In this lengthy discussion of the “National Broadband Network”, it was generally assumed that the existing copper network had essentially reached its speed limit, and that it could never provide the kind of internet performance that Australians will want in the medium-term future, necessitating the construction of the “Rolls-Royce” fibre-to-the-premises solution announced by Kevin Rudd a couple of days ago. But are these two assumptions correct?
Most Australians with “broadband internet access” at home use ADSL. This technology essentially uses your old phone line to send high-speed data. The latest incarnation, which these days most Australians have access to after Telstra finally turned it on a couple of years ago (after deliberately leaving it dormant in their exchanges for some time) is ADSL 2+, which tops out at 24 Mbit/s under ideal conditions. That is, an ADSL 2+modem can carry about 3 megabytes of data per second. That’s the rough equivalent of three books’ worth of text per second, or or a three-minute pop song in MP3 format. Theoretically, it’s enough to stream live about 1.5 HDTV channels, or, if you like, one HDTV and one SDTV channel. The practical capacity is somewhat lower; something closer to one HDTV channel is probably the practical limit of ADSL2+ under good conditions.
But is that the limit of what the copper network, with a few upgrades, might realistically provide? In a word, no.
The government’s original proposal for “fibre to the node” would have involved replacing only part of people’s phone lines with fibre optic cable; the last few hundred metres would have been left alone, and each “node” would have connected to individual buildings using a of a souped-up version of ADSL called VDSL.
So, how fast is VDSL? Considerably faster than ADSL2+, if you put in fibre links to an intermediate point, and just use VDSL for the last bit to individual homes and businesses. The latest version – VDSL2 can support speeds of around 100 megabits per second over a half-kilometer distance to the nodes. Yep, that’s as fast as the government is promising for fibre to the home, and it’s fast enough to deliver four simultaneous high-definition TV channels.
That’s with existing technology. What might we be able to do in the future? this academic paper suggests that, in many cases, it might be possible to do a whole lot better, by borrowing some technology, amusingly enough, from the latest generation of wireless devices. In some places, there are multiple unused wires in the “local loop” to people’s houses – provided in case people want to have second or third phone lines. Those lines are essentially sitting dormant. At a naive level, if you’ve got 4 lines available, you’d think you could increase the speed by a factor of four, right? Wrong. Through magic which I do not claim to understand but is now fairly standard practice in electrical engineering, you can take advantage of the interaction betweenthe signals on the lines to transmit much more information than you’d think – you can add one and one and get more than two. According to their calculations, it’s quite feasible to squeeze ten times the data on to four lines. That’s enough throughput to carry 40 HDTV streams simultaneously.
But there’s a further complicating factor. The unit of measurement I’ve chosen is a high-definition digital television stream, because that’s the most demanding type of data most people will ever try to send or receive on the internet. But, with video, what’s important isn’t how much data is transmitted, it’s that the picture you see looks good, regardless of how much data is actually transmitted. And depending on how you represent the information in the video stream, the amount of information required to make a good-looking picture varies greatly.
No practical digital video system (except for some scientific applications) tries to transmit exactly what the camera sees to the end viewer. What you see on the screen is an approximation. The methods for making those approximations – known as codecs – have improved a lot over the years, and the latest ones can squeeze a video that looks just as good into a lot less space. The standard used for terrestrial digital video transmission known as MPEG-2, is 15 years old. Later schemes, such as MPEG-4, can squish a picture that looks just as good to you, the viewer, into half the space. So our standard ADSL2+ can carry two HDTV streams using MPEG-4, VDSL2 can carry eight, and the super-DSL of the future might conceivably carry 80 or so.
So, in a nutshell, I reckon that we could have gotten enough capacity out of a moderately souped-up copper network, in large parts of Australia, for the medium term future. And we could have done it at much less than 43 billion dollars.
The experts advising the government presumably know this, but they’ve still recommended replacing Telstra’s copper network entirely with fibre optic. And, in my view, the key reason isn’t technical. The expert report on the broadband tenders is concerned the upgrade path from the kind of FTTN network I’ve examined here to the Rolls-Royce FTTP, but the key objection is as follows:
The proposals have also demonstrated that rolling out a single fibre-to-the-node
(FTTN) network is……likely to require exclusive or near-exclusive access to Telstra’s existing copper sub-loop customer access network (CAN), the so called ‘last mile’, thereby confirming that strong equivalence of access arrangements would be essential. As well, providing such access to a party other than Telstra runs a risk of liability to pay compensation to Telstra. The Proposals have this risk remaining with the Commonwealth but they have not addressed the potential cost to the Commonwealth of any such compensation. In any event, the Panel considers that no Proponent could accept the cost risk and continue to have a viable business case.
Of course, if Telstra had never been floated, or the network itself had remained in government hands and just the retail part had been sold, this wouldn’t have been an issue. How many billions of dollars will that little mistake cost Australia?
Have to wonder if it would just be cheaper for the government to give a huge amount of money to Telstra for their last mile infrastructure. Especially while the stockmarket is so depressed. From what I’ve seen there is no general plan or budget to dig conduits for the new cables and there’s going to be all sorts of objections locally to more cable being strung up along poles in the suburbs.
Most of the Transact network in Canberra is FTTN which then uses VDSL. However (and I don’t really know why) its more expensive than ADSL2+ at equivalent speeds. To put a bit of realism into the debate about streaming movies etc, an unlimited 10Mbs connection will cost you around $800/month just in ISP costs (you pay the ISP and Transact separately).
My brand-new ADSL2 has just downloaded Essential Research pdf @ 5Mbs – and that’s on a dedicated Internet line with the modem the only traffic. It’s mid-morning (11.00am). From now until c9.00pm, this is about as fast as it will get (but not yet as slow). This is on past-copper, part fibre-optic.
The government seems pretty keen on throwing money around regardless of the cause. This seems like a better option than the straight to consumer stimulus package(s).
You’ve got to wonder how much more fun this would be if the power companies were as aggressively capitalistic with their poles as Telstra is with it’s infrastructure.
Telstra would value likely their infrastructure at the long-term (i.e ignoring the lower costs due to the GFC) depreciated cost of replacing it. It might end up costing less overall but wouldn’t be any better value for money as you’d be getting aging assets made up of old technology. Additionally the money would end up going straight to Telstra shareholders rather than into stimulating the economy.
DeeCee: but the limit is probably further up the line, not between the exchange and the modem.
We’re so good in Australia at finding reasons not to do things.
Even the poms are ahead of us in solar energy installations for example, as anyone who has visited London recently could support. It’s even a viable proposition in that climate.
And Obama announced that an increase in jobs growth in the US is mainly down to increased employment in the alternative energy sector.
We’ve become so used to our status as a do nothing, complacent and basically negative society that ships innovation offshore and is content to be the world’s quarry, that we can’t recognise a good idea when we see it!
Of course, as long as Telstra continues obstructing this change, it makes perfect sense for KRudd to skip over them entirely. My small town has been begging for ADSL 2+ (the pharmacy, the doctors, the hospital, the high school and one of our major exporters all desperately need faster communication) and Telstra has told us that we’re “never” going to get it.
lilacsigil @ 7 One option is to approach an ISP like Internode. They have a facility on their website where people can register their interest in ADSL and if enough people request it they’ll install their own ADSL2+ DSLAM (as long as there is room in the local Telstra exchange).
This all sounds well and good Robert, but ADSL2+ is currently failing to provide a fast internet connection to people like me. I live in a suburb in Canberra that just happens to be far enough from the exchange that ADSL2+ is marginal at best.
This seems to be a common complaint with ADLS2+, theoretically good speeds .. but .. your attentuation is too high .. you’ve plugged too many telephones in .. there is a source of noise that is affecting the line .. you need a central splitter in your house .. your ADSL2+ modem is not negotiating a high enough connection speed, buy a more expensive one .. etc.
I ended up getting a ’3′ mobile broadband connection that was faster and cheaper. It is currently slowing down rather alarmingly .. I guess a bunch of other people have cottoned on to how cheap it is and the cell tower is having to serve more connections.
Aidan: but the point of a FTTN network with VDSL would be that for places that are “too far from the exchange”, a fiber link is put in to a “node” which is only, at most, half a kilometre away from your home or business.
That’s still a much cheaper proposition than FTTH.
As for multiple phones etc, the idea is that everything, including the phone service, should be run over the VDSL connection.
No. You can not simply assess a technology’s theoretical maximum speed and then divide by 8 and assume real-world conditions. The header and footer for a TCP/IP packet is carried by that 24Mb/s too. This immediately reduces the actual useful data rate (depending on the size of the packet).
More importantly, fibre has no meaningful maximum speed. It’s as fast as the network equipment at either end. Copper has a maximum, and even with newer mechanisms for compressing and/or encoding data, a pair of copper wires won’t give the sorts of speeds you’re referring to. Copper is limited in exactly the way that fibre isn’t.
Obviously, FTTN+xDSL is cheaper. But the key, I think, is that we don’t own the copper, and a public entity owning the fibre infrastructure sidesteps Telstra (and yes, we clearly should not have sold it). No doubt some day post its construction a future Liberal or Labor government will flog Ruddstra off for short-term gains (because they’re idiots, ideological, or desperate [or in the case of NSW Labor, all three]). I don’t immediately see this as an argument against FTTN. I’m happier to discuss alternatives – if FTTN is the best way to drop $43 billion – but you can’t say that as we haven’t reached the theoretical maximums for xDSL we should stick with it.
Alister: Don’t mean to be uppity, but I did point out that the theoretical maximum and the practical maximum are considerably different.
Furthermore, the paper I was referring to was talking about using MIMO technology to take x unshielded twisted pairs with capacity y and get z total data rates, where z >> xy. It’s not my field of expertise, but it was published in an IEEE journal, which generally indicates they’re not talking total rubbish.
Robert, I understand what the FTN proposal is (was?). I’m just pointing out currently ADSL often falls far short of it’s theoretical maximum speed. That their are significant technical hiccups that frequently make it difficult to even achieve the speed that should be possible given the run to the nearest exchange. From Wiki:
Then there is the generally crappy state of some of the copper infrastructure.
I think the FTTH is a good idea, but it could well end up costing ALOT more than $43bn .. how much we’ll see I guess.
(NB: That link above to Kate Lundy’s site should have a target on it (i.e. have a hash in the url) but this breaks on this site and causes all text after it to disappear. So you’ll just have to page down to the story titled “27 May 2003 – $2 billion: the cost of Telstra’s neglect” or copy this link:
http://www.katelundy.com.au/May2003.htm#27May2003
I find this is a common problem when commenting on LP)
In yesterday’s Australian it was reported that Telstra is considering volunteering for structural separation, and selling parts of its backbone to the new network owner. An enormous change in attitude in a very short time.
http://www.australianit.news.com.au/story/0,24897,25331701-15306,00.html
FTTH will be a good long term solution, fibre already has much better upgrade paths than copper and the new network overcomes the commercial and political mess caused by the badly handled Telstra commercialisation and privatisation.
Robert, I know, but then I read it that you referred to theoretical maximums etc. What’s overlooked is the overhead for each packet (so 100Mb/s is not 12MB/s, but closer to 8MB/s). You said that,
But if I understand correctly, 1080i requires at least 15Mb/s, which you’re not going to get from ADSL2 even under good circumstances. And transmitting 1080i flawlessly would require more than that, even if you accepted dropped packets (and so would use UDP/IP which doesn’t require acknowledgement of receipt).
But the real issue for me is that we could try to keep pushing xDSL technologies, or just go for the existing technology we already know works. I’ve no doubt that we can get better rates from copper than we do now, but what about those places where there aren’t four pairs? If we were to recable from the node to the home, we would not do it with copper (even Cat6, which is happily giving me 1Gb/s on the computer I’m tying on). I’m sure the IEEE paper isn’t rubbish, but it’s not in production today, and there’s no guarantee it’ll work reliably and cheaply in real-world conditions (consider IP over power lines).
Lets see now.
