q] = 1.5 [MBytes/month].

The bottom line is for all your bluster, nowhere on the planet is wireless displacing fixed line - even in your precious US. Yet that is what must happen for the NBN to not make money. Surely you must concede it isn't the easiest position for others accept.

@Shadow Minister: Secondly, Which private sources are funding the $23bn

I don't have a clue. Maybe next years NBNCo annual report might say?

Rstuart,

You didn't even bother to look did you. It excluded mobile handsets. This may be Itsy bitsy, but is also irrelevant to what I was talking about.

"Internet subscribers by technology type.

Growth in internet subscribers in Australia continues to be dominated by mobile wireless broadband (dongle, datacard and USB modem based services), which increased by 49 per cent in the 12 months to December 2010. However, the increase in mobile wireless broadband subscriber numbers has not been at the expense of mainstream fixed-line services with ADSL subscribers (covering all copper based access technologies relating to DSL, ADSL and ADSL2+) increasing by approximately seven per cent during the same period. Mobile wireless broadband has continued to gain in popularity such that subscriber numbers are now marginally below ADSL subscriber numbers (see Figure 2). At the end of December 2010, ADSL accounted for 43 per cent of all internet subscribers in Australia marginally down from 44 per cent at the end of June 2010. In comparison, mobile wireless broadband subscribers accounted for 40 per cent of all internet subscribers, up from 36 per cent at the end of June 2010."

The figures you so blithely quote are for broadband including wireless.

"Figure 21 Volume of data downloaded via fixed-line and mobile wireless internet services"

As I said previously, many of these connections are complementary to fixed, line, but in 5 years when the NBN is starting to be available in the city, the data and speed offerings of wireless will start to make the $60 p.m. base offer for the NBN look obsolete.

@Shadow Minister: You didn't even bother to look did you.

Touché. I only looked at the tables and graphs.

Fair 'nuff,

I haven't done so much reading for ages.

@Shadow Minister: I haven't done so much reading for ages.

Me neither. Without your prodding I would know no more about the NBN than the next guy.

That DIDO thing - I gather it only works if the number of transmitting antenna's is at least equal to the number of receiving antenna's, and all transmitting antenna's are connected by a fast backbone, and signal strength and phase shift for seen by each receiver from every transmitter is known to within a small fraction of the wavelength. They also have to be uncorrelated (which in fairness the chances are high).

The transmitters then coordinate, adjusting the signal strength and phase so that their combined signals at each receiver add up to what they want the receiver to see. It works because they have N variables they can adjust (the transmitters), and can thus solve for the N results they want to achieve. So it doesn't break Shannon's law and it deployed it would allow the system spectral efficiency to equal the link spectral efficiency (currently the system spectral efficiency is a factor or 6 below the link spectral efficiency). I'll leave it to you to decide how practical it might be.

rstuart:"That DIDO thing"

But the principle of using constructive interference to create an emergent signal at a specific point is lovely, don't you think? I also can't really see a huge problem in having transmitters almost ubiquitously using direct optical links for the backbone and some suitable radio frequencies for the local stuff. Triangulating the receivers woiuld be trivial and the transmitters would be in a well-known absolute position.

Sound-field processors have done this more or less sucessfully for ages, but the missing link has always been the ability to know the location of the receivers, the ears.