Rstuart,

Your calculations are pathetic. At 1GB/s a monthly usage is consumed in a minute. The data speeds are only used for tiny time slices, and when too many connect, it slows down. Towers have the ability to connect in cells, and pass on mobile units to other cell units.

Unless everyone is trying to download at max speed at the same time it isn't an issue. P.S. the same applies to the NBN nodes. If everyone tries to download at the same time it will also crash.

SM, you nailed it. No network is designed to be 100% available at 100% of capacity to all nodes. To do so is to build in ridiculous amounts of overhead that would be absurdly expensive and unused most of the time.

Every network suffers congestion at full cpacity, from roads to the storm water system to the NBN.

rstuart, I think you're being a little disingenuous. By comparing current technology in mobile services with the proposal for the NBN you're not comparimg apples with apples. For example, due to the poor ability of high-frequency signals to penetrate solid objects and because it's cheaper, some telcos (Optus for sure, not certain about others), are supplying customers with "femtocells" - effectively a personal mobile repreater station that can be installed inside buildings to remove "black spots". There is also the possibility of using antenna systems that are far more directional than the already quite well-focused units in use today, that could be focused down specific pathways, say roads, with femtocells relaying to a small number of houses, once again with directional antennas that direct the signal to each home and not back toward the street. These could be mounted on street light poles with little trouble.

With respect, the spectral efficiency issue is a red herring. Apart from the fact that it has increased massively thanks to advances in both manufacturing technology and the encryption algorithms used, it's only a problem if cells are too large or are poorly designed/located such that there is crosstalk between cells. As I pointed out above, that need not be the case. Further, just as today, a CAN, or City Area Network could easily use fibre in the areas of most use, such as in city buildings. The majors both operate their own CAN already and there's also PIPE.

It seems to me that the main reason for the NBN was/is as a Keynesian stimulus measure, to keep the money flowing for the large contractors who will oversee the work. I doubt it will last as long as the PSTN has.

@Antiseptic: By comparing current technology in mobile services with the proposal for the NBN you're not comparimg apples with apples.

No, I didn't do that. I thought that was pretty clear, but since it apparently it wasn't I'll spell it out in big bold letters.

On the land line on side of the equation I used 1.5M ADSL 1 technology - the stuff that existed 10 years ago. A decade old. Got it?

On the wireless side of the equation I was hopeless optimistic. I used technology that doesn't exist yet, and may never exist. 4G/5G/LTE doesn't achieve what I described. Nor does anything in the laboratory. I assumed we have stuff from the Jetson's era. Got it?

I hope so, because I thought I was pretty clear on the first time around.

@Antiseptic: There is also the possibility of using antenna systems that are far more directional

Something else you evidently missed: to get the 1Gbit/sec speed, I said they used a 16-element phased-array antenna. You obviously don't know what that is, so I will spell it out for you. It is a directional antenna that can electronically alter the direction it points in a fraction of a microsecond. You clamber on the roof and point your high gain yagi antenna at the transmitter once. These things adjust themselves in real time as trees move and cars drive past, to take advantage of changing reflections.

It's bizarre how badly you have got this wrong. It is not a case of me not taking into account directional antenna's, it is a case of using a version so advanced you haven't heard of it.

And you accuse me of using comparing old wireless technology to new land line technology. This is beyond a joke. You are trolling me, right?

@Antiseptic: With respect, the spectral efficiency issue is a red herring.

With respect, all you are saying is they can reduce the cell size. Yes, I know that can do that Antiseptic. My guess is everybody in the industry knows that, and has thought about it. Do you realise that it currently costs 10 times more to get a byte delivered through wireless, than it does via fixed line - and that includes the NBN at current pricing? Check it out - it ain't hard. Your suggestion of flooding the place with $200,000 wireless towers isn't going to help that equation. Besides, high powered radio transmitters aren't popular neighbours.

@Shadow Minister: when too many connect, it slows down.

No, when too many connect the cell collapses. This is a property shared by all broadcast networks, including the old coax ethernet. Put at lot of people on it and they don't slow down proportionately, they collapse in a storm of colliding attempts to resend data lost when it collided last time.

@Shadow Minister: At 1GB/s a monthly usage is consumed in a minute. The data speeds are only used for tiny time slices

Well spotted Shadow. Pity you didn't do some figures to see if invalidated my result. To repeat: wireless can't exceed what we have with ADSL now, thus the idea of wireless providing the next National Broadband Network is a joke.

Yes, they won't all be using it at the same time. But the ISP industry has worked out a formula to allow for that, called the contention ratio. http://en.wikipedia.org/wiki/Contention_ratio A typical contention ratio for residential is 40:1.

So we re-do the figures with a realistic contention ratio. The 1Gbit/sec with a 40:1 contention ratio provides 40 Gbit/sec, or 20 Mbit/sec to 2000 houses in minimum sized 5G/LTE cell of 1 km radius. Notice that is what ADSL can provide at a 1Km radius, without the vagaries of wireless.

Unfortunately that result is for wireless that doesn't exist, and mostly likely will never exist. Lets drop one of the assumptions that made it unrealistic: we reduce the spectrum allowance to a 1/4 to allow for overlapping cells. This is still leaves a number other hopelessly unrealistic assumptions. Even so each house can only expect 5 Mbit/sec, which is worse than what most are in fact getting now.

I don't know whether you two realise it yet, but you are teaching grandma to suck eggs. There are other hints this might be so. The opposition has stopped making noises about wireless in their attacks on the NBN. And you can read what the former CEO of Telstra has to say on the subject: http://www.theaustralian.com.au/business/media/wireless-and-fixed-broadband-tussle-is-far-from-settled/story-e6frg996-1225945853602 I am hoping that even if you continue to think I am a complete Luddite, you might allow that he will know a thing or two about the telecommunications industry.

Dear me, rstuart, you do get upset easily.

You haven't actually addressed any of the things I discussed in my post though.

I'm afraid you ARE comparing the NBN and 3g wireless, because ADLS doesn't compare at all favourably with 3g/HSDPA in tems of bitrate. If your friend is having problems, he may need to install an amplifier, as well as a high-gain antenna, simply because the signal is significantly attenuated by the atmosphere, especially if there's a fine mist of rain close to the transmitter. This is physics, as I'm sure you understand.

However, you're clearly not up to speed with developments in mobile. I refer you to wikipedia:
http://en.wikipedia.org/wiki/4G

"In telecommunications, 4G is the fourth generation of cellular wireless standards. It is a successor to the 3G and 2G families of standards. In 2009, the ITU-R organization specified the IMT-Advanced (International Mobile Telecommunications Advanced) requirements for 4G standards, setting peak speed requirements for 4G service at 100 Mbit/s for high mobility communication (such as from trains and cars) and 1 Gbit/s for low mobility communication (such as pedestrians and stationary users).[1]"

Sounds like a great deal more than 1.5Mb/s to me...

My point regarding using more directional antennas was to do with being able to confine the signal to a narrow-cast to low power microcells proximal to the user. This means that the spectrum can be used more effectively, albeit at a greater cost than the current cellular model, but hey, we're prepared to spend $43 billion, it should be easily affordable. It had nothing to do with the antenna used for the testing of the 3g tech you mentioned. I'm sorrry if this is a little more complex than "fibre good, wireless bad"...