Robert wants ADSL instead of fibre optics because it might work better in the future!!!!!
Point is there is no practical limit to the speed of fibre optics. 100 Gb/s? no problemo.
ADSL maybe 10 GB/s if you can find another 3 phone lines and there is no-body else living within 2 miles ! Get real.
FYI information Robert entire housing developments are being constructed in OZ right now and you know something? Not a single copper wire phone line in any of the homes. Like this is the 21st century – why are you such conservative old fart?
Want the reactionaries want to do is nobble us with ADSL or wireless broadband. Its totally pathetic. They seem to have somewhat belatedly done the numbers on the power required to run very high speed wireless networks and have now backed off and are proffering fucking ADSL instead.
Must close now as I have to polish my button up boots,tell the groom to feed the horses and beat one of the servants half to death. You make me so angry Robert.
BTW check out the Irish system.
Huggy
“why are you such conservative old fart?”
It’s the Australian way. We’re scared of innovation and big projects, unless it’s done overseas, in which case it’s bound to be a good idea.
You’re right Huggybunny, it’s truly pathetic.
Huggybunny and Alistair – there’s a huge difference between ADSL and VDSL. You can do 100Mbit with VDSL2 now. And there’s little point spending all the money on FTTH if the backhaul is so congested that you can actually only use 5Mbit (or less) of it during peak periods.
You can squeeze 1080i into less than 15 mbit/sec, but it starts to look a little crappy at lower bitrates.
But, if you’re transmitting pictures over the internet, you wouldn’t use MPEG-2 compression anyway, as I’ve noted. MPEG-4 H.264 can squeeze a 1080i signal down into roughly half the size with no perceived loss in quality.
What both of you seem to be missing is that a near-universal rollout of fiber to the premises will be colossally expensive, and there are squillions of other bits of infrastructure governments could spend money on – rail, the electrical grid to support renewables, water, bike paths…you name it. Whether it’s worth spending (at a rough estimate) $2000 per Australian on, given alternatives, is a question that deserves a bit closer examination.
Maaate. We’re the wuurrrld chaaaamps at it! Best fugggincountryinawurrrrrld for sayin’ ‘NO’. Only reason they haven’t made NO an Olympic Sport is ‘cos we’d win all the medals, if we could be arsed.
Here’s some other events in which we’re the chaaaampions of the wuuuurld mate:
Holding your horses.
Steadying on.
Not gettingaheadofyuuurself.
And here’s what the crowd chants at the finals (to the rhythm of ‘whadda we want…when do we want it?):
Whadda we need?
THAT FOR!
Why wouldyawanna?
DO THAT!
I can’t access ADSL of any kind. I only recently got access to wireless. I live 35 minutes from the Adelaide GPO. To be honest after paying outrageous line rental to Telstra for a grossly inferior product..ie: a shoddy dial-up connection, that they had no interest in upgrading..I’m not interested in any solution that continues Telstra’s monopoly.
I suspect anyone who supports a solution that includes the existing copper network has been well served by that network and has no idea how frustrating it is to try and get access to a practical and affordable solution, knowing you are paying significantly more than people with faster speeds and unlimited downloads.
Visionary, nation-building projects are only visionary and nation building if they’re good.
The Adelaide to Darwin railway might have been visionary, but so far it hasn’t come close to paying for itself if I understand it correctly.
Eastlink, Regional Fast Rail, the Cross City Tunnel…
Just because something is “visionary”, “nation-building”, and “big” doesn’t make it good.
fb @ 21 – When looking for somewhere to live the last two times, being less than 1.5km from a telephone exchange was high on the list of priorities
Just because something may not provide a commercial return doesn’t make it bad.
Robert, it proves my point that the only projects that you can identify are hardly large scale nation building exercises, and in the case of the Cross City Tunnel were compromised from the start because of the very lack of commitment to innovation that is the cause of the problem.
Can anyone name an example of a genuine nation-building project that this country has committed itself to in the last 50 years? Apart from the obvious, which was begun more than 50 years ago anyway.
What a sorry record.
Chris @ 23: I’d do the same too if I were to move again. internet wasn’t really high on my list of priorities when I moved here 10 years ago, but it will be next time I’m looking for a place. BTW: I’ll also be checking for “pair gaining”…I got the double whammy of broadband road-blocks…unfortunately this never did stop the tele-marketers from Telstra attempting to sell me broadband on an almost weekly basis.
Robert, I’m more than happy to debate FTTH as a matter of opportunity costs. But in terms of technical superiority, there’s no argument. FTTH beats FTTN+xDSL. I think I see this as nation-building infrastructure because it’s recognising existing trends and building on them. No-one’s ever going to say that insert-data-speed-here is fast enough. Data traffic can never be fast enough, and FTTH allows for much faster maximum speeds than xDSL ever will. Sooner or later, FTTH will happen. So why not now?
FTTH is the only serious long term answer. Fibre optic technology is very well developed and reliable. There is no practical limit to its carrying capacity, and hence it also provides excellent future proofing.
Yup, it will be expensive up front to do it properly.
So what?
A one-off cost of $2000 per person is trivial for something that will give major, essential, society wide benefits for many decades to come.
Just. Do. It.
Robert, another thing to remember is that the speeds you’ve mentioned for ADSL2+ (24Mb/s) and thus most likely for VDSL apply within 500m of the exchange. This attenuates fairly rapidly the further you are from an exchance.
Up here in NQ where we have three (yes, that’s 3) exchanges serving a population of 150K, being within 500m of an exchange is somewhat rare. As a result, I know people on ADSL2+ who are receiving 4Mb/s (so nothing near the promised 24Mb/s) and by all accounts, 8Mb/s tends to be the norm.
Of course, with fibre to your door (not that crappy shared stuff that Foxtel sells down south where you’d be lucky to get 5Mb/s), that’s 100Mb/s no matter WHERE you are connected.
Oh, and the other wonderful benefit of high speed broadband at your door is the great Telstra killer, VoIP (Voice over IP). So the $70/month I’ve seen quoted for FTTH is actually more like $40/month (because you can scrap your monthly phone connection) and we’ll end up paying peanuts for voice calls (I’m currently paying 12.5c to ANY landline in OZ and 1.9c/min for international calls with NO monthly service fee – my phone bill has totalled less then $60 for the last year!).
Time to move with the times folks or we’ll all be left behind (well, Telstra might be, but we’re all moving on).
GoTroppo @ 29 – although you only have 3 exchanges, you would have a much greater number of nodes in your area. The existing copper from houses in an area run to these nodes (probably look like grey boxes or poles) and from there run to your exchange. FTTN is about running fibre from the exchange to these node boxes which are fairly close to your house.
btw if the Transact FTTN prices are indicative of what you’d be paying its around $130/month for a 30Mbit connection but this also includes a phone connection. However you also need to pay an ISP (infrastructure is split from the ISP as would happen with the FTTH proposal) so you can factor in say $80/month or so for a 60Gb quota – which you’d use pretty fast at 30Mbits
From what I can see from this proposal its not about making high speed internet access cheaper, its about making it more accessible. Also it doesn’t have to be a choice between FTTN or FTTH – I don’t see why you can’t do FTTN first and then do FTTH later. The big problem about FTTN is access to Telstra’s infrastructure. For new housing estates they might as well install FTTH, but in other areas if FTTN is cheaper then it’d probably be worth spending money on increasing capacity in other areas of the network.
Dare I suggest that we’ve had millions of “nation-building” projects over the past 50 years, from the playground across the street from me in Brunswick up.
And what Chris said about VDSL.
Alister: I acknowledge that fiber is, ultimately, the technically better solution. The question is whether it’s worth spending tens of extra billions of dollars rolling it out to millions of existing premises if the VDSL option was available.
Robert; we have tried to explain to you that your fibre phobia is irrational in the extreme. There is no need to interpose an obsolete and half baked technology between you and the different. Embrace the fibre Robert, the fibre will liberate you, you will be free. Cast aside the chains of the technologically retarded – please. You can be saved.
Get with the program eh?
Huggy
robert – lets face it the majority of australians now hate Telstra so much they will pay almost any price collectively (they’ve been paying it through individual bills for years) to get rid of them or at least bluff them into co-operation and competition. This little move (FTTH) will do that.
“Get with the program eh?”
Yer, Robert, this is the carbon fibre bike of communications to ride into a bright new future.
Huggy,
No point in the fibre setting us free if the bill puts us back in prison.
Huggy: a simple question. Would you prefer:
* 10-15 billion on a FTTN network, and 30 billion on a massive spend on renewable energy/energy efficiency deployment.
* 40 billion on FTTH.
I can live with only four or five simultaneous HDTV streams for the next decade if we used the money to fix our emissions.
FXH: I take that point.
Robert;
“Huggy: a simple question. Would you prefer:
* 10-15 billion on a FTTN network, and 30 billion on a massive spend on renewable energy/energy efficiency deployment.
* 40 billion on FTTH.”
I answer:
The governments of the world will do nothing about “renewable energy/energy efficiency deployment” until it is too late. Nothing will be spent until the sea starts to rise over the shores of North America. At that time money will flow like honey from the hive, like water over the Niagara falls, like the sweet lies that flow from the mouths of Priests and Politicians. Oh yes it will flow.
They will go down on their knees and pray as they open the bank vaults. “please save us” they will cry.
“Too fucking late” will be the response.
In the meantime we will get our fibre optic high speed network, we might as well, as it is the only part of the GW solution to be funded.
Huggy
Chris@30 – ADSL still relies on distance from the exchange so nodes make no difference. It also relies on continuous copper from the exchange to the house so suburbs where RIMs have been installed (because they’re too far from an exchange to lay new copper) have been without ADSL (thus Telstra’s push for wireless broadband).
We also seem to have selective memory loss in that back in the 80′s/90′s the government bankrolled the upgrade of EVERY exchange in OZ to digital – another project that cost billions but has yielded MASSIVE benefits since and delivered the platform for today’s broadband. Do we really think that any of this would’ve happened universally if it was to be solely funded by the private sector?
Where’s option C? 40 billion on FTTH and 30 billion on renewables?
desipis: or how about option D: 10 million on FTTN, 30 billion on renewables, and 30 billion on urban public transport?
The point I’m trying to make is that there is an opportunity cost here, and nobody has bothered to make a sufficiently compelling case, as far as I am concerned, that absent Telstra’s bloody-mindedness it would be worth going for FTTP over FTTN.
Robert – Perhaps you should read this =-from Crikey today.
Rod Tucker, Laureate Professor at the University of Melbourne, writes:
The Rudd Government’s proposed fibre to the premises (FTTP) broadband network has generated a rich variety of ill-informed media commentary. It is time to debunk some of these myths and set the record straight.
Myth number 1: Wireless will provide a competitive alternative to a FTTP network.
The reality: Wireless internet access provides flexibility and mobility. It is useful for people on the move who want to do simple things like accessing e-mails, browsing the web, and downloading small files. But it will never be able to support universal broadband services. The big problem with wireless is that all users have to share the same “airwaves”.
Imagine a politician communicating with constituents at a town hall meeting. Everyone in the room has to share the same sound space to communicate with the politician. If the politician is talking to only one person, that person gets the politician’s undivided attention. But as more people attend the meeting, each person has to wait for their turn to ask a question or express an opinion.
Exactly the same thing happens with so-called wireless broadband. Service providers advertise impressive bandwidths. But the only way a user can get the advertised bandwidth advertised is if there are no other users. As the number of subscribers increases each subscriber is allocated a smaller proportion of the total available bandwidth.
This is already starting to happen with 3G wireless in Australia — a recent surge in subscriptions to 3G internet services has led to a degradation of user experience. Future generations of wireless (4G, LTE etc.) will help to avoid this problem, but they will never be able to provide the bandwidth of a FTTP network.
In addition, both wireless and satellite can provide services to users in remote areas that are difficult to service by other means, and a well-designed national broadband strategy would channel the use of the limited radio and satellite spectrum so as to serve those customers well.
