NBN Myths Busted

[avandonk]

Interesting video explaining the facts of the NBN.

http://www.youtube.com/watch?v=6a2ne1WKxek


Bert

[richardda1st (Richard)]

Thanks for posting this Bert, very interesting but as most commentators don't care about the facts, most of the population will never be aware.

Cheers
Richard

[mswhin63 (Malcolm)]

The one thing I found lacking is a bit more specific allocation of resources. ie: are they going to upgrade dial-up first before ADSL2+ or are they going to leave difficult locations till it is forgotten about.

The presentation was more suited to engineering audience though. Interesting that it talks about new technology in modulation. I have worked with 16QAM and 16OQPSK NRZ microwave link system almost 20 years ago and the next systems to be brought in before I left were higher than that. This presentation suggests the technology is only in it infancy.

[gary]

Hi Bert,

Thanks for posting the link to the Macquarie University Engineering Week lecture
for 2011 by Peter Ferris, planner for the network. I watched this on Monday.

Since the target audience was predominantly engineers, Ferris makes the occasional
reference to some aspects of communications engineering that I suspect might
not be transparent to some who have not formally studied it.

For example, he makes reference to the "Shannon Limit" at time 24:23.
Since the Shannon limit is only presented to electrical engineering undergraduates
in the third year of their studies when they are doing their second year of a
communications theory course, I suspect most people will be unaware of what it is.

Yet knowledge of the Shannon Limit combined with the graph Ferris presents
at 19:27 showing the exponential growth of bandwidth demand is crucial to
appreciating why the methods of data delivery that have been chosen for the NBN
are the right ones.

The Shannon limit is named after a personal hero of mine who is the Father of
Information Theory, the late Claude Shannon.

Earlier this year, on the anniversary of what would have been his 95th birthday
if he had still been alive, I posted this tribute to him -
http://www.iceinspace.com.au/forum/s...ad.php?t=74961

In that post, I make mention of the Shannon Limit in these paragraphs -

Originally Posted by gary

Consider the problem of transmitting information from a source, across
some medium, to a destination. The medium might be wires, wireless or
even the air that carries sound waves. The information might be speech,
television, text ... it matters not. This paper defines for the very
first time what information, in an communications and computer
engineering sense, really is, and it defines it mathematically. Whenever
a message is transmitted through a communications system, the message
can be altered by noise. Before the 1948 paper, engineers would attempt
to overcome noise in ad hoc ways, such as increasing the signal strength
or repeating the message. However Shannon showed that mathematics can be
used to find an optimal way to transmit a message, including optimal
ways to encode it.

What's more, the paper provided the communications equivalent of E=mc^2
with a formula that showed what the maximum rate of information one can
transmit in bits per second is over a medium of a given bandwidth, with
a signal of a given strength and noise of a given strength. For example,
given a medium such as a copper cable, a piece of wireless spectrum or
an optical fiber, there is a maximum rate one can transmit information
over them and which the mathematics tells us we can never transmit any
faster. When this was first introduced in 1948, many engineers scarcely
believed it could be true. However, the engineering endeavors in
communications in the past 60 or so years since then have been largely
about squeezing every bit of channel capacity out of a given medium, to
get as close as possible to the so-called Shannon Limit. For example,
the information that comes over your copper wire ADSL connection is
within a few percent of the theoretical limit and for this reason,
copper cable communications systems are near the end of their
technological life.

More formally, Shannon showed in what is known as the Shannon-Hartley theorem
that -

C = B * log2 (1 + S/N)

where -
C is the channel capacity in bits per second
B is the bandwidth of the channel in Hertz
S/N is the signal to noise ratio

As Ferris makes clear, the NBN architecture consists of next generation wireless,
satellite and optical fiber delivery. What technology will be used will depend upon
where you live.

However, since the bandwidth of optical fibers is thousands of times broader than the
finite amount of radio bandwidth that exists in nature, fiber is capable of delivering
more bits per second that radio ever can, because the fundamental Shannon
Limit at wireless wavelengths is much lower than that at optical wavelengths.

