back to article 'Serious flaws in the Vertigan report' says broadband boffin

Rod Tucker, a laureate emeritus professor at the University of Melbourne, IEEE fellow, and networking researcher has posted an article that offers very heavy criticism of the Vertigan Review's cost-benefit analysis of Australia's National Broadband Network (NBN). Tucker's beef is with analysis provided by Communications …

  1. Paul J Turner

    BS alternatives

    G.fast, reaches high speeds on <250m loops. Not much good for those of us 5km+ from the exchange, unless you think that using this for the copper section of FTTN will somehow magically increase the pitiful provisioning to the node that seems planned.

    HFC is just FTTN strung on poles, how long are the copper sections from node to house and how does that copper used to connect dwellings compare to ancient telephone pairs?

    FTTH has always been the only sane solution with any prospect of satisfying today's requirements for all users (not just city dwellers), let alone tomorrows.

    1. dan1980

      Re: BS alternatives

      @Paul J. Turner

      "FTTH has always been the only sane solution with any prospect of satisfying today's requirements for all users (not just city dwellers), let alone tomorrows."

      I don't agree with the first part. I believe that a combination of technologies can work for today's requirements. And yes, that will include FTTP. If you are saying that we must choose one and only one technology and that single infrastructure must work for everyone then yes, only FTTP/H will do. But, thankfully, there is no need to limit ourselves to a single technology.

      For the future, however, I most emphatically agree - FTTP is the only option currently available and one that can be upgraded as time goes on without having to dig up and re-lay things.

      The simple fact is that the Coalition don't like the idea of publicly-owned infrastructure. This is almost a basic tenet of the party. Whether that is good or bad is another question but the NBN is BIG public infrastructure project and I believe that Tony Abbott would much rather it didn't exist at all.

      A good communications network is as vital for commerce today as a good transport network and any cost-benefit analysis really has to look at the growth and development of the country as we move to a larger population over a wider area. Right now, we still have large business centres surrounded by (increasingly dense) widening urban sprawl.

      We need more private investment in more areas, including 'suburban' and regional, so that there is work closer to home for people. I know for a fact that some smaller companies choose where to relocate based on the available Internet connectivity - the last thing you want is to move your business somewhere suburban to find out that there are no free ADSL2 ports in the exchange. That's just unworkable.

      1. Thorne

        Re: BS alternatives

        "I don't agree with the first part. I believe that a combination of technologies can work for today's requirements. And yes, that will include FTTP. If you are saying that we must choose one and only one technology and that single infrastructure must work for everyone then yes, only FTTP/H will do. But, thankfully, there is no need to limit ourselves to a single technology."

        The flaw in your argument is putting in infrastructure for a system that is already outdated and then having to rip it out and start again.

        There is a use for wireless and a use for satellite both which was included in the original NBN plan. The only difference is the Libs have ripped out fibre to the home for fibre to the node which requires different infrastructure. At some point in the near future fibre will have to be run to the home which means every cent spent on fibre to the node is wasted.

        The Vertigan report is a BS political spin report to hide the fact that fibre to the node is already outdated before they even try to start building it. Telling everyone that we'll only need 15mbs when the NBN's own sales figures show it to be utter rubbish, is just a Liberal lie to hide the fact that they cocked up.

        1. dan1980

          Re: BS alternatives

          I was simply commenting on today's requirements.

          I agree that FTTN will be outdated very quickly. but the technology is able to cater for todays requirements. Whether the specific plan (if there is one) and implementation by the Coalition will achieve that and for how long is a different question.

          I absolutely agree that FTTP is the only real, long-term solution. It is unsurprising, then, that the government (without caring which side) is taking a short-term approach. What is amazing is that we actually got such a forward-thinking project in the first place, though it was ballsed up through incompetent planning and execution, thanks to leadership from two self-aggrandising PMs and management by the (then) stupidest man in politics, Stephen Conroy. So that, at least was comfortingly familiar.