Myth number 2: An FTTP infrastructure will lock Australia into old technology that does not allow future upgrades to new technologies.
The reality: In FTTP networks, a single feeder fibre from the telephone exchange is shared by 32 premises. Near the premises, the feeder fibre splits into 32 fibres that run all the way to each user. The fibre from the telephone exchange carries an incredible 2,500 megabits of data per second.
If all 32 houses simultaneously use the network at its full capacity, each user can obtain one thirty-second of the 2,500 megabits per second, or roughly 100 megabits per second.
Importantly, the experience of each user is not limited by what other users are doing. But this is only the start. The optical feeder and fibre is capable of carrying more than 1,000 times this amount of data. Therefore, as new higher capacity terminal technologies currently in the pipeline become available, it will be possible to upgrade the system to much greater capacity.
The FTTP network could one day deliver many gigabits per second to the home. This could be achieved by just changing some equipment in the exchange and in the home. One compelling advantage of a FTTP network is that the core infrastructure, which constitutes the bulk of the investment — the fibre in the ground or strung from poles — is completely future-proof and will not require any additional upgrades.
Myth number 3: Enhanced DSL and hybrid fibre coax networks (HFC) will be able to provide the same capabilities as FTTP at a fraction of the cost.
The reality: DSL technology has advanced over the years, but unlike fibre, it is already operating close to its full capacity. There is no practical way to provide widespread delivery of 100 megabits per second over DSL. The HFC access networks operated by Telstra and Optus could be upgraded to 100 megabits per second. But like wireless, the 100 megabits per second is shared and user experience on these networks is degraded as more users use the network.
More importantly, the HFC networks cover only a small percentage of Australia’s population. With today’s technology, it is cheaper to roll out FTTP than to roll out new HFC networks. And HFC networks are not greenhouse-gas-friendly. Like wireless access networks, they consume about four times as much energy per user as FTTP.
Myth number 4: FTTP will provide little value because most home users are happy with today’s broadband service.
The reality: Not so long ago, most home users were happy with their 56 kilobit per second service, and more recently with their 0.5 megabit per second ADSL service. We have learned that new opportunities and new services arise when the capacity to support them is made available, and it would be bold indeed to call a halt to further improvements in network capacity because “all of the valuable services have already been invented”.
But more importantly, FTTP is not just for the home. It will enable new services in health, education, business and government services. Video on demand will enrich entertainment opportunities, and video conferencing and other online networking facilities will reduce business travel and cut greenhouse emissions.
Myth number 5: An upgrade of the existing network would be more environmentally friendly than a new FTTP network.
The reality: An upgrade of the existing network to just 10 megabits per second would require a new 200 megawatt power station, presumably coal fired, just to operate the new equipment. The environmental impact of this would be huge. Through savings obtained by removing old equipment, a FTTP network could operate without requiring a new power station.
Per Huggy
BTW the option of wireles broad band would require over 10 times the 200MW.
“But more importantly, FTTP is not just for the home. It will enable new services in health, education, business and government services. Video on demand will enrich entertainment opportunities, and video conferencing and other online networking facilities will reduce business travel and cut greenhouse emissions.”
Gosh huggy, you are a spruiker for this technology!!
I have to give it to you, you have a very deep understanding of the various technological options, and from that perspective this thread has been very informative.
But Robert is right, without looking at the opportunity cost of the full spend within the context of a proper cost-benefit analysis, we simply don’t have enough information to determine whether this is the best use of government money.
For example, it isn’t at all clear to me why the government should be subsidising a further “enrichment of entertainment opportunities” when there look to me to be ample already and besides it is a private benefit and not a public good. As for reducing business travel, I’d like to see a quantification of that, but I’d be surprised if the impact wasn’t very small. Ditto on improved health services.
Point is, you can tally up all the suppposed externalities from delivering FTTP, but without putting a rigorously determined figure on their worth, this discussion is all hot air. Given that households will have alternatives to FTTP, I want to see concrete, independent estimates of likely take-up rates and likely willingness to pay.
I simply find it extraordinary that a decision of this magnitude was made without more evidence based analysis, and until I see some independent analysis (say by the Productivity Commission), I for one will remain sceptical.
Regardless of the likely litigious response of Telstra had Rudd’s option attempted to use existing copper networks (and the FTTP deftly waves that problem away), the manner in which the existing network has been maintained by Telstra is absolutely appalling and would have presented very very real obstacles to even worst case service levels.
Pits with plastic shopping bags “protecting’ joints; EJs with cabling running above ground; pair-gains supporting 4 to 5 phone numbers, so that no form of ADSL can be provided. This isn’t isolated; this is common in rural & urban areas; Telstra have pissed away the worth of the telecommunications infrastructure they ‘inherited’ and frankly I think the government is very very aware that FTTP is necessary as the majority of the copper network is such disrepair that it cannot deliver anything approaching acceptable levels of service.
Bernice, I agree, basically the copper network is buggered as indeed is much of the other infrastructure that we need. The electricity network is in an appalling state as well, unfortunately there is no technology as future proof and out and out reliable as FTTP on the horizon for the electricity network. Neither the telecoms nor the electricity networks have been properly maintained.
I am totally amazed by the inability of otherwise intelligent people to observe the decay that exists all around them.
Huggy
According to Alan Kohler tonight, China is buying up all the world’s copper. For solar technology?
Grace, I think it is for urbanisation and the power network. They build a new power station every week or so. They must need a lot of cooper for generators, bus bars etc. China has basically given up on coper wires for telephone, you get a mobile or nothing.
Huggy
Robert – re. the 4 lines argument.
You’re about 1/2 right, but your conclusion is wrong. There are extra lines available although not many, it’s truer to say that there are vastly more telephone numbers available at the exchange.
There are several issues here:-
1. The exchange can switch many more connections than are actually physically delivered to nodes. This means that if I call a number xxxxxx1 the exchange can route that internally through a series of other numbers xxxxxxx2 xxxxxx3 and so on before finally getting to my house at yyyyyyyy10 (there are network management reasons for doing this – not real good ones, but it happens)
But those numbers are not actually delivered to the street (aka the node) in the form of wire. They’re “virtual” and used for internal routing purposes inside the core network. And used a lot as it happens, in the old days 1800 and 13 connections for example used to be routed through these “virtual” numbers (it works a bit more efficiently now)
2. Taking 1. into account (lots of nuumbers and potential connections) there is a bit of a surplus of wire to the node. So if you want an extra connection to your house the Telstra technician doesn’t have to run new wire from the exchange to the node, only from the node to your house.**
Which leads me to:-
3. In most homes there is only one wire from the node to the house. So if you want an extra line the Telstra guy actually has to string that wire – firstly from the node to the pole outside your hose and then from the pole across your front garden to your house.
So there is nowhere near as much wire available as point #1 would imply. #1 is talking about spare connection capacity (of which there is oodles), #2 is connection capacity to your street (not so good) but #3 (connection capacity to your house) is critical.
There is very little of it. I speak here as someone who actually has 3 physical bits of wire coming into my house. On each of those I can run a phone line and ADSL.
But I’m very unusual in actually having those connections, very few houses have them. And the second two connections (which I had put in) were expensive to install.
So, although there might be a bit of “magic” that allows bonding of separate physical lines, it’s not really practical on a nationalwide level.
There is no extra wire to the house (in 99% of cases), and there is not enough spare wire to the node to service more than a small fraction of the houses in a street.
It would cost a massive amount of money to install extra Exchange->Node capacity as copper (that’s why Fibre-to-the-Node was proposed). Overcoming copper bandwidth issues from the node to the house by running extra copper lines would be astronomical.
Fibre’s the way, really it is. You only have to run it once and you can upgrade its capacity by equipment change at the exchange and the house (or in the old scheme – the node)
The real cost is not so much copper vs. fibre (ie. the material) but the extra cost of physically running the cable no matter what its nature. Once you’re committed to extra cable, you should make it as high capacity as possible and copper just doesn’t cut it.
Secondly (caveat to above) there is a huge cost if you only go to the node. At the node you need transceivers to take the signal off the fibre and put it on copper.
True, you also need those transceivers in the premises if you go straight for the FTTH option (ie. the customers equipment has to do the fibre-to-wire conversion in that case), node based transceivers are a dead loss if you intend to go the house eventually anyway.*
* The house based transeiver (which these days is usually a board in your PC but could equally be a little generic box on the outside of the house) is also cheaper as it doesn’t need the same bandwidth.
** You will have encountered this if you’ve ever installed a phone service where there previously wasn’t one (or added an extra line). The techie will often say something like “I’ll have to check if there’s a loop [wire] available”
JM: Agreed that if you have to run new cable to the premises, you’d be crazy to do anything other than fibre.
The question is how many premises (that includes businesses, which run multiple phone lines much more often than people do) have multiple twisted pairs between the node and the premises.
My house has at least two, I believe. My parents had three active at one time. I would have thought that it would have made sense for Telecom (back in the day) to put in the cabling for at least two lines into every home by default.
Excellent explanation JM.
I was just on the phone to my uncle, who used to do line maintenance for Telstra dating back to when it was the PMG.
According to him, most homes have two twisted pairs running in to them, even if only one phone line is installed.
So while the 4 UTP cables might not be feasible, 2 is, and I still reckon that’s easily fast enough for most people’s needs.
It might be suitable for today, however even that’s problematic.
But why waste money on a solution that will certainly not be suitable for the future?
You certainly seem to have a blind spot regarding this issue, Robert, and no amount of reasoned analysis from those who are experts in this area will change your mind.
aidan: because the reasoned analysis has established one thing: that the potential bandwidth of fiber is practically unlimited. Nobody has established that we can’t tweak a fair bit more speed out of the existing copper and wireless infrastructure with comparatively cheap upgrades. Furthermore, we are all agreed that FTTP will be seriously expensive.
What nobody has bothered to establish, beyond a handwave and “if you build it, they will come”, is that people will want to pay for all that extra speed.
I’m sorry, but I don’t think we should be spending 40 billion dollars without thinking very carefully whether we actually need the services it makes possible; and whether it’s sufficiently urgent that we need to start immediately, given that we have so many other pressing issues to deal with.
Robert, you totally don’t get it.
The existing copper network is basically a heap of shite, much of it is over 50 years old, it is way overdue for replacement and you want to replace it with more of this 19th century technology!!!!!!!!!!!
Perhaps you should actually read the material at
Huggybunny@43
Apr 16th, 2009 at 3:07 pm
It was written by:
Rod Tucker, Laureate Professor at the University of Melbourne, writes:
He totally destroys all your arguments.
You are beginning to sound like all those tired old reactionaries who routinely oppose the new.
Huggy
Huggy: clearly we’re going to have to agree to disagree on this one.
But let’s just take that.
In other words, his argument on demand amounts to two paragraphs of “build it and they will come”.
They built the Concorde; everyone flew by 747 instead.
Exactly, Robert – and in the mean time we will need to be paying the interest on any debt and for the opportunity cost for any other possible use of the funds spent, not least of which could be that they are not taxed off us in the first place.
And in the future we put up with an increasingly unviable, outdated solution which every other country will have abandoned.
And our economy and society suffers as a result.
Anybody who thinks patchwork solutions are the answer should do a bit of travelling and see how backward in so many areas this country really is.
This is yet another example of the kind of thinking that has held this country back for so long.
Robert, you seem to be arguing that we should spend a fair bit now to patch up the rapidly ageing and outdated copper network, and then spend a whole lot later on full fibre optic.
The copper network is simply starting to run into its limits, no matter how much clever (and probably quite expensive) tweaking we do to it.
So why not just do the fibre optic now? We are going to have to anyway eventually, and not that far into the future either.
Also, the standard telcom cable into a standard house/flat is 2 pairs (or used to be, might be more now). But the reality is that many of those cables only have 1 pair working. The 2 pairs are in part a redundancy arrangement for when the first pair fails, which is not infrequent, particularly in areas of high rainfall and humidity.