And the graph (19:27) where Ferris shows the exponential increase in bandwidth
required by households over the last three decades shows that optical
fiber directly to the house is the only technology that is future proof enough
to meet the typical data delivery requirements our society will have in the coming years.

In other words, despite the seemingly miraculous increases in wireless data delivery
engineers have been creating in recent years, we are starting to get within
a few percent of the Shannon Limit. In the years ahead, there will be further
advances in wireless speeds but there is not a lot of head room left and when one
looks at the projected data bandwidth requirements for households in the coming
decades, unfortunately a universal wireless system cannot meet those
requirements. It is also important to remember that what wireless bandwidth
is a limited resource and the future will demand that every spare Hertz of it
should be best allocated for those applications for which wireless is best suited - namely
mobile applications.

As Ferris mentions at 20:50, projections show that video data requirements will
increase enormously as well. Though many of us enjoy HD video today, already
4K (4196 x 2160 pixels) consumer products are about to merge to be followed by
8K (8392x4320 pixels) in the years ahead. One of Ferris's graphics (20:51) describes
8K as cinematographic experience where a 1 hour streaming download will require
anywhere between 13 to 135GB per hour and first consumer offerings are expected
around 2013.

The NBN of course will also replace the entire existing switch telephone network
and all voice traffic will become digital.

It would not surprise me if current terrestrial wireless free to air television
delivery were to switch to NBN optical fiber delivery in the future, thus
freeing up more of the precious wireless spectrum that might then be auctioned
off for large amounts of money to wireless telephone and data carriers.

For very interested readers who would like to learn more about the theory
and practice of digital communications, including concepts such as
data compression, the Shannon Limit, the Nyquist criteria, PAM and QAM
modulation, small signal constellations, Viterbi encoding and so on, MIT
as part of their OpenCourseWare program have the following two undergraduate
electrical engineering courses available online which include video lectures -

6.450 Principles of Digital Communications I
Videos -
http://ocw.mit.edu/courses/electrica...ideo-lectures/
Lecture notes and other resources -
http://ocw.mit.edu/courses/electrica...s-i-fall-2006/

6.451 Principles of Digital Communication II
Videos -
http://ocw.mit.edu/courses/electrica...lecture-notes/
Lecture notes and other resources -
http://ocw.mit.edu/courses/electrica...i-spring-2005/

Thanks again to Bert for posting the link to the Macquarie University lecture.

[Terry B]

I haven't watched the link yet but I do know that the NBN goes past my work but I still can't connect to it as the providers are yet to stop bickering and actually offer a contract.

[gary]

Originally Posted by mswhin63

The one thing I found lacking is a bit more specific allocation of resources. ie: are they going to upgrade dial-up first before ADSL2+ or are they going to leave difficult locations till it is forgotten about.

Hi Malcolm,

Announcements were made for the first and second release sites here -
http://www.nbn.gov.au/follow-the-rol...nland-rollout/
and they have said they finalizing a detailed rollout for all locations in Australia.

It is important to keep in mind that the NBN is not simply an Internet delivery
network. When many people hear the word "broadband" they relate it to mean
a fast Internet connection. However, is also replacing the entire switched
telephone network infrastructure. The network terminating device that will
be installed on your house will still have a phone socket outlet so that you
will still be able to plug in your existing phone handsets and dial a number like
you do today, but the old switched network structure will be ripped out and
gone forever. With this in mind, there will be limited incentive to have islands
of old switched telephone exchanges as it will be more difficult to maintain.

Though there has been some prioritizing for getting the NBN to some areas first,
when you consider the scope and dimension of the project where they plan
on bringing fiber to something like 6000 house per day over the next decade,
the logistics won't simply pan out to upgrade those areas with dial up
Internet connections first, but instead, as the NBN web site itself puts it
succinctly, "For a project of this size and complexity, it is important to get the planning
and design phases right to ensure that the rollout occurs as efficiently as possible".
When one considers the enormous distances involved, spanning optical
fibers across the country between every town and city and the fact that it is
changing the entire communications backbone of the continent, not simply
providing an alternative to ADSL or ADSL2, then it becomes easier to appreciate
the deployment probably borrows more from the logistics of an enormous
military campaign rather than just casual, piecemeal deployment.