          What we needed was a thoroughly-planned and honestly-costed project, headed by competent leaders. Instead we got a serial debacle that continues to this day and beyond. Is anyone surprised?

          1. P. Lee

            Re: BS alternatives

            My understanding is that today's requirements are not just limited to speed requirements.

            There is also a requirement to replace the existing aging copper in many places. So, if we are going to dig it all up, FTTP is a far better replacement. What we don't want is to put in more copper, which basically benefits Telstra by keeping multiple tiers of infrastructure as the norm and provides for market segmentation.

            As with all large infrastructure projects, FTTP is likely to cost more initially, but there are "network effects" benefits from having fast internet which makes it worthwhile. I have at least two relatives in urban and suburban Melbourne who can't get wired internet at all because "the exchange is full." Wireless is rubbish, made worse by colourbond roofing. Someone needs a good kick up the rear end to put new infrastructure in. Telstra isn't doing it, that's for sure.

            1. dan1980

              Re: BS alternatives

              All I can say is that I agree but we need to separate technology from implementation when discussing this.

              I am all for FTTP - not just as a future solution, but a practical one right now.

          2. Abel Adamski

            Re: BS alternatives

            @dan1980

            Guess your post demonstrates the truism of the meme, repeat a lie enough it becomes the "truth".

            It was far from incompetent planning and execution, in fact the development was based on the Germanic methodology of first attend to the foundations and core elements, establish initial builds and continuously evaluate and improve, slow to start, but once best practice and methodolgy established it ramps up very rapidly.

            http://www.theage.com.au/it-pro/government-it/nbn-fibre-rollout-was-going-to-be-cheaper-sooner-pilot-results-show-20140905-10cgdg.html

            to quote

            "The pilot took into account design changes formulated by network builder, NBN Co, last year as then chief executive Mike Quigley undertook a substantial review of the project and identified initiatives to reduce its cost and length."

            "The results confirm Labor's NBN was improving in the lead-up to the election, a point Mr Quigley had pressed in a speech to industry group TelSoc in December.

            He said the NBN was "knocking over the problems one by one and ramping itself into a scale rollout. And it was doing this without allowing any large increases in costs."

            and to reiterate

            "And it was doing this without allowing any large increases in costs"

    2. Fluffy Bunny
      Coffee/keyboard

      Re: BS alternatives

      FTTH is the only sane solution, as long as somebody else is paying for it.

      1. Abel Adamski

        Re: BS alternatives

        Fluffy

        And who may I ask do you think is paying for it.?

        Are you just another mushroom ?

        The ISP'S are paying for it directly and the customer is paying the isp, the higher value customers subsidise the lower value plans. FTTP NBN actually had extremely good take up of 50/100 Mb plans (many for guaranteed upload capability - cloud backup/storage - cloud programs, Video security allowing multiple cameras AND a HD pan zoom and tilt camera wit low compression [max detail])

        Consider even the Questionable NBN Review identified that including Operational and Maintenance costs (with fudged figures and excluding HFC/Copper costs) by 2027 both multimix and FTTP would cost the same, with multimix costing more and more as the years progress.

        FOD . Only just a fibre port off the FTTN module, not GPON. Considering equal costs by 2027, why demand extra cost to the Nation for all the extra FOD's when un necessary under FTTP

        The taxpayer is not paying for the FTTP option, all loans and interest will be paid off and a return generated. This will be less likely regardless of the CBA rhetoric, done by private sector, guaranteed large Taxpayer subsidies.

        The US shows the actual realities. Google only partially connects cities, the Cable Companies are rip offs with a disgusting reputation. Competition is an illusion with incumbents lobbying for and achieving legislation blocking competition and broadband being basically pockets of mini monopolies

  2. Shadow Systems

    The USA is currently enjoying WHAT?

    He claims "... average download and upload speeds in 2023 of around 34 Mbps and 8.5 Mbps, respectively. This is about the same as in the USA today." and I'd like to know where the HELL he's getting those figures from.