Well Lord Kelvin is it true that you said in 1895 that heavier than air flight is impossible? “Oh yes I did say that but what I really meant was that even if it is possible no-one would fly when they can use those really comfortable steam ships”. (I made up the bit in parenthesis)
Well Robert you are in good company.
We are not saying “build it and they will come” we are saying build FTTP instead of refurbishing the 19th century copper wire system.
Huggy
Leaving work this evening, the lift told me that Obama was going to build high speed trains to 7 American cities to ensure that America was as advanced as Japan, Germany etc in public transport infrastructure.
Poor Australia – we’re lucky to get any trains at any speed.
Reminds me of the broadband issue for some reason…
Robert
I found a public copy of the paper you referred to here: http://www.comsoc.org/techfocus/pdfs/sb10g/55tcomm09-lee.pdf?id=11803004
This is the critical sentence (not in the abstract, but in the conclusion of the paper itself – my bold):-
“… symmetric data rates in excess
of 1 Gbps or GDSL may be possible for four twisted pairs over
a 300-m cable length.”
What this sentence is referring to is 4 twisted pairs ie. 8 strands of wire.
What your uncle describes is a typical piece of wire to the house in Australia – 4 strands comprising only 2 twisted pairs.
ie. only half the capacity required for gigabit DSL.
Now. In my case I have 3 such wires (4 strand, 2 twisted pairs) installed. Most houses have 1 such wire installed.
I have enough strands (with caveat below), but the vast majority of premises do not.
Now the caveat. Having enough strands is not enough in itself to make this work – those strands need to be inside the same cable, and they need to be of high quality.
The paper goes on in the second paragraph (about half way through) to say (again my bold):-
“Coordination of the signals in the final drop with good design practices known as vectoring and bonding allow the entire drop binder of two to six wires to be viewed as a single transmission path that can have enormous capacity”
(ignore the “two to six wires” thing for a moment, I’ll explain that in a sec)
What “drop binder” means is that all the wires that are to be “bonded” into a single connection have to be inside the same piece of cable. ie. the arrangement of 3 cables that I have into my house cannot be used. The technique described relies on a single physical cable comprising several twisted pairs inside the same sheath. Very much like CAT 5 LAN cable.
My 3 pieces of cable – each comprising 4 strands, aka 2 twisted pairs – cannot be bonded because they are not physically close enough (for instance they hang a few inches apart as they go across my front garden) and their distance changes (ie. they sway in the wind). Because the distance changes, the noise between them is not constant and cannot be cancelled out to achieve high speeds. They cannot be bonded because they are in separate sheaths.
Now your uncle is right in one respect. In each cable run into an Australian home there are two pairs that can be bonded.
But. You will not get gigabit ADSL out of only 4 strands, you need 8. Now CAT 5 contains 8 strands, but CAT 5 is not what delivers telephone service to houses in Australia. Every phone service in Australia is delivered over CAT 2 (I think that’s right, someone check me on this) which differs in two respects:-
a.) there are only 4 strands, not 8
b.) those strands are not “twisted” but are laid next to each other.
The second point is important. If you cut open CAT 5 you’ll find 8 strands each individually tightly twisted against another in 4 pairs to reject noise. If you cut open your telephone line you’ll find 4 strands loosely laid against each other – not twisted. This arrangement also rejects noise, but nothing like the level tight twisting does.
This effect is known in electronics as “common mode rejection” (a term the paper uses) and it just means that two wires lying next to each other carrying AC signals will shield each other from noise. The more you twist however, the better noise protection you get. CAT 5 twists pretty hard. Common POTS telephony cable doesn’t and is much noisier
So bottom line take away:- yes you may be able to get gigabit DSL on “twisted pair” but you need totally different cable from what is installed in Australian homes. You’d need to replace it all. Firstly because of a.) there aren’t enough strands and secondly because of b.) even if there were, they aren’t of high enough quality cable – they’re not twisted enough.
So you might as well use fibre.
Sorry, I know this is a techie retort, but sometimes when techies object to so-called “business focussed solutions” they’re actually right – the so-called solution simply won’t work.
JM: fair enough.
However, is VDSL the technological limit of what we can squeeze out of current telephone cable? I doubt it.
Robert, Its time to get off that steamship and embrace the wings. JM is talking of the basic physics here, it is an excellent exposition. No amount of frigging with protocols , etc., will change the fact that the copper wires were never designed for high speed data.
They were:
1. Never designed in the first place for high speed data, 3 kHz @-3db bandwidth was considered adequate for telephony back in 1920-1950 something when much of the system was installed.
2. Never laid with data in mind, thus they will be full of kinks, impedance shifts and reflections
The short answer to your question is that it is a nonsense question, its like asking if you can drive at 200 km/h down the length of every road in Australia. The answer is that on some of them you can , but on most of them you can’t, and if you try you will most likely kill some-one or yourself-on the first corner.
You can ACRONYM away as much as you like but it will not save you.
The FTTP network is way overdue and you should stop faffing about, admit your error and start supporting progressive ideas instead of all that reactionary ranting.
You are a leader of men and it is about time to assume your responsibilities.
Huggy
“is VDSL the technological limit of what we can squeeze out of current telephone cable”
Possibly not, but it could very easily be near the limit.
Personally I think we should be careful of technological cargo cult-like arguments that a technical fix will be along real soon now. We’ve suffered from that in the past.
Originally (ie. 20 years ago in the token ring days) UTP data was promised to be deliverable over existing telephony cable (which is CAT 1, not CAT 2 – sorry I got that wrong). But when the technology actually came to market business found that they couldn’t use their existing premises wiring as promised but had to re-wire.
Another example is the technical debacle of digital TV. We were promised that we could retain our existing antennas and so that technological nincompoop (I’m being polite) Senator Alston ran with a technically inferior solution.
What was left unsaid was:-
a.) the new digital channels would use frequencies in between the existing analogue channels meaning that the digital signal strength had to be attenuated by 15db to avoid interference
b.) because of a.) you often have to install a masthead amplifier and lay new higher quality antenna cable inside your house
So while we can thank Senator Alston for saving us $50 for a new antenna, he saddled us with about 4x that cost for new wiring inside the house.
Beware of old technologists bearing buggy whips. And be especially dubious about cable manufactors promoting nifty products that look cheap but aren’t. It’s a basic bait-n-switch “yes you can use your existing infrastructure, …. oh whoops, no you can’t. Sorry, but I can do you an excellent deal on my new improved buggy whip cabling”
Robert, wot JM said. If you allow that the copper network needs replacement sooner rather than later it makes sense to replace it with FTTP. Glass Fibre is less embodied energy and CO2 than copper (probably) and is certainly made from a plentiful and low cost material. The energy cost per bit is far lower than copper (like thousands of times lower).
Perhaps we could have an opt out clause where for a fee you can retain the copper pairs, a liveried flunkey arrives every time you have message and hands you the mouse. Would that make you feel more comfortable?
Huggy
…and meanwhile no-one has really come up with any reason why we need even 100mbps internet connections at home – other than a possible need to watch 4 HDTV streams at the same time.
So, in summary – we are spending at least $42 billion dollars (I would be very, very surprised if it come in at this figure) to get:
1. A pile of debt and a need to pay interest and dividends on it at about $5 Bn per year.
2. The current phone and ADSL system becomes a complete and utter throwaway proposition – along with (possibly) the cable system – meaning still more expense as it is all written off).
3. A system that could well be a throwaway itself if and when wireless is improved.
4. But we can watch 4 HDTVs all at once!!!
Don’t get me wrong – I have worked in countries with 100mbps at home as standard (Korea while I was there) and it is wonderful to have. Great stuff. I am just not comfortable with having to pay an absolute minimum of $80 per month (and probably much, much more – no one can really be sure) to get it. And even if I personally do not get it, I will have to pay for it through my taxes.
“…and meanwhile no-one has really come up with any reason why we need even 100mbps ”
Because there is a technological arms race that is driven by websites not by providers.
About 10 years ago your typical banner site (such as a newspaper site like The Age or NYTimes) would load 1 page and a few ads (ie. about 10-20 http requests). Nowadays the same sites make 50 or 60 or more http requests to load their front page. And the images are larger than they used to be.
What this means is that 10 years ago you could feasibly load such sites over a 33k connection, now you can’t you need ADSL or better.
If you’re still stuck with a 33k modem you are effectively shut out of the modern web because much of it is so slow to load for you that it is unusable.
In other words, if we are to retain the utility of the web we have to keep up with the capacity required to access it effetively as web site providers rapidly scale up their delivery. Every year you stay with low speed access is another year where the utility you are able to derive from the web falls because you are becoming to slow to keep up.
In technology you can’t assume that the utility of something is a fixed value, it is constantly changing. Also costs are constantly changing. Very new technology is very expensive, but so is very old technology. There is a sweet spot in the middle where costs are reasonable and utility good enough, but that sweet spot moves all the time.
This is typical of the technology industry and the metaphor of a surfing a “wave” is often used. You can get too far ahead and get dumped on by committing to a technology that goes nowhere, but if you fall too far behind you fall off the wave and have to paddle hard (spend a shedload) to get back in motion.
The best option is to stay just in front of the crest where you will be carried forward without undue effort or expense (aka “state of the art”). Battalions of tech companies and consultants are engaged in helping companies make these decisions and because they are complex decisions business finds those consultants/magazines/journals efforts worthwhile. Expensive, and like advertising you’re never sure which advice actually works, but worthwhile.
But the classic mistake is to pretend the arms race (“wave”) doesn’t exist and proceed from first principles and assume that requirements will never change. Otherwise you end up saying something like “640k ought to be enough memory for anybody”.
Just another comment on the “wave”
The wave is actually Moore’s Law where technological capacity doubles every 18 months. So for a given investment you’ll get twice the bang for your dollar in 18 months.
But Moore;s Law is just about 100% misunderstood and misrepresented. It’s treated as if it were a law of nature, but it isn’t. It is really the force of the deliberately adopted capital investment strategy of Silicon Valley.
Back in the 1960′s the chip companies in the Valley were faced with a serious financial problem. The capital investment in chip fabrication plants was very substantial but they could be obsoleted by your competitors very quickly. So if you were running Fairchild or Intel you needed to create new plants frequently but you couldn’t be sure how long they would remain viable.
Moore came along and said “look, we can run this race as fast or as slow as we like, but the problem is our competitors might run at a different speed. Either they’ll be faster and obsolete us, or they’ll be slower and have more time to depreciate their plants. So what’s the optimum speed that gives us the best balance of return on capital and risk?
Answer 18 months”
So Fairchild and Intel committed to a technology cycle of doubling the number of transistors on a chip every 18 months. They planned this as a deliberate strategy to optimize use of their capital.
They could have just as easily chosen 3 years or 7 years as the cycle, but they didn’t because Moore made the point that 18 months was a pretty fast product cycle that put their competitors under design stress whereas 3 years would allow competitors to leapfrog them.
This observation was so astute that the whole industry adopted an 18 month product cycle. You build a plant, produce version 1 of your product. The learning curve means you get the bugs ironed out after 18 months so version 2 can come out in 18 months from the same plant. After 3 years you upgrade the plant, do another two 18 month cycles and presto – 6 years of depreciation after which you wheel out your new plant that doubles the number of transistors again and so you start another cycle.
Because of the centrality of technology to the modern economy this investment strategy – derived solely from financial considerations – thus drives the modern economy.
That’s where we are – on a treadmill designed by the venture capitalists of Silicon Valley to maximise their returns. And we don’t have much choice in the matter.
Robert: given the choice, let’s have both.
The solution you’re proposing was the sort of thing Telstra should have done five years ago, had Trujillo been what he promised and seen technology, rather than government relations and spin, as his value-add. If Telstra did this now, it might cast doubt on the fibre rollout but would ultimately enable Telstra to stay in the game.
Again, as I said on Quiggin’s blog – if Telstra has a problem that puts it at a commercial disadvantage, why not urge it to seek a commercial solution? Why the assumption that Telstra’s problem has to always be our problem? The Howard assumption that it could be privatised and, at the same time, be closely controlled by government is absurd. Doesn’t work for the ABC or even, apparently, the Defence Department.