The presentation was more suited to engineering audience though. Interesting that it talks about new technology in modulation. I have worked with 16QAM and 16OQPSK NRZ microwave link system almost 20 years ago and the next systems to be brought in before I left were higher than that. This presentation suggests the technology is only in it infancy.

Not quite ... I think you will find that Ferris, who is well versed in the engineering,
in this context is primarily alluding to the use of methods such as QAM at optical
wavelengths on fiber. So he is not saying QAM is in its infancy, instead, as he says
at 28:44 "so as the radio guys have been investing in technology and going through and
working through an analogue medium to gain the most they possibly can out of the radio
spectrum, the optics guys have looked at that and said 'hey, hang on. That looks
pretty good. We might try a bit of that OFDM. What about this QAM? We could try
a bit of analogue across that.'" In other words, whereas techniques such as QAM
have been around a long time in radio, only in recent years have researchers started
to apply them optically, which requires a different bag of tricks.

[mswhin63 (Malcolm)]

Thanks Gary, went through from 28:44, very easily missed the beginning regarding us radio guys.

Hopefully after my studies I won't be just a radio guy any more. I have decided to do a double degree. : BE in Electronic and Communications and BSc Astronomy and Astrophysics. Stickler for punishment.

[Peter Ward]

Peter clearly knows his stuff. The Physics of Bandwidth is undeniable.

NBN business case is a 7% return. Demand will increase.

But.. At the currently best, or even better future 3D HD res, the decision is to deliver "video" content, in full 1920p, to all who want to view, is IMH nonsense. (to paraphrase...kids, do your homework, then watch "Bugs Bunny" )

I have no doubt Peter and his Engineers can deliver this freeway that can send spectacular content to the entire Australian populace times 8..... but..sorry... WTF??? Aren't there are more pressing problems???

National engineering efforts I'd really like to see:

1) High speed rail on the eastern seaboard (Oz is a generation behind Germany, France...well Europe...and China )
2) Clean power generation (Themal Solar or Thorium...don't care..)
3) Irrigate Australia (sorry, climate change IMHO is real)
4) please fee free to add to the list

[richardda1st (Richard)]

My sky has just clouded over , so back inside and back here again.

I could be wrong but this thread seems to be about the technical merits of the NBN and debunking the myths by using technical facts. I think Peter Ferris pointed this out at the start of his presentation.

Cheers
Richard

[mswhin63 (Malcolm)]

Richard good point, Each government initiative is separate from others so best to create a thread detailing other concern.
Communications is the a requirement for all Australians. I received my experience in Radio data communication in Westrail who used the NBN techniques to provide communication for signalling for trains. So Peter please be aware that without the vast communication infrastructure your first preference is mute, same with the second one. (Not so much the third one)

[Peter Ward]

Originally Posted by mswhin63

..... So Peter please be aware that without the vast communication infrastructure your first preference is mute, same with the second one. (Not so much the third one)

Clearly all, systems in a modern society have pressures on them to be upgraded, but I doubt you need optical fibre to run a train.

Agreed, this is not the thread to soapbox on other priorities. My concern is simply that having a compulsory high speed data link to my home doesn't leave me feeling all warm and fuzzy when I'm sitting in grid-lock or being slugged with onerous utility bills.

[CraigS]

There are many variables which go into the selection of technologies and architectures, by geographic zone. The economic issues cannot be divorced from the technologies, nor can they be divorced from the density of demand for service types.

Ferris makes the point in the beginning that NBN Co are building infrastructure to deliver transport for these application demands however, one clearly cannot divorce the functions of the delivery architecture, from the applications themselves. For example: the bandwidth of particular technology types, facilitates delivery speeds necessary to make those applications viable …. and businesses require higher availability/reliability than others, and also need to be be able to forward business critical information as a priority over less important traffic. All of these functions becomes part of the definition of "service" and thus must be catered for in Ferris' network architecture. Make no mistakes about it, the service delivery capabilities of the NBN network, will be critical to the delivery of applications by 'other' providers.

The business decisions made by NBN Co (internally), will impact everyone's expectations, and the efficiency of the organisation itself begins with how people like Ferris view the needs of their customers … and their shareholders.