    The Federal Communications Commission (FCC) Chairman Mr. Wheeler cited figures that put the average American's "broadband" at less than *Ten* MB/s, and near zero for anything higher. Sure there's places where Fiber is available, but those are statistical blips in a flatline of capability.

    As anecdotal evidence, I'm currently stuck at *Three* MB/s because my ISP (ComCrap) can't be arsed to improve their lines in the area to sustain their vaunted "50MB/s Blast!" service tier. They tried to give that tier to me, and I had to go back to my old/current tier because their own modem kept dropping signal from Ping time outs. Why was it dropping? Because the line would saturate, the modem take a dump, and have to reset itself before reestablishing a connection. Average bandwidth of under *ONE* MB/s because most of the time it was resetting rather than connected. Go back to 3MB/s and suddenly I have a sustained & stable internet connection.

    So the claim that us American's are currently enjoying an average of 35MB/s? I'm not sure where he's getting that figure, unless he's cherry picking his data, because nobody in this area of town can get anything more than a *tenth* that speed without paying extortionate rates for the "priveledge". And I'm right outside "Silicon Valley" where bandwidth is as (supposedly) available as hookers & beer. Damn it, you can keep the beer, but I want my bandwidth! (I'm too close to San Fransisco to trust the hookers.) *Cough* I'll get my hat...

    1. bi-ker-shi

      Re: The USA is currently enjoying WHAT?

      The USA has its National Broadband Plan, we have the NBN. Under the plan private companies like Google are currently rolling out to a few places. Google is in fact rolling out to Kansas and Missouri. They held a competition to see what cities would get it first. Those cities had to make a contribution.

      As far as I am concerned that is infinitely fairer than a bunch of politicians and their servants in a private locked room making the decisions.

      Yes there are people very hungry for broadband. But what you must realise is that the faster you roll it out the more it will cost per unit. So you slow it down a bit and now you are favouring one electorate over another, or you try to speed it up as happened here. That policy proved to be a vote loser. For every satisfied voter there were at least two that were pissed off either because they were no where on the schedule or because the economy turned south and they feared losing their jobs.

    2. Rod Tucker

      Re: The USA is currently enjoying WHAT?

      The original article in the Conversation had a link for that data, but the Register left it off. The data comes from Ookla: http://www.netindex.com/download/

      There are a number of organizations that report broadband speeds by country. I chose to include Ookla data because that's what the Vertigan report used. One of the points I make in the article is that Vertigan is not consistent.

  3. dan1980

    There is one thing that should be evident to everyone and that is that reports and studies and analyses commissioned by a group with a vested interest in one particular outcome will almost inevitably support that desired outcome.

    Certainly if you have a political party that loudly criticised one course of action and campaigned on a competing option, any report they commission is certain to - wonder of wonders - agree with them.

    Rudd not doing more planning was a big mistake. But one must ask the coalition why they commissioned their report after they had already decided what to do and started doing it. What if the cost-benefit-analysis had favoured the previous, FTTP NBN or showed problems with the FTTN vision of the Coalition?

    Would Turnbull have realeased with a big mea culpa, explained that they proceeded in the way they thought best but admit that they were wrong and go back to the original plan?

    Pig's ear.

    I am not saying that an honest report would have said that, of course. There is bias both ways and this shows in the assumptions and scope of the studies/criticisms/replies/statistics/etc...

  4. mathew42

    NBN for average or high end users?

    The fundamental question that requires answering is should the government be providing an NBN service that meets the average needs or high end users?

    If high end users expect the government to subsidise their connections then they should be advocate for everyone in Australia to receive the same speeds on fibre and usage based charging.

    1. Thorne

      Re: NBN for average or high end users?

      Maybe your question would be better off as NBN for high end users or NBN for high end users that are less than 300M away from box.

      High end users pay more for their service thus subsidising the low end users. It's exactly the same effort to install and exactly the same hardware.

      At some point fibre needs to run to the house and personally I'd rather see it done once properly than spending almost as much on a second rate system that will need to be ripped out and done again properly not that far down the track.

    2. dan1980

      Re: NBN for average or high end users?