If you’re going to propose a cheap compromise on the telco issue, you can’t complain about cheap compromises on emissions, workplace regulation or any other issue you care about.
“3. A system that could well be a throwaway itself if and when wireless is improved.”
With respect, that is seriously ignorant nonsense. Wireless can never compete with fibre optic for overall bandwidth capacity. That is just basic physics. You do your case no favours with foolish statements like that.
In particular wireless can’t compete for expandability. Not even close. The current whizz-bang 3G wireless network, which was supposed to solve all problems with wireless, is already getting crowded and adversely impacting performance. The absolute capacity (and reliability) of wireless is inherently limited, no matter how clever the engineering. By stark contrast, even currently available fibre optic technology is, for all practical purposes, infinitely expandable. It provides a level of future proofing that no other currently available or reasonably foreseeable technology can even hope to match, let alone seriously surpass.
Wireless is most certainly not the answer for the vast majority of fixed location users. Never will be. To describe current fibre optic as throwaway (especially in comparison to possible future wireless) is a truly ludicrous assessment.
And if you think that this is all just about getting 3-4 HD channels online, then you have a remarkably short-sighted and ill-informed view of this technology’s potential, and you are a good example of why economists and financial advisers should not be left to make the final call on these decisions, coz all too often they just can’t see past some superficial, abstract, and often highly debatable numbers on a page out into the real world effects.
Well said Hiker and JM.
JM,
Are you seriously saying that web site owners will develop web sites that require 100mbps speeds and are heavier than the capacity of people to receive them? The US, for example, shows no real sign of standardising on 100mbps broadband any time soon. Your contention, therefore, is that web site owners will be content to cut the US, Canada and most other countries that are not (yet) moving to 100mbps broadband from being possible consumers of their content. That would be frankly bizarre.
.
Hilker,
I notice you avoided points one and two – the ones with unambiguos hard evidence. As for my point 3, frankly (and this is why I hedged) I do not know what will be best in the future – but neither does anyone else. In the mean time we are to spend a minimum of $42bn dollars and throwing away most, if not all of the current infrastructure to go to a system that, while I would personally enjoy immensely, I cannot see any real need for. We are all going to pay for this, whether we want to or not. That means your tax bill (if you pay any – I do not know) for years to come will be higher than it needs to be to finance a wonderful possible white elephant.
Sorry – you may be allowing your desire for teh fastest broadband in the world! to blind you to the simple fact that we are all going to be paying for this, even if we do not use it, or I may be blinded by the staggering amount of money this is going to cost, but I just cannot see the value of this.
Test
[moderator note: your previous comment and this one got diverted into the spaminator, AR. I don't know why. ~tigtog]
Thanks, tigtog. I thought it may have been, but I did not think that “Test” was the sort of comment that got diverted. Feel free to delete this one and the “Test” one.
Andrew Reynolds,
Surely the real issue is that the present network is really old and not up to the job? That is the real subtext here. Really high speed networks are the most trans formative technology we could possibly invest in. I am no brave new world starry eyed Utopian but I see this technology as fundamental as road and rail transport. The copper network is just a series 14th century bridle paths through the the dark forest. You may enjoy wandering on foot down these narrow pathways, but it is pure self indulgence.
The positive economic consequences of FTTP will be enormous. I have listed some of them in previous posts.
Huggy
Andrew E: I’m prepared to do a whole lot of cheap, ugly compromises on emissions provided we get the necessary emissions cuts done.
Huggy,
It depends on what job you want it to be up to. It is perfectly capable of carrying data at around 10mbps over most of it right up to the premises. It is also constantly being upgraded – by all of the players. Over the next 10 to 20 years it may be upgraded to the point where it can do what this hugely expensive new system is meant to be able to do.
Do we, as a nation, really need to spend an enormous (and uncertain) amount of money as well as throw away the current system just to get just to get something that has nebulous potential future benefits? To me, that is fine if a commercial company wants to take that risk and potentially lose billions of its shareholders dollars if many people do not want it. Personally, I would prefer that dollars that are to be extracted from me by force not be used for something like this.
To me, the money would be better left in my pocket for me to spend if I judge this as a useful addition to infrastrcture. I could use the money to buy shares or debt instruments. OTOH, if I judge that I do not need to be able to view 4 uncompressed HDTV streams at once I could use it to donate to a worthwhile charity – or get a nice big LCDTV to view the 1 uncompressed HDTV stream that I could get through my current connection – or through the cable connection I already have.
Andrew, yes I am seriously saying that website owners will develop websites requiring faster connections because that’s exactly what they have been doing for the last 20 years. Back when the web started in 1993 I had a 2400 baud modem (and that was considered fast for the time) which was perfectly adequate for the email and web services of the time. Now, I couldn’t even load google’s home page with it (and that’s a comparitively simple page). Why do website owners do this and ignore customers with slow connections? Because customers with slow connections are poor. The only reason why the US “hasn’t moved to 100 Mbps” is because a.) they have a solely market based approach to infrastructure with little government involvement and b.)they are moving to faster infrastructure, they’re just relying on richer consumers to drive the market.*
Secondly, I guess you use Amazon? A US based service. Which has a notoriously heavyweight home page and 15 years after its inception is near unusable without broadband.
The more important point is that other services – web payments for example – won’t work without reasonable speed connections and future services of equal or better utility can’t be created without better network infrastructure. I’ll give you an esoteric example. You’re aware of EFTPOS payments via the little terminals in nearly every shop? They were established over a slow, old fashioned X.25 network run by Telstra. But if you walk into a 7-11 these days you’ll find a terminal that runs over TCP. (You can’t tell by just looking at it so you’ll have to take my word on this.) This new payment service is not universal but it is growing and Telstra are losing market share.
And those terminals are also capable of delivering other services such as selling recharge vouchers for your pre-pay mobile phone. Those are new services created in the last 10 years and made possible by the replacement of older telephony with TCP – something only made possible by the large investments driven by the internet. The financial reason for this is that Telstra charge customers about 10 cents for each 200 byte transaction over x.25 whereas the same packet is near-as-dammit free over the TCP/internet infrastructure.
Voice-over-IP is another example where exactly the same thing is happening. TCP/IP based services are massively cheaper than the old telephone service (something Sol knew very well and spent his entire time trying to distract people from while replacing the core network with TCP so he could secure Testra’s monopoly on the one hand while benefiting from the cost reductions of new technology on the other)
Your argument boils down to this: “I [Andrew] can’t conceive of why anyone would want faster connections, so therefore it’s a bad investment”. The Argument from Personal Incredulity in other words.
My argument is a little subtler. The only reason why you find current technology cheap is because everyone else is at the same point – on or near the crest of the wave. If you personally want to stay with 1M ADSL you’re quite welcome to. But as that technology becomes obsolete, you’ll find it becomes increasingly expensive for you to maintain it, until you can no longer buy ADSL-1 modems (because they aren’t made anymore). You’ll end up bidding on obsolete equipment on eBay, which is something NASA have resorted to in recent years to get spare chips for the Shuttle. I know from personal experience that Telstra have also done the same thing to keep parts of their older network up and running.
Your argument about debt is perfectly valid, but I think we should all be aware of the trade-offs before we blithely assume that some market-based mechanism is suddenly going to bring technological advances to a grinding halt. You may want to let Australia’s telecommunications infrastructure degrade into the future equivalent of Portugal’s rail system (if you’ve ever been there you’ll know what I mean), but I don’t.
But the thing about debt is that the investment is supposed to pay off. So unless you’re seriously suggesting that the almost no-brainer of investment in technology infrastructure – which has paid off ever since the invention of the steam engine – is suddenly going to stop I think you’re ignoring the positive side of the equation.
* In any case, they have less need to. 10 Mbps cable is everywhere in the states and this is conveniently overlooked by those who point to the lack of high speed DSL in the US.
” It is perfectly capable of carrying data at around 10mbps over most of it right up to the premises.”
No it isn’t. ADSL-1 maxes out at 1.5 Mbps and is only available within a couple of kilometers of the exchange. ADSL-2 which is necessary for 10 Mbps is not available at all exchanges (not mine for example and I live in one of the wealthiest suburbs in the country) or all telephone lines even then. Many customers are stuck with bad lines and can’t get DSL even when their neighbours have it. Telstra’s response in those cases is “tough”. They won’t upgrade the line unless you pay for it.
VDSL requires that you be even closer to the exchange.
And Robert’s super-VDSL requires recabling and even then will only reach 300 meters from the exchange maximum. (I didn’t say this before but the paper Robert references is describing a theoretical model that sets an upper limit on cable performance. There is no experimental data in the paper. So the 1 Gbps speed at 300m is the upper theoretical bound and probably not acheivable in practice.)
JM,
I currently have 17mbps* cable to my home, over which I am running a 4 client network, a windows 2003 server handling my email as well as a small, low volume web site. I switched off the ADSL2+ connection I had as it was too unreliable.
Personally I would love to have a 100mbps pipe to my home. I also have 2 VOIP connections coming through that pipe – and the quality is better than my landline.
If I look at your examples I see none of them come even close to needing a more than 1mbps – voip works perfectly well at 64kbps. POS terminals need even less. Amazon loads quickly at home as well.
.
I am not saying I cannot see any requirements for it. What I am objecting to is simple – the use of taxpayers’ money to pay for a particular solution that may or may not prove useful to the general public. If I were entirely selfish I would be really happy to have that pipe to my home and have everyone else pay for it. I like to imagine that I am not that selfish, though. I can virtually gauarantee it will not cost anything like $42bn – particularly after you consider the accellerated run-down of the old network.
.
.
*theoretical limit. Practical limit between 6 and 10mbps
Andrew, we have a figure for the cost of ” the accellerated run-down of the old network.”
$42 billion.
Most of that cost is for recabling. All I’m saying is that the new cable should be fibre.
FFT-node costs 4.5B. Recabling delayed until it is carried out patchwork in the future and paid for by consumers directly and individually. But they will pay for it as the cost cannot be avoided.
Robert suggests we can get faster magic to avoid replacing existing copper. I’ve pointed out that we can’t for various pretty serious technical reasons. There is no magic that will deliver fast connections on the existing copper. In restricted circumstances (close to the exchange, good quality wire) yes we can do it. But only up to a point (which we are pretty close to reaching)
And as a national infrastructure, no. Sticking with the existing copper will fragment the market and hinder development – the opposite of what infrastructure is supposed to facilitate.
So we have to recable. We can use CAT-5 as Robert suggests, or we can use fibre but the cost will be about the same. The technical choice is clear. For the same money we can have the Model-T or the Ferrari. They cost the same, take your choice.
But as you rightly point out, we have a policy choice as to how fast we spend that money. Do we spend it at once in a big national rollout or do we let the market do it, probably slower and with patchier results leading to slower and patchier economic development.
The government is absolutely correct. FTT-node will waste the major portion of the initial investment on switches at the node, which then simply get thrown away when fibre is ultimately rolled out to the last mile. If you’re going to throw something away surely it’s better to throw away the fully depreciated 50 year old copper than the 5 year old switches you put in as a stopgap?
FTT-premises is more expensive up-front but gives a better result faster. Take your pick. Personally I think FTTP is the better way from all perspectives.
Spend a little or a lot now is a valid debate. But it’s not a technical one and hiding behind the technical sophistry of monopolists and buggy whip cable makers is not the right way to approach it. Particularly when in the wash-up copper will turn out less functional and give less payback.
What I object to is the assertion that an obsolete (because it is near its limit) and expensive technology – copper – is somehow in anyway equivalent to a cheaper and much more capable one (fibre). The promotion of copper is simply a stunt where the entrenched interests of old technology are trying to milk their investments for as much as they can get at our expense.
” I can virtually gauarantee it will not cost anything like $42bn”
I can guarantee you it will cost every bit of that. The cost is in sending a man out in a van to dig a hole and lay a piece of cable, not what the cable is made of.
