Many PoI NBN model “insane”, says Hackett


National broadband provider Internode has published a strongly worded argument in favour of NBN Co’s existing model for connecting internet service providers to its future fibre network, backing the incumbent minimalist structure which would see just 14 so-called Points of Interconnect located around the nation, and describing rival proposals as “insane”.

A number of other telcos — such as Telstra and Optus, for example — have argued strongly against the 14 PoI model, with Telstra stating NBN Co should be required to provide interconnection at any technical and operationally feasible location requested, while Optus has claimed the model will kill competition and open NBN Co up to compensation claims.

However, in a blog post published over the weekend, Internode managing director Simon Hackett said the larger telcos’ arguments that there should be several hundreds PoIs was “self-serving”.

“The clear choice of NBNCo (and smaller RSPs) is for there to be a small number of POI’s (circa 14) located in capital cities,” he wrote. “A small number of PoI’s in capital city locations means dramatically lower economic and technical barriers to entry and participation for all RSPs. It will lead to a larger number of more diverse service providers (and services) on the network.”

“The opposite view is held by the very largest RSP’s (Telstra and Optus especially), who have been lobbying for there to be a much larger number of PoI’s (circa 200). This is not surprising, as these are the players who own their own extensive fibre backbone networks that will let them plug into 200 POI’s without significant investment.”

If NBN Co provided a larger number of PoIs, Hackett wrote, smaller telcos would need to buy network access from companies like Telstra and Optus to connect to NBN Co’s network.

“The self-serving benefit of this for those larger players is obvious,” he added.

The Points of Interconnect debate is one issue likely to be raised today as the Federal Government releases NBN Co’s corporate plan. Communications Minister Stephen Conroy has called a press conference at midday to take questions about the document, and to make it available to the public.

Hackett noted he suspected the ongoing debate would result in a likely compromise announcement from the competition regulator and the government on the issue.

“This would represent a decision that can only be framed as being politically motivated — and the classic mistake in politics of choosing the average between two opposing views, instead of being brave enough to make an appropriate decision,” Hackett wrote. “It would ‘half fix’ the commercial drivers of the big guys, while ‘half hurting’ everyone else. Permanently. And in a way that may be impossible to fix later on.”

“It’s quite insane in my view for the NBNCo to spend $43 billion (or whatever it does cost) on the network, to bring it 99 percent of the way to where all the retail service providers are generally located (in capital cities), and to leave that last piece of fibre backhaul expenditure out there to have to be contended for in parallel by every retail service provider.”

Image credits: Internode


  1. I certainly favour a number in the lower end of the range…I’m not convinced that 14 is enough, though Simon does present a compelling argument.

    Certainly, a number in the order of 200 is patently ridiculous.

    The highest number I would be comfortable with is 66 – (or thereabouts) – which maps to the 66 main Call Collection Areas (CCAs), and should be roughly analogous with existing primary backhaul infrastructure.

    Going too far – as Simon correctly suggests – is potentially very dangerous.

  2. Got to wonder why these guys didn’t bother arguing a case weeks ago when ACCC was starting to consider this issue. Why come out only when the decision has been made? Seems a bit “self-serving,” don’t you think?

  3. As someone based in Tasmania, I don’t like the idea that all NBN traffic will be hubbed through a Melbourne POI.

    This could lead to significant performance issues if one of the cables between here and Melbourne is offline, and if there were to be a complete failure, local internet access would effectively be cut off – i.e. no ability to access government services, schools, UTas servers, etc.

    • A cable break between Melbourne and Tasmania would have exactly the same effect as a cable break between Melbourne and Mildura.

      And the exact same effect as if the cable break occurred before the NBN existed.

      Whether it is backhaul from a FSN in Tasmania, or open backhaul traffic, a break is a break. The only way to lessen the impact is to have more and more cables across Bass Strait.

      Or more between Melbourne and Mildura.

      It’s exactly the same.

      • Except that at the moment, there is a PIPE peering point in Hobart which means that not ALL intra-state traffic is routed via Melbourne… although a large amount of Telstra traffic seems to be.

        In any case, judging by today’s document release, there will at the very least be two PoI’s in the state.