      @mathew42

      Ahhh, but that ignores the fact that a patchwork infrastructure like this privileges some more than others.

      The point is the delivery of the underlying infrastructure and the question is whether we should be deploying an infrastructure that can accommodate both average and high-end users. This becomes more relevant when you realise that even low-end users today would be considered high-end users only a few years ago.

      The big problem with Australia is we are nearly totally reliant on copper and thus DSL. For most, that means we are limited to 20/1.5 Mbps on ADSL2. That's a limit - in practice it drops considerably depending on distance from the exchange. Sure, there's VDSL but this is still in its early days and there are relatively few vendors, which means that hardware costs are high and compatibility is low. It also suffers from exactly the same attenuation problem as ADSL/2 but falls off much more sharply.

      And this is the problem with copper - you create a tiered network simply by virtue of exchange placement. In such a system, someone getting 5 Mbps will be paying the same as someone a kilometer away getting 20 Mbps. You could say that low-end (not by choice!) users are subsidising high-end users.

      And that's the beauty of fibre - once it's in place, you can charge people for what you are actually providing. Yes, it is expensive to put in and sure, it might take longer than a less capable solution but it's the only option that is fair and future-proof (so far as is possible).

    3. Fluffy Bunny
      Facepalm

      Re: NBN for average or high end users?

      "should the government be providing an NBN service that meets the average needs or high end users"

      Your question is completely wrong. Why would any competent government provide NBN services for anybody? The best role of government is to govern. That is, it sets the rules around how corporations work and creates an incentive to roll out the infrastructure that it wants.

      When the US government of the day wanted to connect the East and West coasts, it didn't start by creating a monopoly. It subsidized the laying of track. As a result, two companies competed to do the track work - twice the performance - and took all the risks.

      When the KRudd government wanted to buy a bunch of votes, it created a monster - took all the risks - and had no performance. We're just lucky the NBN didn't kill anybody like pink batts - although we're still waiting to hear how those workers that got exposed to asbestos are going.

      So to answer the question you should have asked "should an NBN service that meets average needs or high-end users be provided"... the answer is obviously, "whatever they want to pay for".

      1. dan1980

        Re: NBN for average or high end users?

        @Fluffy Bunny

        "The best role of government is to govern. That is, it sets the rules around how corporations work and creates an incentive to roll out the infrastructure that it wants."

        This is simply an ideological position. As a general rule, left-leaning people and governments believe in public infrastructure and services that are subsidised and provided to all on a relatively equal basis. Right-leaning people and governments believe in privatisation and believe that provision of services should be on a strictly 'user-pays' basis.

        I don't want to get into an ideological argument because that would be pointless; you and I both know where each other sits so we can agree to disagree.

        I would only say that the government is already in the business of providing subsidised public services and infrastructure such as education and health. I believe that this is right because having a smart, healthy population is important for our prosperity and our general happiness. Governments also provide and subsidise transport, which again makes sense because a working transport infrastructure is crucial to the economy. Well, I believe that a decent internet infrastructure is becoming just as important as a decent transport infrastructure so I think that, just as it is a good thing for the government to enable a healthy, educated and mobile population, so too is it a good thing for the government to enable that population to be well-connected.

        But that is my position and neither that nor your position really relate to the thrust of the article, which is that this analysis questions the assumptions and projections used in the Vertigan report. That report was not about whether government should or should not provide a national broadband network, nor was it whether a 'user pays' system is best. It was looking at the capability of the new, FTTN NBN to meet current and future requirements.

      2. Abel Adamski

        Re: NBN for average or high end users?

        @ Fluffy Bunny

        Hi Mushroom

        There has been an identified need for a National Broadband Infrastructure for decades, just the only way anything but high value areas would be covered is by taxpayer subsidies.

        The private sector failed to deliver in fixed broadband Nationally

        The Nationals even produced a National FTTP NBN plan very similar to the Labor FTTP prior to the 2007 election, why they called the initial Labor FTTN plan "Fraudband" (Fiona Nash website).

        No performance.