The real choice is between spend-it-fast or spend-it-slow. And the only way to make that choice is to look at how fast you want the payback.
JM, well put. why all these really old farts want to keep their copper cables is way beyond me. Do they think it will fix their arthritis?
Huggy
Huggy,
With your own money on the line perhaps you could get your arthritis fixed.
.
JM,
Mostly correct in your main analysis, but (IMHO) not in your conclusion. From my reading of this you are happy that every taxpayer in this country should pay for this and take the commercial risk that comes from it, rather than the shareholders, creditors and customers of the companies that would otherwise do so. Fine – if you are happy that people who will not and will never use it are also paying for it and taking the risk.
Personally, I am not.
As for the cost – given the record of such projects in the Government or private sectors I would be willing to give at least a 10 to 1 bet that the cost comes out at least double the original estimate.
As a taxpayer, I’m happy to pay for it. I’ve seen the lack of fairness in the current access issues. Telstra has abandoned my area…and the former governments solution was to offer grants for satellite dishes for people where Telstra infrastructure was inadequate. That grant was for over $3000.00 if I remember correctly…it might seem an attractive option until you realise that the lack of competition in the satellite market, meant you would still pay a minimum of twice what ADSL customers were paying, and for a very energy intensive solution, I was told the dish had to be constantly on..humming away in your lounge room. I wonder if all the advocates of keeping the copper system are even aware of the costs of sticking with that system? Not to mention the issues of equal access if you simply let market forces dictate where technology is upgraded. Sheesh, I already live in a technology ghetto, who knows how bad it’s going to be in 5 years time?
furious balancing,
You may not have noticed, but the new solution provides even less coverage than the original FTTN solution (90% as opposed to 98%) and satelite for the rest. This means that you are more, rather than less, likely to end up with that satelite dish in your loungeroom. Almost none of WA except for Perth and possibly Bunbury, for example, will end up with FTTP.
JM 80,
“ADSL-1 maxes out at 1.5 Mbps and is only available within a couple of kilometers of the exchange.”
ADSL-1 available in Australia maxes out at 8 Mbps, and is available up to 5.5km from the exchange. (I lived about 4km away and still got download speeds over 2Mbps)
“VDSL requires that you be even closer to the exchange”
You’ve said that several times, but not with FTTN it doesn’t (VDSL doesn’t plausibly exist without FTTN). It requires you to be 300m from the SAI to achieve the full *symmetrical* 100Mbps (the 4-UTP GDSL outlined in the paper is also symmetrical btw), and then at 1.5kms – more or less the extent of an FTTN network – that speed drops to half.
So $4b for FTTN, another $40b for the last mile (and I think you misread what Andrew Reynolds meant when he said “it will not cost anything like $42bn”).
“JM, well put. why all these really old farts want to keep their copper cables is way beyond me.”
Huggybunny, do you think all that copper cable is going anywhere any time soon? We were given 5 years (5 years after full DTV implementation) before analogue TV is to be switched off completely. It will be 25 years at least before copper telephone lines are even thought of being phased out.
Most likely now is that in 5-10 years we’ll have a choice between paying $80-100 a month for fiber at 100Mbs, or $10-15 a month for copper w. unlimited usage at good old 1.5-24Mbs, maybe faster, who knows? – and because Telstra will continue to offer those cheaper plans, so will every other company. I know which I’ll choose – and I also know which my mother would choose.
Andrew I agree with your conclusion re. cost overruns (although I think these things are under much better control than they were say 20 years ago)
The argument about public versus private risk however is much more of a philosophical one.
You’re right, personally I am happy with the public carrying the risk for this type of infrastructure investment, and I have a – probably eccentric – reason for that. I think government can capture more of the return from the investment whereas private interests cannot. Therefore government investment will be cheaper overall. Please feel free to shoot me down on this, I’m not an economist and I’m probably missing something important.
The government also borrows for less so there is less cost to me as a taxpayer than as a consumer. The only way you can make your argument stick is to assert that the government is incompetent and will stuff it up technically. My contributions to this thread have been aimed at explaining that I don’t think they’re stuffing it up, and rather that Telstra were with FTTN. They were selling us a technical pig-in-a-poke. They’re a monopoly, they have good reasons for what they were proposing. But it doesn’t mean we should pay for it.
JM,
I am going to try to shoot you down on the overall capture issue – if the private company cannot capture more of the return than the government (a point I would debate) then the “uncaptured” returns would go to one of the consumers (some of us), the debt holders (likely to be many of us with super accounts) or the government (through taxation). They are not lost down some rabbit hole.
I would encourage you to do a basic risk analysis, though. Given the history of this sort of thing cost over-runs (which continue to be a serious problem in this sort of project, particularly as there will be a lot of assumptions in the cost model) are likely to be high. As most of the equity investment is to be government funded, that risk will be likely to fall almost entirely on the taxpayer. As the cost goes up, less and less people are likely to use it – even on current costings and projections the minimum cost before considering the writeoff of the copper network is likely to be over $80 per household per month.
The prospective costs to the taxpayer of a system like this which has, at best, a nebulous economic benefit is, IMHO, just too high.
I have no problem with a commercial company taking this sort of risk. Taking that sort of risk with my tax dollars is, to me, just not on.
Sally, re your fourth paragraph. You’re right to ping me on the “closer to the exchange” bit. You’re right VDSL can service many premises because the relevant distance is to the node not exchange. Sorry.
I won’t quibble with you about what ADSL-1 actually means, but point you to this page from Internode http://www.internode.on.net/residential/internet/home_adsl/extreme/
The theoretical limit for ADSL-1 (as noted on that page) tops out circa 2-3 Mpbs at best at around 4 km. ADSL-2 gives you better theoretical bandwidth but no better distance..
In practice people usually quote 2 and a bit kilometers.
The takeaway is that no ADSL technology (2, 2+ whatever) reaches beyond a theoretical limit of 4k
VDSL is even shorter but IMHO corresponds to about the practical limit of copper – certainly the limit obtainable by any existing technology – so I think I’m being fair in saying copper is near it’s limit.
The GDSL paper Robert references is talking about the theoretical limit of CAT-5. They don’t come right out and say that’s what they’re describing but they do say:
“AWG-25″ Check. That’s the gauge of the strands in CAT-5 (and everything else to be fair)
“twist rates of the four pairs are 3.5, 4.6, 4.4, and 4.2 inches.” Check, that’s typical CAT-5
“each pair is differentially terminated with a 100 ? impedance” Check, that’s CAT-5
“4 pairs”. Check. That’s CAT-5
It then says Theoretical Limit at 300m: 1G. But that’s only theoretical, it’s not any delivered technology and it assumes CAT-5 like cable to the premises.
So to get that we have to be within 300m of a node and install new wire.
Why shouldn’t we just put in fibre?
Andrew: “They [the benefits] are not lost down some rabbit hole.”
No, I agree. But they are lost to the investor. And if they’re lost to the investor the appetite for the risk must be lower, so therefore slower rollout, lower benefit to society and so on. My personal view is that this is an outstanding example of where government can achieve better outcomes for us as a whole.
Sorry I forgot. If you look at Internode’s page you’ll also see that ADSL-2 only delivers its theoretical maximum rates up to about 1.5km and drops pretty rapidly after that.
Copper is a dead end and it has nearly run out of road.
We should get a clue when we’re reduced to arguing about how much market penetration we can get with a 300m* cable budget. We aren’t going to get 98% with that, no way.
Remember, Telstra’s service obligation is only for “basic telephone service” so don’t be deluded into thinking that just because you can force them to run a bit of wire 50km to your farm for voice communication with your local doctor doesn’t mean you can force them to provide technical magic as well.
* which dissappears very quickly when you have to go around corners and up stairwells.
Had a discussion about FTTP today with one of Brisbanes most experienced network guys. The consequences of a really fast network are just mind bending.
LP will make them available free as an incentive to visit 
No need for any software purchase at all by the computer owner would be number one benefit (you want to use MS Word ? well use mine or better still just go to the LP site and you will find all the free tools you need
If you wish you could invest in some higher end hardware and offer to rent it out – like a cyber landlord. Location will be unimportant.I see a huge increase in the sort of free software that allows you to simulate the application of some vendors product.
People will form common interest groups and network their computers with real time full video for discussion and exchange of ideas, art, literature,videos and all that sort of stuff.
In short the economic and social activity of the populace will increase massively.
I predict that entirely new business models will emerge from what is a huge leap in total computing power.
As an economic activity generator the 40 billion is really a good deal. The only people who fear this are the copper wire owners and some incumbents.
Huggy
Robert@77: what if you just got the cheap, ugly compromises anyway? Yes, flowers grow in shit but having a whole lot of shit doesn’t guarantee flowers. Be careful what you wish for.
HuggyBunny,
Add to that list the taxpayers and anyone else who will have to foot the bill.
.
Having thought further through this I think I can be confident in two things.
1. In the unlikely event that even a single sod is turned on this, it will not be built to anything like the specifications given.
2. It will cost vastly more than currently thought – for examples look to almost any government project as I am trying to think of any that have come in on budget (although I am sure there must be one or two).
3. Benefits will take much longer to arrive than enthusiasts seems to think – for example look to most IT projects.
.
I would be really happy to be proven wrong, though. Running 100mbps data to the house would be cool.
“We can use CAT-5 as Robert suggests, or we can use fibre but the cost will be about the same.” etc
JM, forget GDSL – it was acknowledged up-thread that if it means upgrading the cable into every house in Australia, fiber is far and away the winner.
But VDSL2 is designed to work over 1 single twisted-pair of copper. What if the government had decided to implement FTTN, and make VDSL2 available to 90% of Australians, say at 150/50Mbps. At a distance of 500m, that drops to 75/25Mbps. At 1.5km, it drops to 24/8Mbps, which is still faster than the theoretical maximum ADSL2 can deliver.
All for only $4.7 billion. Those who want to upgrade can spend $200 on a VDSL2 modem. Those who don’t, stick with their ADSL/ADSL2 – which, regardless of what happens, is going to become very affordable in the near future.
Over the next 20 years, and as fiber speeds increase, the companies that wish to sell FTTH, install it from the node (which simply has two switches) to the home on a customer-by-customer basis – similar to Verizon in the US which has more than 2 million customers on FiOS, but with much less overhead. 5 to 10 years from now, the government might decide to spend another few billion (at most) installing FTTB to hospitals and schools.
“The real choice is between spend-it-fast or spend-it-slow.”
The real choice was between spending $4.7 billion to deliver very high speed broadband to 90% of Australians, and spending $43 billion to the same, but for every Australian to effectively subsidise the telecommunications industry to the tune of a minimum of $2000 each. Why?
“… forget GDSL – it was acknowledged up-thread … that fibre is the winner.”
If it was I missed it, but I’ll accept the point.
“… At 1.5km, it drops to 24/8Mbps, which is still faster than the theoretical maximum ADSL2 can deliver.”
Strictly speaking its the same. ADSL2+ maxes out at 24 Mbps at approx 1.5 km. So you’re acknowledging that copper – whether it’s using ADSL2+, VDSL or VDSL2 – is limited in range.
In other words, if you can’t get ADSL2+ now you re sure as the sky is blue not getting VDSL2 in the future (nor VDSL which is what we’re discussing)
” … the companies that wish to sell FTTH, install it from the node (which simply has two switches) to the home on a customer-by-customer basis …”
As a spend-it-slow evolutionary strategy that would be fine (sort of, leaving aside coverage arguments and risks which I’ll get to in a sec), except that Telstra have spent quite a lot of effort over the last few years preventing and arguing against exactly that sort of access to their cabinets. Sol and the boys insisted loudly about 18 months ago that hell would freeze over before they let any spotty little reseller into their equipment boxes. (That’s leaving aside the fact that there isn’t room in most cases for the second switch you talk about.)