  4. Cant we do both – a large number of interconnect points will assist the network by providing traffic diversity for Optus and Telstra customers, whereas Internode and smaller RSP’s will route their traffic through the NBN backhaul network to a convenient POI. If NBN backhaul charges are too high it will encourage smaller RSP’s to install their own fibre solutions.

    • The question over the number/location of POIs is more a matter of backhaul competition than a purely technical question.

      The basic argument is the more POIs you have, the more power to influence price you give to the larger players in the backhaul market, who are far more likely to be able to serve a higher number of POIs than a smaller player.

      Simon – (rightly) – argues that the premise of the NBN is to promote level playing field in the wholesale market, and by increasing the number of POIs, you swing the backhaul market heavily back towards the traditional “big boys” in Telstra, and Optus, and to a lesser extent AAPT and Nextgen Networks, and even PowerTel.

      This makes it harder to maintain a uniform pricing structure.

      But reducing the number of POIs – (as low as 14 as NBN Co wanted) – may “strand” huge chunks of infrastructure owned by the “big boys”, opening the government up to compensation for legislating existing investment out of viability.

      I’m surprised they’ve gone as high as 120, but at least it’s not 200.

  5. Am I right in saying if NBN Co have less POI’s (like 14 instead of 200) then the NBN will be more expensive to build as it now has to build more backhaul to connect these points itself?

    • Nope – backhaul is there either way.

      A POI is only a crossover point between the NBN and the external backhaul networks, nothing more. More POIs means backhaul traffic has to cover less distance to a POI.

      It is therefore cheaper for a service provider to access the NBN, because the backhaul from their network into the NBN (via a nearby POI) is over a shorter distance, and therefore cheaper.

      • Ok I think I understand.

        So having more POI’s would mean for example someone from a rural area who is connected to the NBN and opening a webpage to the UK would have to make less hops or travel less distance because there are more exits out of the NBN and onto say Pipe’s network to outside Australia?

        • Not really, and there is more to it as well.

          A big ISP – (and lets use Internode as the example) – may choose to interconnect with the NBN at 20 different POIs around the country. If a single POI goes down, it doesn’t disconnect them from all of their customers, only the ones connected to the failed POI.

          In an environment with a smaller number of POIs, a smaller ISP – (Uncle Bob’s Garage Based ISP for example) – would have to pay for backhaul over a longer (more expensive) distance to get to a POI, and he’ll probably only be able to afford to connect at one POI.

          If that POI fails, he is disconnected from all of his customers, no matter where they are in Australia.

          Increasing the number of POIs makes it easier and cheaper for “lesser” players to get access to a POI – (or multiple POIs), making his network (and business) more resilient.

          Each premise is connected to a single Fibre Service Node (FSN), and each FSN is connected to a single POI. If the POI your fibre joins the core network through goes down, you’re off the network – but that’s more or less the same as now. You can’t fix that unless you have two pieces of fibre to every premise, with each going to a different FSN and POI.

          That’s not going to happen.

          In logical terms, the design of the NBN is not that much different than an ADSL network, it just uses fibre for the last mile, instead of copper.

          The POI is an access level consideration, not a service level consideration.

          • I see, but doesn’t that still mean if there are less POI’s then NBN Co has to now spend more money on backhaul (or internal networking) because NBN Co now has to internally connect more premises up then merge to say a single POI in Canberra instead of what could have been any of the 20 POI’s in Canberra if the POI was 200 instead of 14?

            Just means NBN Co has to buy more cables to connect to the 14 POI’s rather than “cutting” the cables earlier to any of the 200 POI’s.

  6. “IF” the government suceeds in splitting telstra up then possibly that would slightly improve things with the more POI’s. However do we know everything that is included in the buyout from telstra? Proberly not. We are buying the copper we know that much but what else?

    also since they state it that all poi’s will be in areas served by 2 wholesalers or at least very easy to add a second wholesaler that should help with the ACCC controlling monopolistic pricing as they could charge 2 companies if they are not competing for being a “cartel” of sorts.

    The other possibly benefit is better ping times however it is fibre so i doubt it would have much impact on ping times.

Comments are closed.