        Ascertain the facts.

        The current Malcolm's mates crew can look good because the core infrastructure and systems have been built, you know the foundations and framework.

        Major delays are due to Telstra's tardy and patchy remediation, causing delays and economic pressures on Contractors and Sub Contractors as well as questions on contractors project management skills as highlighted by subbie's comments re the people they dealt with which were printed in News Ltd, edited once they realised it was Contractors failing not NBN - good old internet I have a PDF copy of both versions.

        It is a massive task, which we all recognise

        Asbestos.

        That always has been Telstra's responsibility, it had been a issue for over a year which had been reported in the media and was blocking the rollout in some areas. Yet Telstra was ahead of schedule installing backhaul and transit links needed for both options

        Strange it suddenly became an issue plastered over the media just as the install was ramping up just before the elections becoming a political media issue.

        Telstra has financially benefited and now has greater control, also benefiting their Foxtel/News Ltd Partners.

        "So to answer the question you should have asked "should an NBN service that meets average needs or high-end users be provided"... the answer is obviously, "whatever they want to pay for"."

        So depends on what is actually practically/technically possible.

        Depends on whether you view the NBN as Essential National Infrastructure or as a plaything for the jaded rich, as a for the immediate future or the next 50+ years

        Your call

        1. dan1980

          Re: NBN for average or high end users?

          @ Abel Adamski

          "The private sector failed to deliver in fixed broadband Nationally."

          No argument that the private sector hasn't delivered a national fibre network but I am not to keen on throwing around the word 'failed'.

          There's just no way a private company could afford it - big projects like this just have to be government projects. Further, the profit margin wasn't there. The projection was always for this to make money for Australia but it wouldn't have made enough to keep a company's share-holders happy.

          I think the projected figure was around 5-7%.

          And that's the fundamental flaw in the standard right-wing position of privatisation and 'market forces'. The problem is that the Government can create services that run at a small profit and thus deliver better value. They can also work on larger timescales compared to public companies being watched like hawks for growth each and every quarter.

          One of the best things about the Labor FTTP plan was that it would have greatly minimised the Telstra monopoly. That did mean greater upfront costs as there would be new ducting laid but the end result would have been a cheaper system - ongoing. Compare to MalcolmNET, which fills Telstra coffers and gives us 70,000 great, ugly, hulking active nodes full of DSLAM equipments that needs maintaining.

    4. Adam 1

      Re: NBN for average or high end users?

      >should the government be providing an NBN service that meets the average needs or high end users

      Should the government be providing a health service that meets the needs of the average patient or high dependency patients? We could have an orders of magnitude cheaper healthcare system if we had the same attitude.

      Should our public transport system cope with peak hour rush or average load?

      Should our power grid cope with the hot summer peaks when everyone is running air conditioners and pool pumps or just average draw?

      I agree in principle that there is a point where something becomes unviable, or a point of diminishing returns if you like, but the real problem with FTTN is that we are taking about a capacity date in the next decade unless you think that trends will change. That is big coin for infrastructure that will require reinvestment as soon as it finishes. If it were 20-25% of the cost of FTTP then I would certainly be torn, but it is well over half the build cost anyway. More than that, it still relies on the flaky copper from the node to the house. This means you are going to have to relay copper to the node which will need to get torn up again in relatively short time. It is only cheaper if you restrict consideration to the next decade. Longer term it is more expensive to operate, orders of magnitude more expensive to upgrade (at best you are driving around to every node and upgrading hardware there rather than at a central exchange)

  5. Anonymous Coward
    Anonymous Coward

    Show me the hardware.

    Excuse my ignorance. Everybody keeps saying, referring to "FTTN" things like "it'll have to be ripped out to do FTTP". I would presume that the same fiber would have to get to the same node, irrespective of FFTN or FTTP.

    For FTTN, an interface would be required to connect to the existing copper at the node. Job done.