If you’re not worried about the monopolist’s ability to retain its monopoly that’s ok. And if Telstra were still in public hands – or alternatively, separated out as a wholesale provider separate from retail – I wouldn’t regard that as a problem, but Telstra is no longer fully public and Telstra’s behaviour is a problem.
But you’re still stuck futzing around with bleeding edge technologies that have very little rollout (or may not even exist outside the lab) in the futile pursuit of ever-diminishing benefits because of the limitations of copper.
That would give us two risks – real ones, not ideological ones:-
a.) we get stuck backing a technology without a future so we become a backwater market and have to pay too much for equipment that never reaches any economy of scale and eventually ceases manufacture. (CDMA was a great example of that trap)
b.) we deliberately limit ourselves to a service ceiling somewhere around 24 Mbps and 1.5km
Because if we get sandbagged by either of those we’re still up for recabling with fibre a few years down the road having wasted most of our money in the meantime.
Now Sally, be honest. Your argument is more or less this:- by extending Telstra’s core broadband network from the exchange to the node (which is what all this VDSL stuff is about) you entrench the monopoly. You’re apparently not worried about that. Secondly, you argue that placing the extra cost of fibre to the premises on the end-consumer makes those costs disappear. But they don’t disappear. You’re comparing the apple of 4.5B for FTTN with the orange of 42B for FTTH.
Lastly you’re pretending that the monopoly will be just fine with 100 little companies crawling all over their cabinets. You can’t maintain that pretense in the face of Telstra’s outright, public and loudly expressed opposition.
This is a really, really simple case of control of infrastructure and how monopolies maintain themselves.
If you want a free market – you have to make sure it is physically possible
Because if you allow the monopolist to maintain control by keeping the keys to the cabinet it is not going to happen.
(BTW – you sound like you know this stuff fairly well. Do you have an interest? Beyond a few thousand Telstra shares, I don’t)
JM, I’ve no interest in the tech except as a techie, and I don’t own shares in anything except a recently acquired kitten. I spent about 3 hours reading to get up to speed on this.
“So you’re acknowledging that copper – whether it’s using ADSL2+, VDSL or VDSL2 – is limited in range.”
Yes. Short-range is exactly where it continues to be a viable and competitive option, and will be around the world, for at least another decade. That’s not something I’m trying to sell you on – that’s something the larger telcos around the world, including Australia, have already well and truly sold themselves on.
“In other words, if you can’t get ADSL2+ now you re sure as the sky is blue not getting VDSL2 in the future (nor VDSL which is what we’re discussing)”
1) VDSL has been superseded by VDSL2. Providers that rolled out VDSL did so circa 2005. No provider that hasn’t yet rolled out VSDL would consider choosing it over VSDL2 in 2010-11. That’s a moot point.
2) Very different to whether your exchange offers an ADSL or ADSL2+ DSLAM. Again, FTTN and VDSL work hand in hand. To qualify as FTTN, it must bring fiber to within 1.5km of the premise, and offer DSL (or Cable) the last mile. The DSL could be ADSL2+ if you really wanted, but who in 2010-11 would choose to purchase and install tens of thousands of ‘brand-new’ fiber/ADSL2+ converters around the country? 90% of Australians would have received VDSL2.
“But you’re still stuck futzing around with bleeding edge technologies that have very little rollout (or may not even exist outside the lab) in the futile pursuit of ever-diminishing benefits because of the limitations of copper.”
See Wiki’s VDSL2 page for a list of countries whose telcos commercially rolled out VDSL2 in the last two years.
“we deliberately limit ourselves to a service ceiling somewhere around 24 Mbps and 1.5km”
Quickly, ADSL/ADSL2 from the exchange, sharing congested lines with 1000s of other users, versus VDSL2 from the node and sharing it with 10-100s of other users, many of whom will be situated well within 1.5km. With ADSL2+, I don’t expect 24Mbps per second at 1.5km – with VDSL2, (in theory) I certainly do.
It’s not a service ceiling we would have been limiting ourselves to…
“(That’s leaving aside the fact that there isn’t room in most cases for the second switch you talk about.)”
Yes, that was most definitely a throwaway – I have no idea whether it would be simple or practical, or not
Moving on:
“Your argument is more or less this:- by extending Telstra’s core broadband network from the exchange to the node (which is what all this VDSL stuff is about) you entrench the monopoly.”
That’s not my argument.
“Secondly, you argue that placing the extra cost of fibre to the premises on the end-consumer makes those costs disappear. But they don’t disappear.”
That I would argue. It costs Verizon about US$700 to bring fiber to each user, yet they don’t charge each user US$700 for installation. Verizon is investing in its future, and already profiting from that investment.
“Lastly you’re pretending that the monopoly will be just fine with 100 little companies crawling all over their cabinets. You can’t maintain that pretense in the face of Telstra’s outright, public and loudly expressed opposition.”
That’s also not my argument.
My ‘argument’ is more or less Robert’s:
“So, in a nutshell, I reckon that we could have gotten enough capacity out of a moderately souped-up copper network, in large parts of Australia, for the medium term future. And we could have done it at much less than 43 billion dollars.
The experts advising the government presumably know this, but they’ve still recommended replacing Telstra’s copper network entirely with fibre optic. And, in my view, the key reason isn’t technical.”
JM, for many reasons I’m *totally* in support of the Government’s FTTH network, but ‘copper is a 19C, dead-end heap of shite for backwater markets’ etc, just isn’t one of them – at least not in such preclusive terms.
Over the next 10 to 20 years it may be upgraded to the point where it can do what this hugely expensive new system is meant to be able to do.
At the risk of repeating myself: Copper is reaching its limits. It can NEVER compete with fibre optic for maximum capacity and especially for future proofing. FO has no practical limit, compared to any other technology on offer. This is not even up for debate in serious engineering circles.
The real choice was between spending $4.7 billion to deliver very high speed broadband to 90% of Australians, and spending $43 billion to the same,….
No, it is not the same. FO will deliver a guaranteed minimum of 100 Mbps to each connection, with far higher rates easily achievable. Copper cannot do that. If we go down the copper road, we will just have to upgrade to full FO eventually anyway, and not that far down the track either. Why waste all that money on squeezing the last few bits per second from copper? It gets proportionally more expensive to do so the harder you drive the copper, you quickly run into seriously diminishing returns, which is what is starting to happen now (and similar problems are also arising with wireless). FO does not have that problem.
The only criticism I have heard so far of any substance is that the upfront cost for FO will be high. Yes, but not a good enough reason not to do it. It will pay for itself many times over.
This is not optional infrastructure for the 21st century. It has to be done, and fairly soon.
“That I would argue. It costs Verizon about US$700 to bring fiber to each user, yet they don’t charge each user US$700 for installation. Verizon is investing in its future, and already profiting from that investment.”
The money is still being spent and it will have to be recovered somehow. Just because Verizon don’t sting you for it upfront doesn’t mean they don’t get it at all. The only sensible way to think about it is that it will cost 42B no matter how it gets done. The only question is whether
a.) we do it over 8 years or 20.
b.) we do it patchwork and stop somewhere short of 90% coverage like we did with cable.
If we do it over a longer period of time the principal is spread out and more interest is paid but the NPV will come out to be about the same. The only problem is that the benefits will be slower in arriving (and therefore turning up in returns to the investor – whether the government or private doesn’t really matter)
If we do it patchwork, the benefits will be less because the network effect will be reduced and so will the utility of the network.
“That’s not my argument.”
That is your argument even though you’re not aware of it. The reason why telco’s around the world favor FTTN is because it preserves their monopolies.* The real source of the monopoly is connection initiation and termination. Connection in pretty near 100% of all cases involves a Telstra peice of local copper (and if it’s a phone call they get at least 7c for it even if it terminates on another network).
They are desparate to preserve that monopoly and have used every trick in the book over the last 30 years to prevent a.) access to the local loop and b.) replacement of the local loop. Replacement is not something they want to do themselves, but they are desparate to stop anyone else doing it because then they wouldn’t be needed anymore
Telstra’s corporate strategy since 1975 has been to preserve their sole access to the local loop. They’ve done a lot of flim-flamming over the years by distracting people with STD and ISD and a quite a few companies have been broken by pursuing the pot of gold in the wrong place. The pot of gold has always been the local loop.
Now if we go with FTTN there will be no free market in provision of fibre because Telstra will not allow access to the cabinet. We had that argument a while back with Telstra and they very firmly said “over our dead body”. (They also had pretty good reasons – there is no room in most cases and they can’t be expected to meet their service responsibilities if any tom, dick or harry has been in their cabinet and potentially broken something)
So that is also your argument, and even though you’re not aware of it, it’s a fantasy. It’s not practical.
So FTTN simply preserves the monopoly of access to the Service Access Points* at the end of the local copper. That’s why Telco’s want FTTN.
The new government plan bypasses all that and builds a parallel network which has two effects
1.) it kills the Telstra monopoly (and after Sol’s antics they deserve it)
2.) it makes visible the true cost of recabling the infrastructure instead of hiding it away in unaccounted for private expenditure.
The facts are we can’t continue to rely on copper, it’s run out of road and we’re already squeezing the last few drops out of it. It will cost us a lot to replace it, and there is a valid debate about how best to conduct that investment.
But throwing good money after bad to preserve copper (and the monopoly that does with it) is no IMO one of the options.
[Sorry - missing footnotes]
* And it’s interesting that Ausralia gets a guernsey on that Wiki list even though nothing has been rolled out. I’m very dubious about that list. Most are pilot projects and many are of the form “will be rolled out in 2007″, ie. it’s not very up to date and doesn’t have much in the way of recent information.
** Murdering OSI terminology here
“The money is still being spent and it will have to be recovered somehow. Just because Verizon don’t sting you for it upfront doesn’t mean they don’t get it at all.”
It’s recovered in the long run by their subscriptions, but Verizon can’t just increase their rates by x at their whim to profit more quickly. They still compete in a marketplace against every other provider and alternative technology, and the consumer is in no way ‘stung later’ by paying a competitive rate for a competitive service (quite the opposite, in fact).
“The only sensible way to think about it is that it will cost 42B [x billion] no matter how it gets done.”
That’s only true if you take for granted that 90% of Australians want the *premium service* and will pay for it. This is guaranteed not to be the case.
“That is your argument even though you’re not aware of it. The reason why telco’s around the world favor FTTN is because it preserves their monopolies.”
I alluded that it’s overwhelmingly incumbent telcos that have rolled out VDSL2+.
That this would be a bad thing for Australia, for all the Telstra-related reasons you give, is your argument, not mine, and begs the question:
What is the maximum cost to the Australian public you would accept is worthwhile to ‘kill the Telstra monopoly’?
You agreed earlier with Andrew Reynolds that it’s quite possible the NBN might end up costing twice as much as estimated – $100 billion – yet didn’t express having any problem at all with this. You also haven’t indicated any idea as to what the take-up rate for FTTH services might be.
I made clear that DSL technology is not going anywhere for decades. Telstra’s copper network is not being replaced, it’s being circumvented. DSL prices and plans are going to become increasingly affordable and attractive to many Australians and countless Australian businesses. Other service providers will (be forced to) continue to sell alternative plans which still utilise Telstra’s copper network.
I ask again:
What is the maximum cost to the Australian public you would accept is worthwhile to ‘kill the Telstra monopoly’?
What do you (and more importantly, the Government) base this on?
Sally I conceded Andrew’s point only because I can’t really argue against it credibly without getting side-tracked into nitty-gritty. But his $100b figure uses a different number – namely the nominal cost over 8 years, rather than the 2009 dollar cost as adjusted for inflation – it’s a bit of a debating trick and I didn’t want to enter into it. Nominally the cost will be greater over 8 years just as it would be much greater over 20 years of patchwork private investment. The fact is we have an estimate today of 42B in 2009 dollars and no-one has seriously challenged that – Andrew’s cynicism doesn’t either*
Secondly, the choice is this:-
1. spend 4.5B 70% of which is wasted extending the broadband network a relatively short distance and entrenching a monopoly but which increases service speeds only very marginally,
2. spend 42B all of which will refresh the infrastructure, massively improve service levels and reduce costs to end-users if only because the monopoly can be broken
As to 1.) I conceded above it will increase the market penetration a bit. And we have an estimate for that. Telstra claim 98% penetration for FTT-node, the government claims 90% for FTT-premises. ie. we get 8% more, but if your point about “everyone pays for it” is to have any force, then that’s a bit moot. If 90-98% of people want something then treating it as national infrastructure sounds ok to me. If it was only 60% and the rest of us were “carrying the yuppies” I think you’d be on better ground.