I see presentations like this, which focus on pure technology options, as leaning towards being a smokescreen, intended to divert attention away from the fundamental decisions on economic and demographic priorities, being made inside NBN Co. These will ultimately, and primarily, determine the success or failure of this venture over any external governmental influences.

Cheers

[supernova1965 (Warren)]

Originally Posted by Peter Ward

My concern is simply that having a compulsory high speed data link to my home doesn't leave me feeling all warm and fuzzy when I'm sitting in grid-lock or being slugged with onerous utility bills.

For one the NBN isn't compulsory though why you would not use it when it can get you out of the grid-lock and allow you to work remotely from home saving on onerous petrol bills or traveling on the train for that matter. The cost of fibre connection is comparable with the old costs of broadband. As stated you don't need to change to the higher speed unless you want to it isn't being forced on us and seeing as it won't cost more than what we pay now I can't understand why it is such a problem for anyone to accept.

[Omaroo (Chris Malikoff)]

Originally Posted by supernova1965

As stated you don't need to change to the higher speed unless you want to it isn't being forced on us and seeing as it won't cost more than what we pay now I can't understand why it is such a problem for anyone to accept.

I guess that at $43bn, it's already cost every man woman and child in this country $2,260 (at today's guestimate) to merely install. That's before we get to pay for its usage. It ain't going in for free Warren, and it is being forced on us as we're all paying for it even if we individually aren't going to use it.

[supernova1965 (Warren)]

Originally Posted by Omaroo

I guess that at $43bn it's already cost every man woman and child in this country $2,260 (at today's guestimate) to merely install. That's before we get to pay for its usage. It ain't going in for free Warren.

I reckon for what it will deliver I for one am happy with $2,260 for the giant leap foward in communication infrastructure we should have had this for atleast 20 years in my opinion it is well overdue. And waiting for the private sector to do it didn't work so I applaud the government for doing it.

[Omaroo (Chris Malikoff)]

I'm glad you're happy, so you can pay for my share too Warren.

Getting away from the central post here. Sorry Bert.

Being in the comms business in one form or another for the past 30+ years, I find that its slightly comforting knowing that the billions spent (or misspent if you prefer) might deliver something that has a chance of at least working on a technical level.

[supernova1965 (Warren)]

Originally Posted by Omaroo

Then you can pay for my share Warren.

In Mackay our public transport is virtually non existant and my taxes pay for the infrastructure for it in all area's of Australia and I have no problem with that as it is my country and when I travel I get use of it. The amount of benifit that will come from the NBN isn't just limited to internet connection it will revolutionise medical care for all allowing access to specialists from all over the world.

[Omaroo (Chris Malikoff)]

I'm not going to get into an argument with you Warren.

[avandonk]

That is fine with me omaroo We could also consider the four billion spent on private health insurance by taxpayer subsidy every year and growing. Not to mention the twelve billion in lost revenue to support negative gearing. The six billion gifted to the wealthiest schools should also come under scrutiny.

The government investment is about 28 billion in the NBN over ten years. It will be amortised by cost recovery.

Bert

[Omaroo (Chris Malikoff)]

As is the case with all of this Bert - there are arguments to support and deny all major points of publicly-funded expenditure - and there are many. We could also stop giving so damned much to Indonesia.

I'm actually reasonably happy with the notion of the NBN itself - if it's implemented correctly and has a solid and purposeful takeup (which is by no means guaranteed) then I'm probably even half good with it. What I'm not happy with is the "management" team at the top who are currently in charge of splashing these funds around with, apparently, little actual managerial or legal skill. Today there is talk within the ranks of replacing our "Managing Director" once again - over the latest of a long string of incredibly expensive bungles. Billions are the "new millions". Doesn't sit well with me I'll say. It's lots and lots of our money at risk so how about waiting until someone with the capacity to manage the implementation correctly comes along first. I'm understandably nervous as I firmly believe that the current crew simply aren't capable of seeing it through - as has been proven so many times now. If this is denied by anyone they are delusional.

Once again Bert - I apologise for contributing to the divergence of your original discussion.