    Lets say in 5 or 10 years that there's an upgrade for the area to FTTP, then fibers would have to be laid to each property and a new interface between the main fiber and the individual fibers. The old interface and the labor connecting it to the copper are the only thing replaced / wasted. And amortized over 5 ~ 10 years. Big deal.

    Spend the money not wasted on FTTP on fixing the shitty areas first. (ones with full exchanges or long runs etc.)

    If the copper is cactus. Replace with fiber. (Everybody loves to add throwaway lines (no pun intended) about ageing copper. old tech yadda yadda yadda. If it works. use it. if not replace it. )

    And that's presuming that in 5~10 years that: a) We want or need faster than that. b) There isn't some other technology.

    Bottom line. I don't care how the data gets here. Spend my taxpayer money wisely.

    1. Charles 9

      Re: Show me the hardware.

      "Excuse my ignorance. Everybody keeps saying, referring to "FTTN" things like "it'll have to be ripped out to do FTTP". I would presume that the same fiber would have to get to the same node, irrespective of FFTN or FTTP."

      The problem is that FTTP requires a completely different topology from FTTN, meaning most FTTN equipment can't be used in FTTP. If you use FTTP now, then in five or ten years time when more bandwidth is needed, it's MUCH easier (and less expensive) to build on an existing setup than it is tearing up the FTTN setup to replace it with a FTTP one. IOW, FTTP has a higher up-front cost but is more future-resistant.

      1. Dr Project

        Re: Show me the hardware.

        Can you post some links? I'd like to see the difference.

      2. Fluffy Bunny
        Holmes

        Re: Show me the hardware.

        "most FTTN equipment can't be used in FTTP"

        That's just rubbish. The main cost is in the cable - whether copper or fibre. Yes, fibre costs more than copper per metre, but the real cost is in laying it. Digging trenches, running conduit and filling it with the cable.

        Ultimately, all it is, is a means of getting bits from one place to another. The protocols used depend on the equipment you connect into it, which are largely standardized worldwide. Certainly the topology is no different.

        The only difference between FTTN and FTTP (equipment wise), is that there is a converter from Ethernet over fibre to Ethernet over copper (and there are several to choose from).

        Going back to the original point, that equipment will all need to be replaced in 5-10 years anyway and so is money we would already have to spend. If you choose to replace the cable at the same time, fine. But don't claim we need to do it now when we don't.

        One final point. People keep talking about aging copper, as if it's bad. If the copper really needs to be replaced, then it is either because it's due for normal maintenance work, or it wasn't maintained properly. I'll take that silence as an answer, shall I?

        1. dan1980

          Re: Show me the hardware.

          @Fluffy Bunny

          FTTN is active, whereas FTTP is passive.

          Can you reuse the power cabling, batteries and cooling that each active FTTN node needs? No, because FTTP uses passive 'nodes' and so doesn't need all that. Can you reuse the fibre ports in the Exchange? No, because it won't work with a PON. The VDSL hardware in the nodes? Nope? The customer 'modems'? Nope.

          So what equipment, exactly, can be reused?

          1. dan1980

            Re: Show me the hardware.

            @Fluffy Bunny

            I'll take that silence as an answer, shall I?

            : )

        2. This post has been deleted by its author

        3. dan1980

          Re: Show me the hardware.

          @Fluffy Bunny

          "The only difference between FTTN and FTTP (equipment wise), is that there is a converter from Ethernet over fibre to Ethernet over copper (and there are several to choose from)."

          Sounds insignificant when presented like that but it's only when you see images that you realise just what this difference means.

          Here is a FTTN cabinet. The unlabelled mass in the bottom-left are batteries, I believe. Here's another, smaller unit from Telstra, with parts hidden behind panels but the battery packs clearly visible.

          Here are a few close-up views of the kit required.

          In contrast, these are some FTTP cabinets. Here's another, and here's one partially-populated, showing how you can start with a bit and add more. The little black box in the bottom-left is the (passive) splitter.

          Here's one strapped to a pole (that tiny box about 2/3rds of the way up on the left is the splitter) and here's it's bigger, ground-dwelling brother, missing the splitters and the cables.