No, the fact is we are talking infrastructure here and it is legitimate for me to argue that the considerably increased bang of fibre is worth the cost over the damp squid of FFT-node.
We all pay for it either way, the question is how much we pay and how much we get. My argument is that FTT-node gives us about three-fifths of not very much so we’ll have to pay for FTT-premises anyway.
And all the handwaving about new magical protocols that don’t yet exist is just a pretense to cover up the fact the copper is limited by the laws of physics and it is just not going to get much faster. If at all, since in practice VDSL2 is no faster than ADSL2, and reaches no further from the switch. Putting the broadband switch in the node rather than the exchange is just a conjuring trick.
Anyone who tells you that a single pair of copper will ever carry significantly more that 24Mbps much further than 1.5km is lying to you. I haven’t got my books with me at the moment so I can’t do a back-of-the-envelope calculation of the maximum, but the basic thing is this:- copper as a transmission line has much, much less bandwidth than fibre. Heaps less, and unless the laws of the universe flip overnight and its dialectric constant changes it’s never going to either. And given that VDSL2 only gives a marginal improvement over ADSL2 (if any at all) I think we can safely conclude that “that’s it baby, that’s all we got”
Now let’s consider the practicalities as evidenced by the Wiki page for VSDL2
Belgium: “20 Mbps … 60% of households …. prices up to EUR85/month … planned”
France: “Fibre to the basement (ie. FTT-premises) with VDSL2 after. Pilot project to public housing in Paris.” Paris is actually rolling out fibre
to about 3 million households so this VDSL2 rollout is just leveraging that bigger project
Germany: Planned 50Mpbs leveraging FTT-curb (ie. closer than FTT-node, fibre doing the heavy lifting)
Italy: Planned but then delayed until Q2 2009
Ireland: Planned but then delayed until new announcement (not seen yet) in Feb 2009
Norway: VDSL2 only offered where multiple customers request it, max distance 800m from the exchange
Holland: Testing during 2006, nothing since
Portugal: FTT-house with VDSL2 (I think they mean Main Distribution Points within flats – ie. to the premises – it doesn’t make sense otherwise)
Spain: Tested but went with FTTH
Slovenia: VDSL2 to enterprises, also consumers by another company
Sweden: Multiple providers offering between 20-40 Mbps (distance unstated) cost about EUR50 or AUD80/month
Turkey: Announced but still unavailable
UK: Only going as far as ADSL2+, and BT prefer FTT-curb, no final decision on VDSL2
Hong Kong: Still not deployed, but 50Mbps claimed
Malaysia: “at least 50Mbps” but not when I was there last year, they were still advertising heavily 1.5M ADSL (which is what I had in my 5 star hotel)
Singapore: Being trialled
Macau: Being trialled “in 2 buildings”
Taiwan: Contract awarded for 340000 households (now we’re talking)
Canada: Saskatchewan and ATT claim large deployment with 20Mbps downstream within 900M (you have to look at the link to see those numbers)
US: Planned, planned planned for 2010
Australia: “On April 7th 2009 the Australian Government announced …. [implied rollout of VDSL2]” Eghhhh, not very credible
New Zealand: “30 Mbps $399/month” “… $50-$60/month” “… lab testing … VDSL2 will augment ADSL2+ over time …. within 1km”
Note the prevalence of fibre in these “solutions”. Note the small size of many (not all). Note that many seem to have gone nowhere. Note that most speeds offered are in the region of 20Mbps (when they’re offered at all and not simply promised). Note the short distances. Note the price is circa $80/month.
This just doesn’t look like a technology that’s going gangbusters. It looks like something people are experimenting with, delaying and then backing away from. Slovenia look to be about the most enthusiastic.
I wouldn’t recommend it to my clients (unless it was Telstra – “yeah guys, go for it, extend the monopoly”)
How much should we spend? That’s a legitimate question. But as I said before, I don’t think we should kid ourselves that a dead-end technology (remember the laws of physics here) of dubious viability is an option. Particularly when we’ve got a really good one (fibre) that will definitely last us for another century and we don’t have to put through hoops and handwaving to get it rolled out.
Copper was the high-speed physical transport of its day, but its day is finished. Fibre’s the new one. Copper got 100 years to depreciate itself, I see no reason why fibre won’t have a century to itself either.
“I made clear that DSL technology is not going anywhere for decades.”
I think that’s my point. It’s not going anywhere, ever. It’s at the end of its run.
“Telstra’s copper network is not being replaced, it’s being circumvented.”
Yes. And I don’t think that’s a bad thing. Is the copper network sacred? I know Telstra think it is, but I don’t have to agree with them.
* The “IT projects always run over” line is bollocks in this context. Rolling out fibre is a basic construction task – just dig a lot of holes – so the risks are nowhere near as great as a greenfields development project where the coders don’t know quite what they’re doing at the start. Construction projects are pretty tightly controlled and easy to estimate by comparison.
“The fact is we have an estimate today of 42B in 2009 dollars and no-one has seriously challenged that”
I agree, and I’m not challenging it either. Just interested that you appeared to agree with Andrew’s $100 billion, but that didn’t affect your views about the project’s worthiness.
“If 90-98% of people want something then treating it as national infrastructure sounds ok to me.”
You’re radically confusing ‘having access to something’, with ‘wanting something (enough to pay for it)’.
“2. spend 42B all of which will refresh the infrastructure, massively improve service levels and reduce costs to end-users if only because the monopoly can be broken”
It’s worth remembering that the Government is not planning on footing the entire (“up to”) $43 billion we’re all discussing. I’d be very interested in knowing how much private capital they’ve estimated will be injected. Any guesses on a realistic (achievable) figure?
“but that didn’t affect your views about the project’s worthiness.”
No, but that’s because copper really is at the end of the road and I personally believe there are benefits in better infrastructure, but as I said it’s a debatable point. We can be like Portugal who dragged their feet initially and then stopped their rail infrastructure with a few privately funded low speed lines that don’t even form a complete network somewhere circa 1930 and paid the price in lost opportunity for decades afterwards, or we can be like Germany who went gangbusters on rail right from the start several decades earlier. Different outcomes for different investments.
“You’re radically confusing ‘having access to something’, with ‘wanting something (enough to pay for it)’.”
No I’m not. We’re talking about networks, how they get rolled out and what benefits accrue to the economy as a whole. If we go the Portugal route and treat the network as a premium service to be paid for peicemeal by private investment we are very likely to get a similar low grade result. If we follow Germany’s example we’re more likely to get pretty rapid price drops, rapid take-up and much, much more development as a result.
The main value of a network only appears when nearly everyone is connected. This is one version of Adam Smith’s invisible hand (again this is an eccentric view and Andrew can weigh in if he wishes) – network effects pay off, big time, because they allow specialization and the emergence of services that are otherwise impossible if the potential market is limited to only a few.
“I’d be very interested in knowing how much private capital they’ve estimated will be injected. Any guesses on a realistic (achievable) figure?”
No idea and apart from my little contretemps with Andrew earlier re. capture of benefits I don’t really care. Someone will be spending it and someone will have to get the return. But the return will be more certain and faster with near-universality.
Though I didn’t attend it myself, apparently this was very interesting, and certainly sounds pertinent to this debate.
A fascinating discussion in the comments, nearly better than the originating article. Yes I confess to reading the whole lot. Yes I should have been doing something else, ah the internet.
However, case in point, I have spent considerable time reading this. I didn’t need even 56kbaud to do that. You only need large bandwidth and speed to access large bandwidth media such as video, online gaming (same thing) etc. While these things are good for an entertainment point of view they are not, in my view, essential for business.
There is a limited application for video conferencing but phone calls work good too. I can’t see actually being able to get a specialist online to do a medical teleconference, while it would be possible, getting the appointment would take 6 months just like it does to get to see one in 3D. I have actually been involved in a teleconference with a medical professional about 15 years ago so we don’t need fiber to do that either.
At home we already have fiber to the node, and are limited to ADSL, we cannot get ADSL2 or above because of the node limitations so I’m in agreement that to progress we need to go to every home.
The questions I have are related to the cost for access and the cost to connect. You will not be able to use your current house wiring for maximum speed because you run into the same limitations that have been discussed with the copper network. I think this means that everyone who wants this will need a new cable modem and then tapped into your home network. Cat 5 or Cat 6 would be the go but few homes will have that in. Wireless has the bandwidth limitations also previously discussed.
With the proposal to connect to 90% of the population this really limits the NBN to only the major centers on the East Coast. Since the plan is to provide wireless or satellite everywhere else the claim that this will bring Rural Australia into the 21st Century is possibly misguided.
In the process, by bypassing the Telstra network and coupled with the increasing use of VoIP, this has the distinct possibility of shutting down the copper network completely which will shut down communications in the Country unless they sign up to the NBN.
I am keen to get faster access but not at any price. I can’t see the NBN coming in on budget, for historical reasons just look at the budget overruns on the Building Education Revolution and the Computers in Schools programs. No different on any Government project that I have ever seen.
I understand your argument re this is a construction project rather than a technology project but, since I am an estimator for capital construction projects, I can assure you that cost blowouts are a fact of life in any construction project. If you go underground you have problems with rock, swamp and sand and all that other infrastructure already in the ground to get around and through (water, sewage, gas, electricity, telecoms, fiber etc.). If you go overhead you either have to put up your own poles or rent someone elses’.
And then there’s the maintenance of the network. Cables will get broken overhead and underground, overhead will have to be shifted constantly as the host poles get moved, upgraded or replaced because they’ve been broken.
Logistically it can all be managed but there is far more to this than just plugging in some new fiber and I can’t see any of this being considered in any of the statements from the Rudd team.
I watch with interest.
Its important to understand the $43b is the “worst case” exposure by the taxpayer. The government will only wear 100% of that if there is NO private investment interest in the project, and it is now looking increasingly as though there will be very substantial such investment – from Telstra for one.
I think people miss the point with widely available ultra-high speed braodband. Such a beast is not mainly about entertainment, but business. With routine high quality video conferencing and real-time secure document transfer in every home telecommuting will become very widespread (you’ll probably get more carbon abatement from that alone than you would from spending the money on renewables). Business travel will sharply decline. And, as Huggybunny said, you can bet that whole new business models will emerge.
Can we also add “education” to the list of advantages (entertainment and business) that could arise and evolve from ultra high-speed broadband? We should be able to look forward to the expansion of educational opportunities for all, that can come from the direct delivery of high quality (and if necessary government regulated, to weed out the nutters) knowledge services, to children and adults…something television could have started doing back in the fifties, but got side-tracked into entertainment alone, and then went rapidly down the sewer into the dross we have now… hope lives again.
I’m with Derrida Derider on this one as a lot the cable for the NBN has previously been laid, it just isn’t active. It needs to be activated, so even that should bring the price under $43B.
Also wholeheartedly agree with telecommuting.
If we all contacted our MP’s and demanded that the USO (Universal Service Obligation) http://www.acma.gov.au/scripts/nc.dll?WEB/STANDARD/1001/pc=PC_2491 , that the Government imposes on Telstra, be lifted from 31.2 kbps (2400 baud). This would MAKE Telstra via legislation remove the pair gain and RIM technology. (Telstra quote it will cost them $2 Billion to remove pair gain.. RIMs? much much more..) DSL will now be available to EVERYONE! (at Telstra’s expense.. Well.. Business is business.. and Telstra have it coming..) Mwaaaah ha haaa (Evil laugh)