          Can you show me where you would reuse all the FTTN equipment in here?

        4. Thorne

          Re: Show me the hardware.

          "One final point. People keep talking about aging copper, as if it's bad. If the copper really needs to be replaced, then it is either because it's due for normal maintenance work, or it wasn't maintained properly. I'll take that silence as an answer, shall I?"

          The problem is that Telstra has to legally supply you with a phone line capable of phone calls. Broadband isn't covered and they are not going to replace anything that they aren't forced to.

          1. dan1980

            Re: Show me the hardware.

            Just to clarify Thorne's very valid point (as it took me a couple of reads) . . .

            DSL and PSTN services run over the same sets of copper twisted pair telephone lines that run from your location back to the exchange.

            The difference is that DSL services are more finicky than PSTN services and so require a better connection. What this means in practice is that while the twisted pair cabling running to your building will be fine for making telephone calls, it may be inadequate for a DSL service.

            Going into more detail, this is generally due to noise on the line, which is, in turn, usually due to bad connections and taps or just poor quality copper.

            If you think of PSTN calls as just data (which they are) then they are a much lower data rate than DSL services. In increasing rates, it would go: PSTN < ADSL < ADSL2 < ADSL2+.

            As with ethernet networking, higher data rates (10Mb, 100MB, 1000MB) require better cables and this requirement is increased as distance increases. In other words, the higher the data rate and the longer the signal is travelling, the better the cable needs to be. In networking, CAT5 will work fine for 100Mbps and will likely serve for gigabit over short differences if the cable is good and interference low. Over longer distances, however, CAT5 will likely be unable to support 1000Mbps. CAT5e will.

            People with noisy lines are sometimes limited to ADSL (1) or, if they can get ADSL2/2+, then it is likely to be effectively limited to near-ADSL speeds.

            Thorne's point was that such a situation is of no concern to Telstra as they only need to have cabling good enough for the low-data rate PSTN services.

            Directly countering Fluffy Bunny's point, it is perfectly possible for the copper to be fully within spec so far as Telstra is concerned (and thus not in need of maintenance or replacement) and yet be unsuitable for speedy or reliable Internet access.

    2. dan1980

      Re: Show me the hardware.

      @AC

      Here's the deal . . .

      FTTN means fibre from the exchange to multiple streetside nodes, with each node then connected to a bunch of subscribers - 100-300 - via a non-fibre connection such as twisted-pair or coax. I could be wrong on the numbers but it's not really important for the moment. This much is largely understood.

      What is less-widely understood is what FTTP entails. There are several different ways to do this and many people believe that there would have been fibre directly from the Exchange to each and every home. Not so. While this is an option, the old NBN plan was for a GPON - Gigabit Passive Optical Network.

      This involves a passive 'splitter' that splits the optical signal out to multiple 'last mile' fibres connected to the end points. If one were to draw FTTN vs FTTP (GPON) then they would look similar on the surface.

      There are two big differences, however, which is that a GPON FTTP splitter services fewer endpoints (32 is the standard) and is passive. This is a big advantage as all the active electronics in a FTTN node makes them more expensive to build and maintain.

      As regards the first difference - the number of endpoints - this is the number per splitter and a node may contain multiple splitters. This is a flexible situation. You can drop a (very small) 32-node splitter in an existing pit without any problems - they are small and passive. Or, you can set up a cabinet and add as many splitters as you like.

      Here is an image of a single, 32-port splitter, of the type that could be dropped in a pit. (In a suitable enclosure.)

      Larger, more manageable units resemble network switches and these are installed in (usually above-ground) cabinets. As there is no real maintenance overhead nor distance restriction on the 'last mile', deployment is very flexible, so one can have small, medium or large cabinets depending on requirements. You can strap 'em to telephone poles if you like.

      This is in contrast to the FTTN nodes, which require complicated, active electronics to convert the incoming optical signal into an electrical one and off to VDSL (essentially making each node an exchange) or coax. For these units to perform well using VDSL, which is the main push, they must be no more than around 400m from any given endpoint, due to the attenuation of copper and the subsequent sharp fall-off of VDSL speeds*.

      So, what happens if one wants to transition from a FTTN to a GPON FTTP?

      The biggest issue is that you no longer need all those active nodes. You still need somewhere to house the GPON splitters, of course, so the node cabinets could be re-used, but the upgrade would have to happen in parallel sou you'd need enough spare room in there. If you have a look at a FTTN node, you see that there isn't all that much room, though the small size of the GPON splitters and the required close-spacing of FTTN cabinets to achieve decent speeds means that you wouldn't need to put too many splitters in each cabinet.

      BUT, remember that the ratio is 1:32 for GPON, meaning 1 incoming fibre from the exchange split to 32 endpoints. So, you need to lay more fibre loops to each of the nodes. And, as these nodes are numerous, that's lots of separate runs of small bundles, rather than one run of a larger bundle if you build a new GPON node. But again, that's a new node - new infrastructure!

      But, once you've cut-over, what do you do with all that obsolete hardware? You've got heaps these huge cabinets - far larger and more numerous than are now needed - either filled with quiescent electronics or large empty spaces. That's a wasted investement - it serves no purpose now.

      BUT, it's not just the hardware in the nodes that needs replacing - you also have to replace everyone's (now obsolete) VDSL/coax 'modems'! And again, that old hardware is useless now.

      But, there's another problem, and that is whatever commercial arrangements have been made for third party ISPs, which will be delivering their offerings over combinations of twisted-pair and coax. I can't speak to those but you would imagine that it wouldn't be a smooth transition.

      And, of course, you still have to actually lay the fibre to the end points.

      * - Just as ADSL2 quickly falls to ADSL speeds, so too does VDSL fall to ADSL2 (and then to ADSL) speeds.

      1. Abel Adamski

        Re: Show me the hardware.

        @dan1980

        With respect

        "BUT, remember that the ratio is 1:32 for GPON, meaning 1 incoming fibre from the exchange split to 32 endpoints. So, you need to lay more fibre loops to each of the nodes. And, as these nodes are numerous, that's lots of separate runs of small bundles, rather than one run of a larger bundle if you build a new GPON node. But again, that's a new node - new infrastructure!"

        Labor FTTP NBN as you state uses 32 split nodes, they generally populate between 26-28 leaving spares for increased services (property development etc).

        However in fact x64/x128/256 splitters are available as are x8/x16.

        Knowing the practical incompetence of the current Government and it's politically and ideologically committed NBN management, I suspect they would take the cheap option of x256 split and proclaim how FTTP is of marginal improvement over FTTN/HFC rather than run in the extra fibre needed for a x32 split. With lower split ratios for business precincts, thus crippling innovation, small start up innovative companies work from home and can't afford the high business premises rent on top of everything else

        1. dan1980

          Re: Show me the hardware.

          @ Abel Adamski

          Yes - I should have been a bit clearer. I meant that in the context of the people asserting that the FTTN was able to be efficiently upgraded to the original FTTP spec. The reason being that some people are saying that the FTTN is clearly the best option because it can be a kind of middle-ground stepping stone to a full FTTP.

          In pointing out the 1:32 ratio of the GPON network, I was implying much the same as you are saying, which is that if you split the existing inbound fibre out to the same number of endpoints as each FTTN cabinet would have, then it wouldn't be the same as the original GPON FTTP plan as that is 1:32 and thus would require more inbound fibre.

          Again - should have been more specific.

  6. Medixstiff

    It's all moot either way really.

    After all, the Coalition are only paying until 2018, after that I expect they will sell it off similar to Telstra. So who really cares whether it's FTTP or FTTN, as after 2018 whoever is on NBN will be screwed by a monopoly owner, just like Telstra customers.

    I'll still be happy on my ADSL2+ iiNet connection that maxes out at 13 Meg thanks to the crappy copper in West Perth, as long as no-one takes iiNet over that is. I've never had a problem in nearly 12 years being with iiNet in multiple premises.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like